💡┃suggestions-web

[formal bobcat]

When trying to organise my library, and adding workspaces, there should be an option to move things in bulk rather than one at a time. If the option already exists, finding it is impossible.

[scarlet cairn]

It is possible, right click after selecting multiple songs

[formal bobcat]

How can I select multiple songs?

[low hinge]

Use Kden live to make your own videos. Dont think i would ever use or have use for an srt file, but you can use Kden live to customize anything you want in a video.

[scarlet cairn]

Shift+click

[strange blade]

The Stem separation is after The song is generated. It's the same thing we can do with other software. They need to train the AI on separating the instruments as it's making the song like being tracked in a recording studio.Please we need this!!!!!!!

[elder talon]

please let us know why the lyrics dont work

my lyrics fr have 0 "bad words" and still not working

[blissful sail]

There’s a growing number of really amazing Norwegian song makers on Suno. They really deserve a Norwegian trending. More Norwegian music to the people please 🙏

[fiery wave]

Option to set 'new' via filter as default view @profile page Good sunday & new week after o/ - Subber -

[dull notch]

Daily credits for free account to create/remaster to V4.

[novel gate]

A creativity metric perhaps in a numerical or slider format, like midjourney has, for music and voice seperately. By that I mean, if it was set to 0 creativity for music, it strictly adheres to the Style of Music prompt and if set to 100, it can loosely adhere to the Style of Music prompt. Also if the creativity is set to 0 for the voice, it strictly adheres to the voice of the Persona, and being set to 100 it can alter the voice to something different from the persona within reason.

[half heath]

I’d love to have a bulk delete of songs rather than one at a time. Is that in the works? Thanks

[twin bear]

STOP THE ABUSE, I've HAD IT with all these people abusing the genres and repeatedly entering the genre type into the prompt field. It robs other user's of actually getting to enjoy the various genres in SUNO

I would suggest to suno DEVS to limit the genre type in the prompt field to 1 time or 1x

• Variation should still be allowed for example: Funk, Funk Pop, Soul Funk, etc

But, limit the number of times it can be repeated to just 1 time.

For the existing songs that have flooded the categories, I would suggest bumping them down to a least sorted category so that way it promotes more variation at the top

Please and Thank you

[peak knoll]

Something is really wrong with track edit/replacement. I typed out the lyric to be replaced but it made up gibberish lyrics out of thin air.

[spark egret]

I'm trying to make a persona from the audio I updated personally, but it doesn't turn grey. What's the problem?

be using a pro-plan

[polar gorge]

Create music from images and videos

[summer depot]

I am facing same issue

[azure plover]

Please change the crop tool to remove the selected section and not the selected section being the part you keep. If we need to remove a duplicate verse or just remove something in the middle of a song it's impossible to do that currently. Thank you.

[sturdy wave]

I thought I was going crazy looking for the link on Desktop. Why are Workspaces not easy to navigate to? Please add @fresh shard

never mind I think I figured it out

[tepid trench]

You're only supposed to upload audio that you own the rights to. Uploading soundtracks from anime episodes would be against the Terms of Service. You would have previously agreed to these terms before being allowed to upload audio:

[slow walrus]

I bet the GEMA did ignore it, did upload copyright and now pretends the AI does copyright songs

[twin bear]

Allow Personas in the Extensions

Very simple...

[wind raptor]

What if you could provide an instruction in the lyrics section using { }

i.e

[Chorus] {Loud, pounding drums. High guitar riffs, helicopter blades in the distance}

Insert lyrics here (Backgound Lyrics)

That could allow for user to input a descriptive direction to the music, further refining the user's vision

[twin bear]

SUNO's AI especially in V4 cannot maintain a consistent regional accent in English: Prompted for British English Accent, some words it gets the accent correct, but on others, it defaults to a generic American accent.

[novel gate]

I've had this exact issue as well.

[nimble shore]

Or add it as "Delete Section" option

[keen helm]

make ppl not get in trouble #💬┃general-chat message

✅

[upbeat mantle]

just a thing i wanted to have in the lyric box are a grammar correction system so that u now you´re spelling is correct before you make the new gens

[stark inlet]

What’s so difficult about bulk downloading

[sick dagger]

I have noticed in the Romanian language that if I use the word "cum", which means "how", it is also used in certain words that include "cum", such as "acum", which means "now". It's quite annoying to keep changing the lyrics when it's not even an obscene word. Try to optimize the AI for all languages.

[median wave]

try spelling it with a K. If that works and you get the sound you want, edit the displayed lyrics back to the Romanian spelling.

[sick dagger]

aight thanks gonna try it

[peak knoll]

Please fix the edit/replace feature. AI is not able to read the words I type in the lyric. It either ignores it or vocalizes gibberish.

[arctic python]

It would be ideal if we could choose a vocalist, or simply choose male or female as I have wasted literally hundreds of credits on a fully female song only for a male metal vocal to get into the song for no reason at all. And yes I have tried every possible conceivable way to get around this. Especially as the "exclude" box does the opposite and includes it

[radiant shell]

Seems like our profile image gets compressed quite a bit once uploaded. I can use high quality images for my song images, would be nice if we can upload a high quality image for our profile.

[radiant shell]

I'm sure you've tried this since you said you tried every possible conceivable way, but did you try something like this? [Verse 1: Male, double-time bounce]? Can even do something like [Intro: Soulful female vocal sample, warped synths, deep 808s pulsing]. Seems to work pretty consistently for me. Not perfect, but pretty good.

[karmic sigil]

New Edit mode needs a "Reverse Extend" feature to make intros on songs that need a new one because Replace won't do it properly.

[nimble narwhal]

Suggestion: At least for paid credits code in a way to critic SUNO's takes for example, "Right Voice, Melody too soft, No distortion on guitar". That sort of thing. My hope is the Large Language Model will try to understand and implement the necessary changes.

[upbeat mantle]

One feature to improve is a small one but its needed a function to let u expand the Style box and see all the genres at once so you dont need to scroll and squint

[viral ember]

There will be or is a tool for that use case for premium users afaik.

[arctic python]

Yes I have added [Female vocal..etc] into the lyric sheet, next to the verses etc as well as the prompt and also the genre box, and somehow somewhere a male vocal comes in, but weirdly it's like metalcore male vocal often it's singing the last few lines, or randomly in the middle.. I found using a persona can get around this but that is quite limiting depending which song the persona is created from e.g. no intro and specific style different from the specified prompt

[bright cliff]

I have a suggestion : Using a search function on a profile, so you can find a song more easily on a suno channel.

This makes it easier to find songs of others with titles who are already taken.

When you search for a song of others in the song search function with a title who is used by others but higher in the algo you gotta need to scroll a lot, putting this search on a suno profile will save all the hassle

[vernal pagoda]

Please add an option for the creation of .lrc synchronized lyrics files

[floral wyvern]

I would love the ability to change (add, remove, modify) the "style" tags to better reflect the final song output, or a "tags" system untied from the styles for searching purposes.

This would greatly help those searching for the types of music we generate. In my instance, my music is a blend of genres, and there are combinations of style tags that I use to generate my sound that diverge from the final product.

I worry that while simple style tags such as "rock", work for searching purposes, it lacks the control over sound something like: "chill nu-jazz soul jam" has, yet is so specific that it cuts my song out from queries looking for simply: "jazz" or "soul" or "chill"

[unkempt berry]

Is there a paid subscription thinkable like plant a tree, but then contribute for clean water in relation to server usage?

[kind shore]

Hello Good morning. before the rant "PIN ONE SONG TO PROFILE " this could be a signature song . **But this comment is mainly directed at a welcome introduction to "Who we are" and "What we're about" an instrumental with context ? ** Mine is ready 😉

[severe glen]

SUGGESTION: I want to swap the voice from a generated cover to mine by uploading my sample voice....

[stray pebble]

This! If it can! The interface seems to be able to do karaoke lyrics highlighting already, we just need "rest of the owl" support for separate instruments and vocals when generating a track the first time. Oh, and the lyrics highlighting needs to be per word and not per line. Would help if it were bigger.

[worthy vale]

give us the ability to hum our own extensions to our songs with the extend tool
🤝

or whistle, or whatever else

[pseudo inlet]

Is there a way to bypass the 120-second limit when uploading sample tracks?

[prisma aurora]

Yes, this. Get rid of 120 second limit. Do it based on size instead

[worn moth]

add feature where we can delete the bad generations and to earn like 50% of the credits that we lost (to can generate more songs to improve the AI model)

[woven berry]

Any chance to remove rate limit on disliking our own creations? Most of the time I do batches of songs, and then remove em quickly, and get the rate limit notification each time.

[pearl carbon]

They just need to add this to mobile

[pearl carbon]

They had an intro feature but didn't include it with the latest drop for some reason

[bright cliff]

This balloon sticks for more than 10 seconds, now it's around 30 which consumes a lot of time trying to work on a project, i was a lot more satisfied with the older timer.

Applying a cross on it will make it better

Just like in the remastering balloon

[modern summit]

seriously for all the credits we waste especially these days with all the messed up generations/extentions/replacements, i'm sending this suggestion again:

• put dislikes in the trash automatically. i don't care if i eventually may come back to it later on second thought, it just polutes my library
• get 1 credit per "permanently deleted" songs from the trash. make it really useful for once, please. i discovered i'm litteraly having 200 trash pages! why! just why!

[low hinge]

I don't want auto delete. We have a search feature in our library so i can easily pull up 99 percent of the tracks i am looking for. sometimes i have to scroll a few pages but to me its not a big deal. I have never deleted any track including the "trash" generations. I can still use them with other programs and cut out and re use drum tracks, bass or vocals basically any good parts that i still would consider using. Sometimes i don't have time or attention to listen to the full track and later on when i am walking or working out i can hear the entire song. i actually discovered like halfway through a crap track one of the most beautiful and amazing breakdowns and songs that i would have never heard if i discarded it automatically.

[lucid light]

Can the "Following" feed sorting be adjusted to be by the date/time the track was made public, instead of the date of creation? Right now it's too easy to miss something if a creator didn't toggle that switch soon after creating the track. (And to prevent abuse, it could be the date the track was first made public).

[static salmon]

💡: With Suno having millions of users, how that music is presented needs to change radically.
Because every day we see the same artists don't we?
Yeah, so where are the rest of the millions of Sunoers?
That's also making music never to see the light of day.
Shared but never shown to the rest of us.
I'd suggest those who DO NOT write their own lyrics not be prioritized. Or shown as an optional sub-category.
I reckon they are doing it for fun, for their own sake.
Not to be recognized then would be writing own lyrics.
Anyone can specify the Style of Music and press a button.

[rigid void]

Sheet music generator would be cool for those who would like to play their pieces on an instrument.

[haughty dome]

Download a playlist (MP3's) - #💡┃suggestions-web message

[haughty dome]

I have several British personas. Try using prompts like "Second Wave Ska, Two Tone" "70s British Boogie Rock" "UK Glam Rock, 70s British Rock, male vocals" "70s British Rock, Art, Eclectic, male vocals" - "early 1960s uk merseybeat merseysound" Once you have the voice you may want to re-use, make a Persona out of it.

[rapid willow]

1️⃣ Suno could make some voices without copyright available to choose, and in each of them, being able to make changes such as: softer, more rebellious, powerful, more pitch, etc.

2️⃣ This already helps a lot that you want to sell or release a music album, because it will always have the same voice.

3️⃣ This until they organize a way to check our sent voices, as there are many people who will send voices like Michael Jackson or Freddy Mercury, and this cannot because of copyright and should be checked.

[finite lark]

hi, so i kinda want to see a feature in suno where you can dub existing songs to another language, ik theres elevenlabs, but its limited asf, plus it wont even give good results when dubbing musics, so yeah

[sleek vigil]

I need to be able to replace 1-2 seconds not just 3

[peak knoll]

Does any veteran know how to properly edit/replace to have the same tone, style, and singer as the current track? I have 4 lines for 1 stanza and changing only 3 words in one line changed the entire tone and style of those 3 words. They're not compatible.

[tired vault]

Extend the 120s limit to idk 240

[honest bridge]

This may be an Apple issue instead of a Suno one, but when using the default iOS browser to use the mobile Suno website a significant portion of the page is blocked by banners for the iOS app. The mobile web app is much more functional than the iOS app so it ends up just being a disruptive reminder that there’s something worse available. Sorry to sound so harsh but I think that was a pretty accurate take.

[dusky cedar]

You may have tried this, but it might work better when a smaller region of time is selected (e.g. around just the words you want to replace). Let me know if that helps?

[peak knoll]

I've tried it. Spent a lot of credits. I have a handful of songs that's unfinished because the edit/replace AI is designed the same as the main AI to generate songs. It suppose to follow the same format/tone/singer/style as the current track but it's not. Changing a couple words will make it "innovate" differently. Also, if I make it too short it ignored and skipped my word I want to change completely.

[dusky cedar]

^ going to follow up in dm

[azure plover]

I'll be very very happy when that is added

[sleek vigil]

Could we get a basic amplify feature to make parts louder or more silent manually

[sleek vigil]

Remove the style of the replacement and keep it empty or only type something like remaster
Then generate around 400 takes and there will be one that fits if you keep it at 3 seconds you get credits refund

[radiant shell]

It's certainly a lot of trial and error. Will be nice when one day we can really fine tune things be specific like ensure male/female vocals, being able to use the exact same voice on multiple songs, etc. I get frustrated all the time with Suno, but try to remind myself that we are in the early days of using AI to make music, and will get much better with time. Good chattin with you, take care!

[torn verge]

multiselect for items to be deleted please, also clear trashcan option... also "download complete workspace" or somethin'. Also...maybe add a way to monetize/ donate to your fav creator directly on Suno Website? And let suno take % of the cash? as is currently there is hardly any reason to set commercial license songs public, and everyone struggles with youtube flagging harmless Suno music for theft, same with Streaming Services, it's a mess. Or have suno distribute song option too

[pulsar pagoda]

How about the ability to Cover one song with the persona of another?
Currently copy/pasting styles doesn't quite do the trick

[gray barn]

Hello, I have 2 suggestions: a trash bin for the deleted songs and new improvements into the remaster, for example: a variation mode to adjust how different the remaster is from the original. I'm spending a lot of credits in the remaster mode trying to get better versions of my v3 songs, some are really good and I don't want lot of changes but V4 is changing too much and most of results are not so good. 🙂

[midnight jungle]

Suggestion for the entire platform: much like the remaster button, but without correcting or recreating the track… it would be awesome to have a button to normalize the levels to the top. As it stands most tracks as about -3db to -6db depending on dynamic styles. Eitherway, I think a lot of people are exporting their music in order to normalize the music instead of using the mastering feature. 1) song sometimes slighting changes 2) uses credits when all people want is to make it louder / normalized… 3) if remastering is required to get normalized exports, can we select various mastering styles? Similar to Lander.

Let’s get a free Normalize button or at least something normalized during export? Thanks y’all!

[midnight jungle]

Additional idea - being able to annotate the mastering style with {} or [] with different types of mastering… say u want it mastered for Vinyl or mastered for 1990, 2000s, 2010s or 2020s - I’ve experienced with this before and it does quite well in the style section, but there’s times when I want to have a track mastered specifically not relative to the music style. Just a thought, and another reason I’d love to work for Suno!

[midnight jungle]

I second this! I told my friends to not dislike a song as it completely removes the song from view… having a trash bin could allow to keep the song and bring it back or to remaster/cover the song… this could be fixed by when u hit the dislike button, not removing it from the workspace.

[sleek vigil]

I would like to be able to consolidate public songs into "albums", like subfolders you can create to organise your profile, with a nice album title and albumcover

[potent plank]

Playlist feature KIIIIIND of does this already but yours sounds like a a slightly different ui and flow

[spiral escarp]

yeah a better way to sort the mess on the profiles is needed without having to sort into playlists and make them private. We need tabs

I would very much like to sort my song topics so anyone looking can find what they are looking for and not have to scroll through all of it

[stable sigil]

Remove global trending

[arctic python]

This would be good to have I really needed a way to add an intro like extend but at the start.
Also I wish we could set the time of an extend as the whole 2 mins is often way too long and we still can't edit out bits from the mid section (not replace)

[sleek vigil]

Suggestion for the Edit/Replace Section list

[tacit mist]

i would like if the animated 3 lines play sign was easier to see on white album art. everything else is perfect thanks!

[autumn basin]

why isnt it possible to buy another 2500 credits for the same ammount you buy the month? thats annoying. thats keeping me from buying more....

[delicate cedar]

Extra Credits are one which persist between the monthly subscription and only come into play when the monthly allotment has been exhausted, making them more valuable. So they cost more.

Perhaps you would be happier with the Premier Plan instead? That comes with 10,000 credits,

[echo tapir]

Hi All, Not sure if this has been suggested before but would be nice to have a notes type column so you could add notes in all the generated songs as I generate tons of them listening for parts or changes i like to then combine into the final song. Something like this

[delicate cedar]

It has been but I hope each time someone else suggests it the developers see again how valuable it is considered to be by the users.

[hybrid pasture]

First of all I love Suno! Then, a lot of songs start from second .5 and it feels rough playing a song and going immeditely into 100% density. Just like the new fade out feature, it would be cool to add a fade in or a silence pause before the songs we publish so it starts smoother. ❤️ Great job team Suno is the best development ever.

[stray pebble]

would also be neat if an instrumental can be made to loop

[sleek vigil]

i need this on suno at least this

[rocky jay]

Not sure if this was brought up but can we please please move the "confirm & generate" and "apply replacement" buttons away from each other. Too many times i've mistakely clicked the button since it's in the same spot when i meant to click generate and instead click apply.

regardless of color, the placement is the same and can be confusing. too many times i've mistakely done that.

[upbeat rock]

you shouldnt have the option to CROP a PART 2 extensions if the crop will then make it invalid for "flattening" to the whole song

[humble river]

I know I will be a minority, but I would actually love if you could add a light mode to the browser page. As someone with astigmatism, watching a dark themed page for a long time is so tiring and I constantly use suno for fun or work.

[echo tapir]

ah thanks mate for the info

[long robin]

I would like to suggest to be able to add a Creative Commons Licence for things if I wish to - so that folk can download & re-use without it being copyright striked on other platforms (or if it is, they can point to my original and note that it's okay to use).

[knotty rain]

I would really like to be able to rearrange the order of songs in a playlist.

[cunning blaze]

I would love a feature that lets you generate the song in a specific tempo or BPM

[dense dock]

Add 100 BPM to promt

🙂

^^^

Example

[next ridge]

You can use darkreader extension to make it light theme pretty easy

[next ridge]

You can just drag and drop them (at least on PC to re-arrange them)

Only playlists not workspaces

[pearl orchid]

on the new "edit" page, can the ai list the style it chose in generating the song? it would be very helpful when generating a new randomized song without any chosen style input.

[next ridge]

If you don't choose a style it doesn't use one on the edit page. It just gets the style from the music

It's how you can create almost identical copies

Becomes more important where you start/stop your replacement if you don't have style prompt

Also if you just keep same style prompt there it seems to kinda "stack" almost overdoing it

[tepid umbra]

How can i replace all the lyrics ? Can you help me /

[hot pier]

it would be useful to have several functionalities in create which would avoid then going into other AI such as MOISE or CHORDIF

1 - a field where we indicate a BPM so that the composition is in this BPM
2 - it would be useful to indicate the BPM in the creations already made
3 it would also be good to have the dial tone

4 it would be good to have the chords in PDF for example of the songs created

thanks for these additions if you can

and the possibility of exporting a midi file of the song's chords as chordify also does

thanks

[knotty rain]

Weird... I tried that before, and it didn't work. Now it does. Hm.

[delicate cedar]

Please do not repost the same thing multiple times, it comes off as spamming. Thank you.

[twin lotus]

Hello, I paid for a profi plan but when I want to buy another credit it keeps loading and won't let me buy another credit, I don't understand why, it used to work and now it doesn't, if it doesn't work I don't want to pay next time already. what a shame.

[low hinge]

Can we get a showcase for strictly instrumentals. I don't care for most people or AI singing and when I hear a voice or a singing style that doesn't match or line up it ruins the entire vibe of the track for me.

[granite crypt]

Add a "Notes" section (to be seen only by the user) to each song listing that shows up under the song title where the user can list modifications and alterations, etc that help easily identify song variation differences at a glance, if that makes sense

[maiden sonnet]

why the support dont answers emails? I have the same Problem i cant remastered v 3,5 to v 4 can someone help me?

[wicked galleon]

Any V4 for free users (no promotion credits, permanent like v3.5) ETA?

Any V4.5 ETA? I imagine in May/June (V4.5) and November/December (V5).

Wishlist:

1- To be able to create Personas with uploaded audios
2- To be able to make collaborations between several Personas and to be able to choose in which part each one sings, this is necessary to create groups/bands
3- To be able to create songs of genres like plugg, pluggnb, new jazz, jerk, rage trap, jersey club I have been asking for this for a long time
4- One free remaster per day

I would like the Suno app to have all the functions of the web app.

[clever marten]

Suggestion:

Ability to use Persona for Covers

[tropic girder]

I want to see shuffling and repeat-all added to the player.

[polar gorge]

Criar música A partir de vídeo e imagem

[low hinge]

I would like to be able to hide the terms used for the styles and excluded styles used to create a song. Currently we can edit a page to hide lyrics or the prompts used to create a track. I don't want people stealing my "recipes", I think this is a big reason I won't share tracks publicly. Thanks.

[vast dawn]

Increase the size of the "Style of Music" text box. The visible area is too small, even for 200 characters.

[grizzled kelp]

Suggestion: More payment methods so more people can subscribe and use 4.0 sobs in German

[little sleet]

i am begging y’all to allow us to add longer clips of audio in the next version!!!

[kindred hull]

The redesign of the UI is cool, but I feel that the space for the space allowed for the "Style of Music" box is too little. Moreover, the suggested tags overlap the text to be entered:

[chilly fable]

I'm new to Suno, but I'm noticing a lot of things that either don't work as expected/advertised or just plain don't work. Is Suno the company having troubles? Or did I arrive at particularly buggy time beacuse of new feature releases?

[kindred hull]

Can you be more specific? I've been for nearly an year using Suno, and it's only getting better. The point is that you have to learn what does it works. For instance, creating a list of the styles you like best is a good starting point.

[runic pier]

You don't need more space, you can put extra info in the lyrics tab in [square brackets] for the same effect with even more control.
[Instructions: fast tempo, upbeat. {Instruments: clarinet, trumpet, flamenco guitar}]

[night pond]

can we monetize our song by creating it using Suno AI?

[gray barn]

I noticed a trash bin in the Library section, I don't know if it's new or if it was already there, but I wanna thank Suno for that. Most of my deleted songs are there 🙂

[runic pier]

Depends what you want to do. If you're using V3.5 then you probably have the free version and the voice is pretty hit and miss, especially at the start of gens. So consider starting with a [short instrumental] or [instrumental hook] to give it a little bit of a buffer. If you have no paid tools available then you are pretty limited and it's been a while since I have used that version so I'm not sure of your options. But if you have a subscription, you can make some pretty decent stuff. gotta echo what the other guy said though, what Specifically is the issue you're having.

[runic pier]

You sort of can. I know it is advertised that you need a subscription to be able to monetize your gens but, 100% AI generated music is not copyrightable. So if you said "AI make a Instrumental track" with zero input of your own, nobody owns it. Not you, not Suno so you can do what you want with it but it will be public domain. If you write lyrics then you have more ownership over the creation but you might need a subscription to monetize if suno has something in their terms about their service "enhancing the lyrics you put in" but I haven't read them, I ain't a lawyer, and laws change so depends where you live. Legal talk aside though, Suno is really not gonna start pursuing you for monetizing a gen without a subscription unless you really make a load of money on it.

---

SUGESTION:

Allow people to add covers to the prompt when extending.
Every sensible suggestion is just about the users wanting more control. This is the best way to directly tell the AI to "sing it like this" or "make it sound like that." and it's a feature you already have so I would assume that it wouldn't require much extra work to make happen. (probably)

and while you're at it.. allow our original lyrics to be public if sung in an uploaded audio. If people abuse the system and upload copyrighted songs, is that not just equivalent to someone uploading an inappropriate video to youtube? I don't understand the reason Suno would be held responsible for Users breaking terms of service.

[rare garnet]

"You're out of free songs for today, but you'll get 50 more every day."

Might be better to add credits somewhere in this message

[static salmon]

The "For You" section only works if you create a new account and choose what genres you like. I know this because on my main account I see nothing but slant rhyme one click songs. On a new account I don't...easy to understand right? Because on our current account(s) we have NOT been asked to select what type off genres we like. Is this yet another rushed oversight by you devs? or a bug.

[bright cliff]

I got a perfect solution to keep the cloud space a bit clean!.

Credit refund when you delete files from your trash bin.

What does this mean?

It means that if you clean out your trashcan on SUNO you can get 1 credit back per song.

This will keep the website it's cloud space clean and motivate others to clean up some space on the website.

^ Now this will permanently delete clips from the website, but it also saves space and we get a little credit for keeping the website's cache clean

[analog phoenix]

Please create some opt out function.
I'm a browser user , I have NO interest in using the app.
Now every time I refresh my page via mobile. I'm getting the pop up, Suno is better on app , go to app or continue on browser.
Appreciate you want to push the app, but this is too much and very intrusive.
Give me an opt out profile function that I don't want app and stop spamming me with features that is only going to drive me away.

[stable sigil]

Remove the android app popup for browser users it's so annoying

[twin bear]

STOP!!! 🛑 with the SUNO App advertising in the Web Browser and Mobile Browser platforms!
It's annoying and uncalled for!!!

[jagged chasm]

FEATURE REQUEST

Just had a great idea: Suno has to have a "Reuse Styles" feature. Basically, we can save a set of themes and name it to, for example, "Rap 1", and instead of entering the styles manually into Suno, we already have a presaved style.

I know there is that "Reuse Prompt" button, but you get the whole song reused and you have to scroll through all of your songs to find the one you need and its just a headache.

Thanks Suno team!

[opaque gull]

Hi,
can you please add production date to each song? (DD/MM/YYYY)

[chilly fable]

I have the pro version (the paid first tier). I could put a lot of work in (as a product manager) and enumerate and prioritize all the issues I’ve experienced, but I won’t because by the sound of it nothing will come of it. However, because you asked a few critical issues off the top of my head I’ve experienced in the past 24 hours are— the lyric editing seem to be broken completely eg doesn’t actually work. Extending seems to not actually just extend the songs but rewrite them sometimes in a different voice, style etc.

[eternal flame]

"Style of Music" dialogue box is very small for 200 characters. Is it possible instead of dragging out to the right, to actually make it taller so all 200 characters fit in the box on the screen? I have to always use notepad to put my styles, and paste into it. And can't see the full list to change it later.

[runic pier]

If by lyric editing, you mean the "replace section" then I agree, it's terrible and I haven't used it since it stopped being free. Editing lyrics might be possible through covering, first you need to edit the song details of the song you are covering to the ew lyrics, then edit your normal lyric box to the same so that they both match and generate a new song. I think this should give you the best chance at a word changing but even then some things just might not be possible, It's likely that similar sounds will cover in to each other easier because a cover has a bias to keep the sounds of the lyrics the same. so like 'fray' would easier be turned in to 'lay' than 'cut' for example. If you want to change more than a single word here and there then 😬 it's not really something I've ever tried so maybe someone else knows how good your chances are with this method but my bet is 'not great.'
In terms of extending, you're right that you get a new sounding generation but this is actually a good thing. Suno seeds are very restrictive in general so songs can sound repetitive a lot of the time, especially instrumental tracks. So a way to keep it from getting boring is to let it play out a verse or two and then use the extend to generate the next part, not to actually extend keeping everything the same. in terms of the voice though, you should be able to just cover the song again as soon as you have your full song stitched together and, if you put your prompts right, like:
[verse 3, {Bob, male vocals}]
blah blah blah,
this is my song bit,

Do that consistently above every verse and you should be able to get a cover that has a consistent voice. you again, might need to change the lyrics in the Song details for better results with this but I haven't tested that so it could actually have the opposite effect and be less helpful.

[runic pier]

You don't need more space, you can put extra info in the lyrics tab in [square brackets] for the same effect with even more control.
[Instructions: fast tempo, upbeat. {Instruments: clarinet, trumpet, flamenco guitar}]

[vast dawn]

Woot! 50 free songs a day... oh. They meant credits. 😭

[vast dawn]

Oh, what an optimist you are! It ignores your method just as surely as it ignore everything else.

[runic pier]

no, it works, I've done it many times.
But you shouldn't need to use all 200 characters anyway, all that will be doing is diluting the tags that would be most important in achieving what you want. Tags are not like a check list that will all be included. (I should have realised this before but didn't, oh well) The AI cannot juggle everything you put in, it seems to just look out for tags of certain types (tempo, lyrics, instruments, genre and probably some others) and give the first of each type a bias, like it has been instructed to "take (genre, lyrical style etc) in to account." it then gives weights to the tags (based on the order they are written for the first gen but changes them around with subsequent gens) So if you have 15 tags, you're giving it too much. You gotta be more concise with how you describe what you want. what does your tag box look like? and what do you want you song to sound like?

[trim smelt]

There is a issue with my Suno in my browser when I use it on my mobile. I can't upload the audio I created. "Upload Audio" is not highlighted if I try to upload from my device.. But all Suno files are showing there. Any algorithm change or it is a bug? I could it to few days back.

[tame fiber]

Srsly suno, remove that annoying new banner on mobile browsers ... im not gonna use your bad app.
This aggressive and annoying banner just makes me wanna cancel my subscription.

And the banner is hueting my eyes everytime ... black website with a fat white banner popping up ..

[vast dawn]

It literally ignored me when I tried it. The issue is it is too generic when following "instructions" which at most are suggestions to it. Are you familiar with the iconic piece from Tubular Bells where each instrument is played in turn. Suno can't do that. You get one or two at best. The closest I have got so far is : https://suno.com/song/0faa0d8c-402f-4590-b836-fe7d4f9f866c?sh=YxmhORuJXH7vVP66 and I had to use generic text, because when I used text related to the instruments themselves, it was announcing the wrong instruments, or the same instrument over and over again such as here: https://suno.com/song/85f8527b-1aa0-4ed3-937f-ea5f5a1c15d9?sh=crkTlkBA4llJp1rL.

[wild pumice]

New Editor has much omproovd options to report what is wrong with geenrated song. But outside of it Bug reprot options is not updated. Please, make Bug report have same reporting options.

Song listings would benefit if there is option to sort songs by various criteria (title and date of creation are a must).

It alos is needed to have good and elaborate options for filtering song list. Filtering by keyword in title, lyrics and styles, date of creation, type of manipulation (first creation, crop, replacement, extension, cover, persona...), bug report type, like, dislike, especially if they could be mixed, would help a lot.

[low hinge]

more space is always better. plus it cuts off so you cannot see the full 200 characters in the style box. Love to see an expansion of the 200 character limit in the style/ excluded styles and more than 3000 character limits currently in place in the lyrics box would be great progress moving forward. i use square brackets in the lyrics box but like 10k in the lyrics and 3k in the style would be nice. I never have enough space and have to try to go back and edit out specific sections or ideas. like i want to give commands to use specific vst or instruments played in specific patterns or styles and it can take allot of text to get these instructions to be understood by the ai clearly. sometimes i give it instructions like a teacher to a student advising on a live performance, i want more subtle breakdowns, complicated variations, details and nuance and need more space ie a 10k character limit to help achieve this.

[distant sigil]

(1) A way to mute or pause Suno profiles for when people wanna take a break from Suno especially for people that’s just using it as a hobby. The only suggestion I can think of other than what I suggested already.

[old trench]

why upgraded to VIP but cant produce whole song? even using v4, v3.5 all cant produce whole sone, only 2:30 max, but using free account can produce 3:50 whole song, why?? and support no any reply. what can i do now?

[obsidian marten]

have "extend" feature ACTUALLY extend the song, rather than making two new ones (as well as actually adding an ending to it)

[severe cipher]

Add a way to make really small, but focused changes to a song.. Let's say you make a song and it turns out there's a word that comes out wrong.. Then you can go to some kind of "micromanage" option in the "Edit" menu, and mark that lyric you want to change in a certain way, like using {} brackets for instance. This keeps the song as is, but attempts to change just what's in those brackets.. Say max 2-4 uses per generation.

[wild pumice]

This actualy worked for me today in new Editor. I marked single word, typed in replacement, generate replacement and to my surprise it did what I expected.

I tried the same on other word in the same song and it warned me that i should select at least 6 seconds for reliable replacement. I followed that and it did it again! I am stunned, new Editor did not work for me at all so far.

[wild pumice]

I get songs duration up to 4 minutes. I guess it heavily depends on styles and especially on lyrics. Try inserting some instrumental segments in text to fill in more time space.

I frequently use [instrumental], [instrumental break] and [instrumental prelude] but i also saw people are using just [music] where they want some instrumental part to show up.

I always use [intro][instrumental] and [outro instrumental][end] at the end of a song.

You can try specifying slower rhythm in styles.

If song ends up shorter than you want, you can use extensions to make it longer.

[obsidian marten]

Extension doesnt make it longer though

[wild pumice]

You first create extension and then use option Get Whole Song to get everything packed up.
Are you using extensionoption form the Menu or in new Editor?

[obsidian marten]

I'm using the suno website, no idea wtf you on about

[wild pumice]

Well, then you have to describe how do you use Extension option.

[obsidian marten]

i select the extend option then press extend from 3:59 where the song ends

it just makes a brand new 2 minute song that also doesnt end

[halcyon linden]

I am pro subscriber and need help logging into my account

[wild pumice]

Well, there are two mistakes you make:

1. do not extend form the very end, back it up a bit, at least one second so it can catch vibe of the song you are extending. Make some logical point to discard rest and start new part of the song - Extension. When creating Extension there is filed to set time point of source song from which you want to extend.

2. After extension is created it indeed is new song (i mean new file) but it will have mark it is extension. That is not full song, just extension part. In Menu of extension you will have option Get Whole Song. Use that to merge first part with Extension. That would produce next file.

[obsidian marten]

Why does it need to catch the vibe it already has it in the main song, it just needs to finish the original song off. If extending isnt just extending and finishing the original song it needs to be named something else

but even so... it doesnt end the extension in the 2 minutes anyway

theres no point to extending if it isnt going to put a proper end of the song

[wild pumice]

It does not work that way. If you extend form the very end you are extending from silence so it has no ide what to do - thus it creates new song. Suno has to have context to work with.

[wild pumice]

You can ask for proper end using [End] tag at the end of lyrics.

[obsidian marten]

but it doesn't get to the end of the lyrics before the 4m point

[wild pumice]

Suno tries to understand what you want, but you have to be clear, make your best to explain it in a way it can understand, and even then it does not have to comply to all you ask for.

If you get song that abruptly ends ad 4 minutes, it usually means you put to much text in lyrics and it could not fit. If that happens make extension but from the point where song is ok, discard rest that was abrupted and set new lyrics just for extension part to continue at designated point. You will get song longer than 4 minutes.

[runic pier]

put <end song> in every part of your prompt when extending you final part. Put instructions to end the song in the lyrics box, put it in the title box with the title and "outro" in the tags, put [outro] before your last verse and put <end song> at the bottom. you need to overload the prompt to bias it towards that. ending is annoying in suno but this usually gives shorter gens.

[wild pumice]

Recommended way to create song is to use part of the lyrics up to and including the first Chorus. Let it create song. Once you are satisfied with music, make extension from the point behind Chorus, and add next part of the song (again up to next Chorus) and repeat until you get to the end of song.

[obsidian marten]

lol all this nonsense to have to do instead of "extend song" just doin the blatantly obvious thing

[wild pumice]

Well, tool has its own principle of working. It does not have to match what you expect. It is far from nonsense. It is up to you to learn how tool works or not to use it.

I see we are just wasting time trying to explain to you what to do, as you are not accepting our suggestions but insist that tool shuld work in a way you want. That will not happen.

[wild pumice]

I was thinking to offer to remotely access to your PC to quickly show you on example, but it seems you are not willing to learn how to use Suno, just to complain it does not work way you think it should.

[wild pumice]

First thing you have to accept is that machines newer do what you want, but what they understand you want. And their understanding solely depends on how you explain it to them. If you want to talk to machine, you have to learn machine's language as it will never speak yours.

[runic pier]

Yeah, like I said, the AI does not go through everything like a check list, it can't hold all that information. and because the seeds constrain the music, it might be hard to add multiple layers like in Tubular bells, although I've never tried it. (To do that you would need to use extends or uploaded audio, which it seems you don't have access to.) The AI instead looks at your prompt and gives you something more general. you should also look up what each type of Bracket means to the AI when prompting. And look up the exact terms for what you want in your piece of music, like "Ostinato" or possibly a "cannon" (idk if that's what you want but it's definitely a better tag than "drums and other instruments" - that is just a waste of space.) The tag "switching between instruments when indicated" does nothing, That's not how AI prompting works. You don't need the tag indicating an instrumental because that is implicit in the fact you have no lyrics in your lyrics box (anything in these brackets counts as lyrics btw) Basically your first step is to learn the musical words for what you want. which you can do with a bit of googling.

[wild pumice]

It seems @obsidian marten gave up 😦

[runic pier]

Good, so have I 😁

[wild pumice]

I have to remember to try that <end song> thing to see if it can fix endings. Suno is really odd in ending songs. It used to be quite good.

[runic pier]

it's not even 100% it just makes it more likely, as with all things AI. usually give me 0:30 to 1:30 long Gens, depending on the song and maybe the point of extension.

[wild pumice]

Today I even saw someone used <chorus start> and <chorus end> but I disregarded it as it was strange syntax to me.

[runic pier]

I never seen that, It doesn't seem too useful. you wouldn't need to tell it where to end a specific part, you just start a different part if it's half way through the song. also it seems too specific for the AI to even listen to. I don't trust it either.

[wild pumice]

Maybe it can help if chorus is not just one verse, to make it understand that next part also belongs to chorus.

New editor is helpful in that manner as it displays song structure as it understands it.

[runic pier]

In that case i would probably do something like:
{[Chorus]
blah blah blah,
song lyrics in one verse

and another down here
so the song doesn't sound worse}

haven't used the new editor much, I'll check it out.

[wild pumice]

Curled bracket can be used in that manner?

[runic pier]

used to say that different parts are connected, yeah.

[kind grove]

Is there a planned option in the edit section where individual words can be changed so that only that specific word is re-sung/spoken? The background is that individual words are often mispronounced. Unfortunately, the replace option doesn't really help either, as it often changes the overall sound of the section. It's really frustrating when you've generated a song that you otherwise like, but it becomes unusable because some important words are sung incorrectly—especially problematic with proper names.

[crude river]

Yeah and I think we should Combine two audio files but it has the first audio is genre and it's correctly stable in speed

[kind grove]

Is there a planned option in the edit section where individual words can be changed so that only that specific word is re-sung/spoken? The background is that individual words are often mispronounced. Unfortunately, the replace option doesn't really help either, as it often changes the overall sound of the section. It's really frustrating when you've generated a song that you otherwise like, but it becomes unusable because some important words are sung incorrectly—especially problematic with proper names.

It would also be great if users could create their own voice for singing, which could then be used in the song (possibly even through tags in the lyrics: [Verse 1: <MyOwnVoice_1>] ... [Verse 2: <MyOwnVoice_2>]). This would help maintain consistency across an album or define different characters who could sing against or with each other. When browsing Suno, many of the singers tend to sound very similar or even identical. It would be amazing if users could use their custom voice to re-sing an existing song without changing the melodies and instruments.

[wild pumice]

Edit field for Style of music is to small. I am not talking about number of characters limit but it is visually so small, hardly two lines can be seen and there is no scrolling feature on the side. It should be higher allowing more lines to be seen.

Today New Edit Replace works most of the time but after two replacements it is lost - starts creating wrong lyrics, not one specified, or even some gibberish words. It seems to me that for some reason, Suno messes synchronization between lyrics text and actual music. It is easilz seen in lyrics of song after replacament, it is messed up, missing parts, mixed stuf and so on.

[kind shore]

I would like to share some feedback regarding my experience on your platform, particularly concerning language accessibility.

As an English speaker, I’ve noticed that a significant portion of the content I encounter is created by users from diverse linguistic backgrounds, such as Chinese, Russian, and German. While I appreciate the rich variety of music and artistic expression, I often find myself at a disadvantage, unable to understand the lyrics or engage meaningfully with this content.

I believe that having a built-in auto-translate feature would greatly enhance the user experience for many individuals, myself included. This addition would not only foster inclusivity but also allow listeners to connect more deeply with the diverse musical creations available.

I value the cultural exchange that occurs here, but I feel that without a way to bridge the language gap, my experience can become limiting. I would prefer not to feel constrained to a specific audience or genre due to language barriers.

If the goal is to promote multicultural engagement, I urge you to consider implementing translation capabilities to create a more inclusive and enriching environment for all users.

Thank you for taking my concerns into consideration. I look forward to seeing how the platform evolves in the future.

[wild pumice]

I tried to do Remastering and even Cover after Replacements are done. It seems to polish and even up overall music.

[kind grove]

Unfortunately, it has only rarely led to success for me.

[kind shore]

cover the song, make sure to remove all the prompts / negatives/ mingled verse instructions, leaving only lyrics. It might take 3 generations to change a misspell lyric, but if its just a suno misread lyric then the song should generate as you expected the first time.

[kind grove]

Thanks for the tip, I haven't tried that yet. I'll do it soon. I hope it also works for a fantasy word name. He is supposed to pronounce it in German but very often falls back into English pronunciation.

[kind shore]

try writing your German lyric with english phonetic spelling

[kind shore]

“In the still of the night, with a gootoo see-no, >> sounds like ("gut suno")
I hear the whispers of dreams as they flow…”

[kind grove]

Ok, the cover and the spelling of the name/word helped so far. But the cover is shorter than the original and is cut off in the middle. I have now created a number of covers but they were all exactly 3:20 long instead of 3:57 like the original. So now I have to try to restore the rest with Extend. For another 10 credits each... That's not fair Suno...

[kind shore]

i cant speak for suno, we all have this issue .. my comment was just something ive found helpful.
this frustration your experiencing == many people have advised to just generate songs using extend extend extend method. like Udio ??

[kind grove]

Yes I know. The part about "That's not fair" was aimed directly at Suno and not at you. 🙂

Here is the covered and remastered song. (I translated the lyrics into English for you)
https://suno.com/song/bddffb0c-f421-49af-9857-85c0d1725170?sh=SFC2Xbvp1JRkTfQj

[wide cape]

I know v4 is there but can there be like song dubbing like change the song into an different language
like japanese, spanish, korea, and etc

and please do not lock it behind a paywall

[vast dawn]

Especially as Sono's paywall is legendary for how bad it is. You won't get a penny out of me until reviews improve.

[severe cipher]

Yeah that's just not working.. It's clearly struggling and trying, but I wanted to change the word eleven to twelve, and it ends up like elewle

[kind shore]

weekends seem to provide inconsistent generations , maybe weekend brings a tsunami of new creators playing, loading the system?

[dense gate]

A copy melody function where you can make a song using a melody Like in mureka

[gloomy python]

I COULD MAKE MORE THAN JUST CONCEPTS If you would just generate lead sheets as a byproduct and let us EDIT them before generating more iterations! WOW. If it could regenerate off of a lead sheet, game over. I win. You win.

[untold grove]

I would like to request a new feature..... Im a game dev and suno makes great music... I would love for the ability to have suno make a song that will "Seamless loop" as if the song never ends.

[peak knoll]

Is there a way to upload a track in multiple sections and combine them as one track?

[wind thicket]

I missing on the browser "shuffle Modus" to playing all good songs in random. I use SUNO generated Songs in my Livestream and would be enjoy to shuffle. And the really best is Support for Stream Deck to pause and skip the next song.

And ofc what i wish is a Desktop App and not just in browser.

[peak knoll]

Please add an option to edit/replace to select instrumental only. I want to select an instrumental section for editing, but it keeps adding singing to it.

[kind shore]

Can you please turn off comments by default ..so these massive external discord groups aren't poaching all your regular followers.

[wild pumice]

I was surprised that it worked for me. However, I had to replace the same word in three places in a song. Two worked. Third one was a mess. It seems like Suno just messes up synchronization among audio and lyrics.

I then tried using Cover. In Cover I changed nothing but that one word and miraculously - it worked! Totally unexpectedly, it kept whole same song except that one word fixed.

[glad cypress]

I'd like to suggest that after an upload Suno could give you text to describe the style. That way you could create new similar tracks without using extend.

[feral grove]

Please consider implementing that on MP3-Generation for Download the cover picture is taken over.

[broken canyon]

Can we have a playlist / Album showcase aswell as their being a normal showcase.
In its current start, the showcase section has just being a dropbox dumping ground.
I used to love listening to stuff in there but the over exposure of tracks has now lead it to being a deterrant for exposure.
I believe reducing the ammount of links and the time for posts in Showcase will allow people share and appreciate each others posts again.
Having a seperate showcase for larger projects to spread its wings and would cater to those who want to show said projects.

[upbeat finch]

esta para ti, andy.... tiene mas de 12 años....

[upbeat mantle]

how about adding a Title generator kinda like the image one ?

[crisp rock]

I have just seen this filter ! It is useful, but hidden. Is there any notification about this ?

[wintry perch]

Do we know if there is any plans to have a combine two versions button soon? I would love to just take one part of a song and put it into another part of the song and just blend the two seamlessly. Sometimes one song has a killer opening but a poo chorus and an even more poo ending. but you want to save that amazing start. Then another song with have a great middle and end but a poo start. If we had a button that allowed us to either hit combine so it was random how it mixes together. Or a button then said combine and then in edit mode you could pick the parts to splice together, having both options would be hecking awesome!

[wintry perch]

OMG THIS! this would be amazing, sometimes i hate the voice, but i love the way its been sung, i love the beat of the track, and I love the cadence and tempo speed. But then the voice is utterly terrible. This would be an amazing feature. This I have been trying and playing around with Cover song and it sorta works to keep the tempo and keep the cadence but change the voice. But sometimes it still changes the song iteself too much along with the voice. And sometimes it does literally nothing and it all sounds the same so cover is sorta like this, but seems inconsistent with the results! so a button where its like change voice only. Would be great. That way we could take a song that was generated in female and turn the voice only male. Or stuff like that!

[crisp rock]

Please add here as a button also "ONLY FULL SONGS" so it is VISIBLE for anybody and FASTER this way than accessing the pull down menu.

[fair turret]

SEARCH FOR GENRE -- Please add this to the Suno Search Options not sure how this was overlooked

ADDITIONALY: Add an advanced search feature please where we can search for songs with a combination of metatags this will give us much more reach into the huge amount of songs in SUno

thanks

[gray barn]

WASTING CREDITS ON BETA STUFF: I think Suno should review the way how the users are using their credits on the platform, because it doesn't make sense at all we (users) spending a lot of credits on tools that are still in beta (especially in "cover" mode)...

[mortal pecan]

im having issues with uploads

getting a weird message like the artwork already exsits and it lags,

[high forge]

Maybe add an option to write things you do NOT want in your song. E.g. "guitar, flute, duet", and then it will avoid having these things in the song when generating

[tepid trench]

"Exclude styles" exists for this purpose (though it's not guaranteed to omit those things, just less likely to include them.)

[high forge]

No idea how I missed this but thanks

[raven patrol]

i need opinions c: https://suno.com/song/c584e56e-9a67-429c-a936-94adff39d54c?sh=s3lMjvReAiMw8YnA

[peak knoll]

Singing too low vs the music.

[fresh moon]

where is upload-vocal entry?

[empty thorn]

I might get a lot of hate for this suggestion but .... Optional monetization of PERSONAS for creators.

[prisma geode]

Dear Suno team,

If you happen to read this, could you please set it so that when visiting a profile, the newest songs are displayed by default instead of the most popular ones? Right now, new songs don’t get much attention because they’re pushed to the back, and users have to manually change the sorting. It would be amazing if the newest releases were shown first automatically.

Thank you for your time, and have a wonderful day! 😊✨

[haughty dome]

Let's have a review of the need to create 2 songs every time, 2 possible extensions every time, 2 possible replacements every time. Maybe we're feeling confident enough to go with 1 creation. Usecases: 1. . a short happy birthday song. 2. trying out a new genre with the current lyrics.

[upbeat rock]

I wish there was a way to be notified when your track was featued in a playlist. I'll notice a track getting some attention and have no clue why or how

[gaunt forum]

Official Admin API

I know there have been plenty of questions about Suno's API, and I know that there have been problems with malicious third-parties exploiting unauthorized access to it.

I 1000% understand not wanting to give easy access to music generation endpoints. It serves to avoid overloading, misuse and reselling that would impact negatively both Suno itself and its users.

Would it be possible to offer API access to purely administrative tasks?

Those would be actions performed on your own account that do not involve the use of generation models. Things like:

• Listing your songs, playlists, workspaces, personas...
• Editing song details (title, displayed lyrics, visibility status...)
• Adding, moving and deleting songs from playlists, workspaces, the trash...

It's not something unheard of, I've seen other services having separate "product" and "admin" APIs. Or offering API access only for menial organizational tasks, and restricting the main use of the product to its UI.

Since these are actions that work directly on a particular user's account, they can be easily protected and monitored with Auth Tokens... and most importantly, having a safe, official, authorized way to do it would benefit real Suno users, and be of very little use for bad actors looking to mass-generate songs 😄

[gaunt wave]

You can do this, but with other apps

[kind grove]

I'm sure... But I didn't ask for a work around. My suggestion was to do it in Suno 😉

Suggestion: An insert function in the edit section to add parts in the middle of a song. For example: Insert another verse. The previous extend function is unsuitable for this because it would also change the remaining song. Replace doesn't work properly either. Even if it's too insecure.

[small kayak]

I'm sure this has been mentioned but I would love a search function for our trash bin. I've accidentally deleted before and there seems to be no rhyme or reason and I go through endless pages to find it - but usually just give up.

[naive venture]

Right, I wanna suggest the function of combining multiple clips together, let’s say I have a beautiful build up in one song, and another song has a sick drop, I wanna combine them together, so that I can have that nice drops.

[magic flower]

I'm really baffled that Suno doesn't have a simple toggle that says I either want a female singer, male singer, or A/B duo

[sleek vigil]

Or none of that and something else entirely

[glad bramble]

get stems doesnt work super well which is a bummer, I wonder if the beat is literally just generated along with the lyrics?

[timid ingot]

Will we ever have an update that show us exactly what is "flagged for moderation" ?? Cause right now, the only way to know, is waste countless credits by cutting up the song part by part, then removing what you think is flagged, paste the song back together, only to find out it's flagged again (for no reason) since all parts cut were fine, but when the song is put together again, it's not fine?

[modern remnant]

Mass delete option for trash can.

[empty thorn]

Single word replacements would be amazing, sometimes songs are 1 or 2 words away from been perfect.

[valid dawn]

💡 I have a request, (unless someone else has asked?) could we have an update that features a band-in-a-box type section that allows for backtracks. No melodies, no vocals, just basic song-ideas.

This feature would include the song-writing capability where one could ask for a key signature, beat type, sections (Verse, Chorus, etc.), tempo, etc.

Thanks!👍

[faint geode]

你好

[atomic lintel]

a huge suggestion for suno in general, will we ever be able to upload clips longer than two minutes?

[wintry perch]

you already can, just shift click multiple files click hamburger menu click delete, it deletes all highlighted files at once. but can only do this one page at a time.

[grizzled dock]

Pretty sure this have been asked and suggested several times, I feel like it's about time to implement something like it. The suggestion is that we can get some points back when deleting/trashing a song. I dislike/trash more songs than I can use due to vocal errors, personas not working as they should, promts not followed, styles not followed, songs cutting off because they are more than 4 mins etc. Out of my 2500 points per month maybe 500 is somewhat useful, rest is just trashed. Just 1 point back would be better than nothing.
Thanks

[thin sleet]

Bring v4 to basic plan

Because we only get to use a few songs with it and 1 free remaster honestly just give everyone infinite remasters and bring V4 to basic plan please

[haughty dome]

I've mentioned it. But I think it should be renamed 'Archive'.

[haughty dome]

你好！

[junior tapir]

Make using a specific persona inside a song possible, lets say that I want to have a band of singers for a song, 3 personas, A,B and C, so I want them each to sing a verse, ofc they have the voice but they will follow the rhythm and style of the song, for example:

[Verse]
@personaA

[Verse]
@personaB

[Chorus]
@PersonaC

This will be incredible to compose songs!

Also please allow us to create persona from song extended from uploaded file, I made a song using my recorded voice, it was awesome but sadly I couldn't add my voice as a persona.

[twin grove]

I'd like to be able to transfer comments/likes from one song to another in the same chain. For example, I used the new Remove tool to cut down an extra long rest in a song like I did in a DAW, but the result is a 'new' clip, which resets its stats. Since it's obviously the same music, there should be some way to import the stats of the older one

[marsh lichen]

Well, now you can at least select a line for 3 seconds. Most times it works if you want to replace a single word.

[river trail]

i see there is limit of 3000 characters. My suggestion is to remove that for Premium Users. I have some songs goes beyond that

[oblique cradle]

What also would be great to produce real acapella with stems for bass, tenor, alt and sopran

[sick tapir]

can we have a group delete instead of one by one, need a box to tick to select all that is in the rubbish bin and so you can delete all.

[untold shuttle]

Plz add a checkbox for every track so we can delete multiple tracks at once & also add multiple tracks to playlists. It's so time consuming to do it one by one.

[untold shuttle]

So funny.. I didn't read your post before I post mine 😅

[untold shuttle]

Got a tip on how to do that right now... you can select multiple of them by holding Shift... then delete them on a page (or send to trash)

[sick tapir]

Thanks

[steady river]

Would love a list of "people i follow (or maybe people i used to listen to), that i haven't listened to in a while" as a front page category or list to access. Would also love to be able to filter my "for you" list by Genre/Mood.... as i don't always want to listen to the same kinds of things!

[peak knoll]

Edit/Replace ignored what was highlighted and changed everything past my highlighted section. Please take a look when you have a chance.

[viscid token]

could we have a duet function where we can assign two persona's to a song while having an overarching prompt, which must be similar to one or both of the persona, then these could be assigned verses/chorus through things like [Verse 1][Persona 1]

[fallow snow]

anyone know how to adjust length of replacement parts after generation?

[tribal raptor]

that would be cool if you could choose an already existing song and than you put your own lyrics and the song will be styled like the ,,already existing one,, you choosed

[unique imp]

The possibility of setting: Part of a song (Chorus, Verse etc.) before and especially the total minute length of the song or the individual part.

[dark magnet]

Hey @tulip sinew,
Suno.ai has already positioned itself as a leader in AI-generated music, empowering creators, businesses, and developers with cutting-edge technology. To further expand its impact and increase its profit margin, offering an exclusive API for Suno.ai subscribers would provide a strategic advantage, driving both user engagement and revenue growth.

By enabling API access exclusively for paying subscribers, Suno.ai would:

• Increase its profit margin by adding an additional value proposition that encourages more users to subscribe and remain engaged.

• Extend its technology beyond individual users, allowing developers and businesses to integrate AI-generated music into their own applications, games, and platforms.

• Strengthen its presence in content creation and live streaming, providing adaptive, copyright-free soundtracks that enhance user experiences.

• Unlock new opportunities in marketing and advertising, offering brands the ability to generate unique audio assets for digital campaigns.

• Support automation and personalization, enabling industries such as fitness, wellness, and virtual events to leverage AI-generated music seamlessly.

Also, another additional opportunity to expand Suno.ai's brand influence is through

integration with Lumia Stream, a leading platform for interactive live-streaming experiences. By integrating Suno.ai with Lumia Stream, creators and streamers could:

• Enhance audience engagement with AI-generated music that reacts to stream events, chat interactions, and alerts.

• Create dynamic, real-time audio experiences, allowing music to shift based on gameplay, mood, or audience input.

• Offer streamers an exclusive toolset, positioning Suno.ai as the go-to platform for interactive, adaptive music.

• Drive adoption among content creators, reinforcing Suno.ai’s value within the streaming and gaming industries.

By combining the power of an exclusive API for subscribers with Lumia Stream integration, Suno.ai would solidify its role as an essential tool for content creators, developers, and businesses alike. This strategy not only enhances** brand visibility and market influence** but also ensures sustainable growth and profitability.

[kind shore]

Hi there, Happy Saturday 👋

[kind shore]

Proposal for Album Presentation on Profile Page

I have created an album—a theatrical series of songs—that I would prefer to display differently than a traditional profile page.

Proposed Features:

**Defined Order of Presentation:**  
    The compositions should play in a predefined order.
    It would be beneficial to have the ability to pin the songs horizontally across the page, starting with the intro soundtrack and followed by consecutive songs (1, 2, 3, etc.).

**Album Orientation Preference:**
    This album orientation could be offered as a separate preference switch from the standard profile layout, allowing users to choose based on their individual preferences.

Rationale:

**Current Profile Limitations:**
 Currently, profile songs are ordered by the number of plays, which we have no control over. This system can create confusion, as it does not accurately reflect the intended presentation of carefully orchestrated albums.

**Individual Needs:**
 Adopting this new presentation method would allow for greater flexibility and better alignment with individual needs. Some listeners may prefer one song over another, and the default order can sometimes be disheartening and misleading.

I believe implementing these changes will enhance the user experience and provide a more authentic representation of the work.

Thank you for considering this proposal.

[karmic quiver]

Being able to pull in other users Personas would be pretty awesome too.

[Verse 1: usernameA@personaName]

[Verse 2: usernameB@personaName]

[unique imp]

The possibility of emptying the entire trash

[open pond]

Being able to change the style field for visibility / discovery. Sometimes I throw a lot of information into the box to get the sound I want, but want it catagorized differently in public.

[manic locust]

Workspace SPLICER!
You got Parts of the song but now you think it needs a longer Intro but it simply just can't...
Well, now it can just stem, Sample, and Add your edit to your main song by splicing. Once the splice is done it will remaster it seemlessly without adding or adlibing anything between the two spliced audios.

[void raptor]

Yes splicing in the editor is in opinion really needed, likewise as you have a fade out, having a fade in would also be necessary in my opion to have more control over audio creation process. To add this, another reason to implement this, is that if you have make changes to make these kinds of changes, then these will never reflect on the platform as we can't upload finished tracks to Suno for showcasing / sharing purposes on the platform.

[dim vault]

Why can't we use Personas whenever we want? Like...say if you created a really good instrumental and want to add vocals to it. Currently, you can't add a persona when you extend a Gen.

It would be so much better if we could do this.

[void raptor]

Also, persona's don't really work the way you think.
If you use a persona with the exact same seed prompt as the track it's based on, you will get an identical iteration of track.
I think having more control over various elements like voice, instrumentals, layers and so, would be useful and really enable users to become more creative.
So for example, if you could pre-define the sections, with length and what to contain, along with being able to insert section post generation to create new iterations based on the users vision for the track.

For example, say you had an into in the prompt but the AI ignored/failed to interpret it and thus didn't use it, but the rest of the prompt was used and generated satisfactory. As it is currently, you're stuck with the sections it created, but if the user had more control like having the option to insert new sections and specify parameters for the sections, then it would be much easier to fix tracks without having generate a complete new track, which most likely would be very different.

[upbeat mantle]

Been wondering a while why this function have not been added to v3.5 and that is remaster option like v4 which would improve v3.5 alot but at the moment we are locked out that option

[fair turret]

NEW FEATURE BEAT INDICATOR -- Pleasa add a beat indicator to the timlines for easier editing in all th tools. Give the creators the opton to snap to the beat to guarantee cuts replaces etc

[fair turret]

probably because they are not going to update a an older version of the software with a feature that is an attraction for the new version. This would require them to provide support for what they consider and obsolete version number

that said it would be nice if they did

[peak knoll]

Suggestion for song generation: Have the AI pre-read the generated soundtrack before delivering it to the user. If the words do not match the lyrics the user provided, don't display it and regenerate again. Or better yet, regenerate only that word and replace it behind the scene. This will save a lot of headaches for the users and keep your customers. So many have incomplete songs burning up credits because of 1 or 2 words.

[spiral pulsar]

Will there ever be a way to split the generated song into stems and reedit the stems independently from each other?

[sweet flicker]

I recently saw that it's possible to generate much faster AI results using dLLM (diffusion Large Language Models like inceptionlabs.ai) as it is diffusion based generation (which generates the entire results and then refines it in multiple steps) vs token based (which is much slower as it predicts the next token)

This got me thinking shouldn't this be possible for music models?
Most likely you probably have thought about this but if not perhaps it is something Suno could try to make,

The big advantage aside from the speed is that it refines it each step further meaning it can eliminate hallucinations and most likely degradation too (which i think is the biggest issue in the current model) keeping it coherent over the whole song.
Anyways just food for thought ^^

with each new step you could save the audio so you can kind of control the level of hallucinations

[zenith needle]

The option to extract the instrumental for pro users because the instrumental stem sounds like garbage, and some of these tracks I'd like to do live. I do use my voice in some of my tracks in Suno but lack the patience to sit here and learn FL and make my own beats.

[spiral pulsar]

Functionality : split track into stems and regenerate parts of individual stems or make changes to them in a piano roll tool

Functionality: stems to midi stems and back

[fair turret]

NEW FEATURE: SEARCH BY DATE RANGE - Please implement this in the library to make life easier thanks

[haughty dome]

Sorry, but how does this suggestion differ from presenting as a Playlist like this: https://suno.com/playlist/fa74417c-e56a-41c5-92ed-af300c86854c

[wicked galleon]

I need Stems in Studio quality

I need make personas with uploaded audios

[peak knoll]

Is there a way to avoid strong bass sound that drown out the voice?

[grave pasture]

This is a suggestion to notification tab, would be nice if we could separate the 'like, comments and new songs' inside notification tab, this way we can look over stuff without scrolling through.

[solemn forge]

please add bulk wav download from playlist , it takes too much time!

and if there is solution please inform me , i need to download 100 songs its unbearable

[golden bobcat]

Suggestions:

• TEMPO BPM (Slider Control on Hallucination to respect or not the TEMPO/BPM of the Reference Audio when creating COVER) -- Currently it is very Random, and sometimes it even varies on the same track. (It will allow us to create more controlled Samples or Songs in order to use it as a tool for creators)

• KEY/TONE (Slider Control on Hallucination to respecto or not the KEY/TONE of the Reference Audio when creating COVER) --Sometimes it changes the Key or tone to some semitones above or below the original track. (Same as the first one, it would be useful to have more control over the creations to use it as a tool for composers-producers)

• Properly Stems Separation. (Currently there are better separation tools outside SUNO,

• API (as pay per usage)

[broken canyon]

When doing cover's, can we get it so we can use persona's to do the covers. That way it is like people are actually doing covers with established AI artists

[pliant tendon]

Can we get a "Add to Queue" button for songs we're listening to akin to Spotify? Sometimes I want to add random songs to a listening queue that aren't necessarily part of the same playlist. 😊

[kind shore]

Playlists dont display on the profile page in a static predetermined order. their positions will change hourly.
As you know , Profile page song order is created by the amount of thumbs acumulated by each song.

[uneven lichen]

Add a search songs option

[wise urchin]

Hi! It would be great if you could adjust loudness of instrumental VS vocals, sometimes vocals are too loud or too quiet, although really good

[cobalt sail]

Hi, I have a problem now with the V4 update. I have tried to use the COVER function, but it no longer respects the vocal musical structure, since it makes the harmony as it wants. In that case, it is no longer a cover, it is just another simple random creation.

[peak knoll]

[Verse 1: Male] [Verse 2: Female] is this the format? My male verse is sung by a female it's not working.

[fair turret]

SITE FEATURE: Give us the ability to keep a home page or our favorites accounts(list/page toggled private/public). Followers are necessary but hard to find stuff when you get a bunch. This will allow us to have a handy shortist on site.

[tulip fox]

Notes Field for Song Generations

Currently, I add notes directly into the song title to keep track of details within each generation (as shown in the attached screenshot). However, having a dedicated notes field next to the title, accessible with a single click, would be much more convenient.

This would allow users to organise, track, and compare versions easily.

Would love to see this implemented!

[wicked ravine]

https://www.youtube.com/watch?v=EUzg__m39Sg

[tidal panther]

would love it if this was fixed.

[dusty relic]

There's so much awesome stuff going on with this program it's insane. I would like to suggest, if possible, improving the solo piano samples. It sounds very tinny, especially on higher notes. I prompted it with the description of Chopin's Raindrop Prelude to see what it would come up with https://suno.com/song/017fd1a9-15c9-43fb-be16-a68638b7eb34?sh=fHHgjAMj2jsHqrTs

[strong topaz]

could we get a feature where we can take pieces of one generation and mix it with another generation? kind of like a REPLACE WITH and you can choose sections from other songs within the same project... It's so often that one generation gets the chorus right and the other generation gets the verse right

[upbeat rock]

add "remaster" to the edit features so that we can select a timestamp to remaster sections within a track

[signal roost]

any one can help, how to get child vocals for rhymes, what prompt is needed in the style of music. always I am getting professional vocals.

[marsh lichen]

Guys, if you add some feature, at least announce it somewhere. For example, you added the Remove feature to the Edit section yesterday.

[twilit hedge]

It's been a week or so, not yesterday but yea, to your point... they should have announced any updates... even if just bug fixes or performance fixes.

[marsh lichen]

Okay, it just means that I only saw it yesterday.

[wheat dragon]

It almost changed my whole layout haha 😂 my song are all jumbled in my list and now the media player don't keep playing, when you switch between pages

[wheat dragon]

I would also like to know how to do this, I've been trying to make a DJ Intro with a child's voice rhyming

[low flicker]

you can create videos by api for the songs, users will add apikey from third party ai video generators, so when song completed there will be also videoclip created

[pulsar estuary]

this is redundant... why not just make the Style section editable?

[twilit hedge]

Still haven't figured out what the use for it...

maybe as note? to label the clip? maybe useful for that

[lucid light]

On the right side panel which pops up with the song lyrics, can we please have the datetime stamp back at the top? And the font size for the lyrics was perfect before; now it's way too large. And a link to the comments is missing now too.

[marsh lichen]

Maybe if they announced and explained it, there would be no questions like this?

[twilit hedge]

They did announce it.
#🚀┃updates message

[marsh lichen]

Oh, sorry. I didn't receive an update, I don't know why. Usually Discord informs about them.

[fleet cliff]

In the new create layout, the 'how much characters you have used' number should always be visible, not that it's only when your full.

[wary loom]

Anything official from Suno about this? can you share a link please?

[pliant tendon]

Wut, I've been treating this as a description box since it came out.

lol

Was the intended function describing song styles?

[modern summit]

i don't know who suggested the idea of replacing a section of a clip by another clip from another generation if this is technically the same song, cause i completely agree. i have an amazing transition and verse in one generation and i wish i could use it specifically in my song.

it's a shame we can't do that, especially if we are working differently on the same base song. we might have idea and perhaps soetimes extent differently and find something cool, but what if we want to mix things smoothly?

[pulsar estuary]

description box, comment field, text box, doesn't matter - people have been putting random text in style box anyway. only difference is style box uses hyperlinks. might as well make style box editable since most people want that anyway I think

[midnight shoal]

this is beyond unacceptable

please make this scalable, I'm losing my mind and squinting to type the prompt

how can the literal most important thing in Suno be this primitive?

hello?

[pine hinge]

why do ppl use 20 genres

[glacial vector]

Because you can, of course

[thorny dragon]

I really suggest this feature, make us free users get 70 credits a day, sometimes I don't even have the right beat from songs but I run out of credits.

[outer star]

hi 👋 Can someone help out. How to control the length of a song ? any suggestions 🤔 for example I would love to have 4 min results but instead I'm getting 1 min or 2 or 3 and really 4min . I'm using V4 model . and I'm promoting structure of the tune . for example

[Intro]

[Build-up]

[Drop 1]

[Breakdown]

[Build-up 2]

[Drop 2]

[Bridge/Breakdown ]

[extended Drop 3/Final Drop]

[Outro]

What I would love to find out is it possible to control length of the intro or Drop .... by setting how many bars it should be for example

[ intro , 32bars]
[ build-up , 16 bars ]
[Drop 1 , 64 bars ]
[Breakdown , 16bars]
[Build-up 2 , 32bars ]
........ and so on

Really appreciate for any suggestions thanks 😊 🙏

[junior vector]

Hi, I have a question. I'm a premium user on Suno. Can I upload songs to Spotify?

[cerulean coyote]

Is there an option to pay subscription fee Via Google play since app on Playstore?

[twilit hedge]

I heard it's possible. Look in the app see if you see the option.

[balmy zealot]

how many v0 users back from early 2023 would use a Bulk v4 Remaster for Liked songs?.. Would I be the only one that had gone back to v0 material and Liked tracks after the website had launched later? I have hundreds lol and I have to reupload every single one for them to work with remaster... as it is now that Remaster feature doesn't work and neither does the Editor for early v0 tracks made in discord. There is also no Get Full Song for the stuff I made threads for .. "Part 13" lol awe

[queen wadi]

Remove MADE BY SUNO From video for paying users

[carmine ruin]

Add back the [Extend] button in workspace view. I use extend way more than edit (which redirects to another page, meaning any written lyrics and such will be gone from the creation page).
Edit can still be there, there's enough room to have both.

[candid lance]

Real simple one. Can you make the “extend” option a one-click deal?

[fleet cliff]

can there be a way to sort by not just top and new but also most viewed

[latent scarab]

PLEASE get rid of the banner on the right side of the screen or make hide ACTUALLY hide it

[river haven]

🤖 Hello is Possible to ad an options to download the captions (lyrics) with the time stamp (Srt.file) to edit inside premiere i mean just sync or editing it with different styles, 👍 👊

[twin grove]

Here's some ideas for some easy UI Wins:

1. Port Filters to the Library. Should be easy enough to do, I'd think.

2. Change 'Snap to Beats' in Editor to a checkbox that can be enabled or disabled for all tools, not just Remove. (Hard Mode: allow easier snapping to beats 2-N (depending on time signature))

[slow walrus]

Can we have the option for 16:9 videos?
The 9:16 videos are way too often too narrow so that a lyric line is too long and is missing then

I make own lyric videos anyway now, but:
For those who want to upload these lyric videos by suno to youtube they will end up as a short with 9:16. and shorts get low views for many people because many people block shorts

[fleet cliff]

a direct message system would be nice for suno. it could be used to chat with other users about songs or styles that can be used, and can be used for help.

[undone knoll]

Could y'all please disable offclicks dismissing modals? I'm trying to look things up while creating a persona. If I click the wrong thing I loose whatever I've filled out.

[fair turret]

i hate when i accidentally navigate off the page and it gets wiped

[pliant tendon]

Has anybody suggested an actual dedicated language field on Suno? If I were to write "Bulgarian" or "Български" in said field, would it steer suno towards better pronunciation without getting confused with other languages?

Well I'm suggesting it either way.

[slate roost]

Lyric Checker/Highlighter:

This feature checks similarities of lyrics ensuring words match the previous line or lines only if they are repeated before pressing the create button. This will be helpful when there's only one one word error or line in the lyrics and Replace or Extend don't work that well when the flow of lyrics differ. Here's how it would look like.

[slate roost]

I tested out the bars thing which reaches to 3-4 minutes length but with the 4 minutes one could end up incomplete or complete depending on how many things you add. I also tried adding something like [Intro, 10sec] etc which leads to a 2+ minutes length track however this will differ per genre I assume

[crisp rock]

dear Suno, when do you want to understand that we need more song rating possibilities than the thumbs up and thumbs down? Many proposals have been made already! Why it so difficult to help us ?

[west elk]

This (image) is annoying a sh!t. Make the box expandable so I can see the text I've entered instead of hiding it behind a bunch of suggested genres and the character counter. Good grief, this is infuriating.

[twilit hedge]

There's a new UI update. See #🚀┃updates message
The new Create UI will have this fixed. You see it on your browser soon. Some users already received it.

[west elk]

I don't see anything in that announcement about fixing this issue. Have the people that already have the update confirmed this is fixed?

[twilit hedge]

yes it seems fixing it for me

[limber rose]

When using the full song option to generate lyrics, the generated title isn't recorded. It has to be done manually. This has been happening to me for the past two days. Please correct it. I'm creating different songs with repeated titles. leave it as it was

[slow walrus]

rightclick - Create - Extend

keep in mind though that this one is inaccurate with just 1second step accuracy of the timestamp
With the new Editor you also can see at which point of the lyrics the extend would start. So the new extend editor has its advantages

[twilit hedge]

I think the Create > Extend on the new UI also has a waveform, at least on Desktop version

[slow walrus]

yep and with sample accurate accuracy 😄

the zoom in is incredible 🙂

[pine haven]

Hello! I would like to suggest an Affiliate Program or a Creators Partner Program! Definitely something to look into 😄

[charred turret]

A good idea to implement is that to add our own voice , and then the possibility to make a personna with that voice and make songs with SUNO with that voice

like this

[haughty dome]

Add a Caption button = It was an earlier request. It's an optional feature, so that people who want to, don't have to put info about a song within the Lyrics section.

[upbeat mantle]

how about adding a generator title like for the image that goes after the song and make a fitting name from it

[twilit hedge]

I remember it used to do that at one point and when I forgot to put my tittle... it gave me one and I was confused whose song was it 😀

[dusty relic]

Hey, you have the liturgical text from the "Dies Irae" as being copyright protected. The text is 400 years old. I was kind of hoping to do a Dies Irae or Lacrimosa 😄

I really wish we could actually interact with the suno as a chatbot as we are creating music. It would make it a collaborative process with the AI and you could specifically tell it what to do and where

[violet sierra]

Come on, For You. You gotta know I like videogame covers by now...

[oak steppe]

In the character module, the character characteristics are set and the song standards corresponding to the characters are selected. However, the song produced under the character style is still the lyrics memorized by the previous songs when the lyrics are modified, so why won't the lyrics be recompiled into a new song? What is the difference between setting characters and setting prompt words? Or how do we get this character to generate songs inside other lyrics? Is there a custodian or a technician?

[haughty dome]

It may depend which country you are in, or where the composer composed. Suggested reading: https://www.mozartproject.org/mozarts-requiem-public-domain-or-copyleft/

[weak heron]

We need more control over how to create a song.

This includes, for instance:

1. When the song should end or how long it should be.

2. In which parts of the song I want lyrics.

3. How strictly Suno should follow my lyrics.

4. How strictly Suno should adhere to my chosen music category.

5. Whether to use a male, female, or both voices.

6. How long do I want to extend the song.

7. How long should the song be created to hear if you like it.

[gaunt wave]

We need quality, nothing else

[crisp rock]

While other music AIs permit up to 3000 characters or more for the prompt, Suno allows max. 200. Is there any reasonable motiv for this limitation ? I cannot find any.
200 is way too little! Why should we let the machine to control our songs by random instead of us influencing the song in the way we want?
And why is Suno's model interpreter not even listening to that 200 characters ?????

[sleek vigil]

Well I cannot really see it follow even 50 symbols reliably so I guess more wouldn't make it better

[crisp rock]

you cannot take a conclusion about others only from your own experience. Sometimes you need only to name some instruments, but there is no place for that

[sleek vigil]

that would still make around 25 to be honest, do you type stuff into the lyrics panel as well to guide it like with [tags]

[tepid trench]

Are you using "Custom" mode? You can name genres, moods, instruments etc. in the "style" prompt. And then you get 3,000 characters in the "lyrics" box as well for words, or other tags you want to include (e.g. "[guitar solo]")

[crisp rock]

I always use the custom mode. It is limited to 200 chars. It is a struggle to add tags in the lyrics. You have to do it for every song. Other AIs don't limit that field and they also have lyrics tags.
So why ?

after 100 days, we still have no control for COVER. Hmmm...
Cover does only what it wants! You cannot even tell it to make more similar and more different than the original. Nothing, no control at all. Random rules and the credits are wasted. Is this on purpose like this ?
All other AIs offer a slider for the COVER where you can choose what you want. Only Suno does not have any.

[oak steppe]

It's really annoying. Why would programming change lyrics make my music shorter? I only changed one piece of the lyrics and replaced some of the lyrics with the function of the replacement part and re-arranged the song, and I actually shortened the part that I didn't replace. Why did this happen?

[dusty relic]

It's insane that anyone to me that anyone could hold control over anything composed by Mozart. "Dies Irae" goes back significantly further than that. The music (not the lyrics) shows up in random music composed to this day. The theme to "The Shining" is the "Dies Irae" music and it is used in many other places. I was just hoping I could use the lyrics. Thank you for the reading though! Here is the wiki about Dies Irae if you were curious https://en.wikipedia.org/wiki/Dies_irae Also edit to note: AI isn't quite where I want it to be to collaborate with AI to make my own Requiem or Dies Irae anyway since it doesn't seem to handle too many meta tags at once. If AI does its own thing with dynamics, tempo, and orchestration then anyone could just plug in the Dies Irae lyrics and it would be ALL AI composing just another Dies Irae rather than the user inputting their own thing, if that makes sense

[upper canopy]

is anyone from the webpage team online?

[upper canopy]

i'll take that as a 'no'

[sleek vigil]

it's saturday and its overall kinda unlikely someone checks this chat exactly the moment you ask

[upper canopy]

indeed

[autumn basin]

ok the new update or whatever it is really sucks.... suno wont add lyrics if there is anything "evil" or so. i dont wanna make music for 6 year olds. if this site is getting woke i go elsewhere

[mild cairn]

Please bring back the ability to see the full STYLES added to a track on the right bar when clicking it. Now I have to open the track in a new tab to see all the tags I used. If you want a cleaner aesthetic - then just add a 'Read More' thing so we can expand the list of tags we used please. This was such a useful feature to have access to on the main page when creating tracks.

[fast fog]

Anyone else find changing playlist imgae or a track image buggy on browser? I can cange a track image on the ios app, but ther eis no option to do so for a playlist. In the browser, it just flatout doesnt work. The playlsit image just freezes when i hit save. For the tracks i get "Network Error"

[slow walrus]

5 needs massive improvement yeah. shouldnt be suggestions, should be commands, also I want ability to change persona on an extend PLEASE. This is absolutely mandatory when you have 2 singers and I need this function badly!!! 🙏

For options affecting the generation like variety and other options which other platforms have - only if they have an "auto" switch. Because this kills for me generations on other platforms too because you HAVE TO use these sliders, because they have no Auto function and then are always at neutral if you dont touch them, so without an Auto function I would not like it

[midnight whale]

V4.5 coming in May/June

[patent glen]

An opt in/opt out of beta testing through your account settings. I love testing out new features and such and I love the new create page, however a lot of these changes come with a whole heap of bugs and have ruined creating songs for me as my persona has completely broken among others I've seen/spoken to. It feels as if we are forced into beta testing which I understand isn't the goal with the changes, however from the bugs that come with the changes it's why it feels that way.

It's why I believe an opt in/opt out setting through your account is the best way forward

[jagged spruce]

I was just thinking about that. 😅

For context I only have the mobile app on my iPhone to use.
I’m burning through my trying to get female vocals and the renders are almost always male. Which sounds okay for the style, but the lyrics are written from a female perspective.😅
suggestion:
add a vocals prompt box with the style, lyrics, and title boxes or some kind of toggle like the Instrumental feature.

[shell merlin]

Hey guys, it would be great if we could make the interactions predictable; lost a bunch of credits because of changing one thing which suddenly changes 3 other things.

Case and point, added a persona to something I was working on, and it changed lyrics, styles and excluded styles which I only noticed afterwards, and so it not only lost credits but also time having to recreate what I worked on.

It's great to have a Persona addition change other stuff if those are empty, but when working on something it's quite cumbersome working with various features not knowing what each one will change and when. 🙏

[fleet cliff]

make it so workspaces can actually be deleted instead of just having them in trash only

[fair nimbus]

Stats and analytics. I want to see my top listeners, and what songs are getting the most plays and likes for a time period

[jaunty raptor]

I got the idea to add a feature that allows users to change the words of the singer even after generation

[woven pilot]

When the song should end or how long it should be. NEEDS TO BE AN OPTION!!! every damn song ends up 4mins

[woven pilot]

this is crazy

[zenith needle]

PEOPLE lets stop with the SUGGESTIONS alltogether until they fix it for good. No point in adding anything until V4 works flawlsessly with good production quality level.

[clever falcon]

I appreciate the work that has gone into Suno AI! However, I noticed that when extending a track, it always generates two versions instead of one. While this is useful in the initial creation stage, it often leads to unnecessary credit usage during extensions. It would be great to have an option to generate only one version when extending a track. This would help users save credits while still refining their music. Thank you for considering this!

[crisp rock]

Suno has the uggliest FONTS ever seen in a web application. Uggly and crowded, squeezed fonts. Even a 1st year student in graphic design would do something better. And also some weird colors.
It is not nice at all to watch the User interface. Take any student, pay him with 300 $ and will do much better!

[upper canopy]

Hello Web Team, - Since I am a content creator, the 'Home' page is quite useless for me. It appears to be made for consumers, to help people find music to listen to. However I am a creator and not a consumer. I am interested mostly in my own stuff. I don't want to listen to other people's stuff. Instead I want to see statistics of my own songs, for example my songs ranked by total view count and what has received attention recently. Could you add that to the 'Home' page? Or maybe, could there be a separate 'Home' page for creators? Btw, I assume that most people viewing the suno.com 'Home' page are creators and not consumers, with similar interests like mine.

[crisp rock]

We need much more often to edit the song title than to edit the audio of the song . Why don't you put (also) a edit button there for editing "song details" ????

[crisp rock]

Why does the edit lyrics button has to be AT THE END OF THE LYRICS ? Do I have to scroll first the long lyrics to be able to press a button?
What is the reasen behind this waste of time for the user and lack of visability?
Which programmer is programming such non-sense????

[sleek vigil]

you could also use this if it fits your fancy more i think the lyrics edit down there is good because i rarely need it and it's out of the way for the rare occasions where i need it

[slow walrus]

Especially since the notifications bell is mostly out of order in regard notificate about new songs from people you follow:

Can we have the "Following" list a bit more easy accessible than somewhere in the middle of the "home" page?
It should be easier made to let people see when new songs are released. Nobody did even know that I did upload a new song because bell broken.
Same for me: I did open the following list - 14 new songs - bell: 1 song
I can imagine you want to prevent new song spam when you follow many people, but come on: more than 1 song should be doable by the bell. So I think the bell is just broken

[leaden marlin]

Aha, I see there is indeed a feedback channel, so I will repeat: Suno. PLEASE move the 'song deleted' popup away from being directly over the music controls, then I have to wait for it to go away to keep working. 😔

Support for this update shared by two other users in #General just in the short time it was onscreen

Also, allow us to select multiple items to delete, this is AI generating, having to delete one at a time is tedious,

(or is there a way to do that with a hotkey etc?)

[slow walrus]

thats possible in the library, click the first one, hold Shift key and click the last one
trash bin for some reason doesnt work though. But I just keep the trash bin full

[leaden marlin]

Oh neat thanks!
Library browsing could use some userfriendlly updates. Grid view, collapsed list view (no art icons) with all gens or way more gens seekable with a scroll bar, plain text with the ... drop menu.

[slow walrus]

oh looks like trash bin meanwhile multiselect works now too

[leaden marlin]

I just discovered there is page jumping, you can type in the page you want, that will make it less annoying when updating old things which spawn on page one then jumping back over to the last page

organization folders for the library sound like a good idea, not just a single master feed which gets quite long. keep all my ' genre A' stuff in one library, and all my 'genre B' stuff in another library directory etc.

[slow walrus]

I use "Create" as my library

there is workspaces

which is like folders

[marsh lichen]

I asked several times to give us a possibility to select one version to be created at least for extensions, replacements, and remasters.

[midnight whale]

I got the best mf suggestions for v4.5/v5

1- To be able to generate sound effects and to be able to add them in specific parts of the song.

2- To be able to upload 2 reference audios, one for the melody and the other for the voice.

3- To be able to make collaborations between Personas and choose in which part each one sings.

4- Add a reference song with a YouTube link.

[slow walrus]

to be able to get the correct gender in a verse reliable, same for duets
to be able to change the persona in an extension
to be able to apply any extension on any song on any timestamp
to be able to download stems which are for real stems and not filtered stems from the final mix

how about these? 😄

[leaden marlin]

not bad, i tried that out. so that will work for now. should not be at the bottom of the sidebar, it's an organization feature not a song option, so it should be at the top of the main panel

or on the very leftmost panel under Library

[leaden marlin]

the last one idk if the AI works like so, but that will be lovely once we have the tech
the rest, yes, especially #2, i'd like to see genre-bending music.

afaik the AI generates the tracks together, not individual instruments sequenced in the black box

[leaden marlin]

just release a desktop sequencer, basically - much easier to make it this powerful. 😄

I prefer to use Magix Vegas Pro for offline fine tuning, FYI, do recommend.

[haughty dome]

works for me in browser (Gener8 based on Chrome)

[fast fog]

Yeah, i hit the save button and that greys out but it doesnt save i have tried chrome safari and duckduckgo so far

[fast fog]

Oh and i am uploading my own image, not generating one

[slow walrus]

the rest, yes, especially #2, i'd like to see genre-bending music.

Actually my song "Lucy's Totalschaden" I got it happening, click my discord profile to see my suno channel if interested.

But I want it more for the use case "having 2 singers" because I have 2 singers and one of them would always have a new voice because I cant change the persona in an extension

[sinful cedar]

I am experiencing an issue with my Suno interface that differs from my colleague's, even though we both have the Pro version and are using the same software version across different browsers.

Specifically, my version is missing certain features, such as the "Add Audio" function, which is available on my colleague's interface. I have attached a screenshot for reference to illustrate the differences.

Do somebody know what is the rootcause?

My interface wih missing functions

Expected Interface

This issue is linked to my account. Verified with his computer and my account

[warm tide]

Yo is there a chance to toggle the voice? Just the music and lyrics

[tropic cypress]

能不能和团队反馈一下，提供一个局部重唱的功能。就是现在的replace section+remaster，这样子哪怕歌词唱错了，让它重新唱一边就好了。
消耗积分都没有关系的，不然现在改歌词太消耗时间和精力了，而且还改不到位，心态都受影响了😩

Excuse me, may I suggest that to add replace section + remaster feature on Suno? Because I had struggled for a long time to fix my song based on the unsatisfied Chinese pronunciation of the results.

I wish to fix certain part which has poor Chinese pronunciation perfectly

[void sand]

Can we also get stems, not instrumental and vocal, like drums, bass, guitar, etc OR OR, midi files that go along with it

[proud verge]

I'm native speaker of Estonian language. It is very small nation. I'm willing to audio upload my poetry to improve the pronouniation. Is it possible that Suno clings to the pronunciation the author of lyrics is giving? It would encourage people to upload their lyrics, singing or reading it. It would massively support the small languages or people who prefer to sing with their own accent. Even members of large nations sometimes feel that pronounciation is the issue. Tries are using up a lot credit though, when it's struggle to get the lyrics sound right. Disadvantage to us. And also, when I'm deleting the song I'm willing to give feedback, why. Unhappy with music/lyrics at least. Currently Suno don't know why the creator disliked the song.

[granite rose]

So the lytic replacer is nice, when it works, but its not what I (and I expct many others) want. Its way too much work, and then bugs make it even harder to use.

What I really want is just to dump the lyrics, Let me rewrite them, and replace the old lyrics over the whole song with my rewritten ones,

I dont know how you are building the song, but my guess is that there is a wave form generator for the music, and and an LLM whose output is fed into TTS and then something like an auto-tune.
Then the tracks are mixed.

If you kept all the parameters for the generators, it shoiuldnt be hard to sythesize the singer again and just remix?

[sleek vigil]

If they were done separately you'd be able to get clean stems which we don't have

[slow walrus]

the current stems are filtered from the final mix, so like the usual audio editor does - extracting the center channel, then there is still little bit voice heard and the instrumental in bad quality
thats exactly what suno does too with the stems. So tbh I dont see a proper bass, guitar, whatever stem xD

[void sand]

what they could do is render them separately before the final mix so its better quality but it would use more cpu and thats why they don't. But they should add it so its better quality.

[slow walrus]

proper instrumental and vocal stem would be amazing, but these filtered ones are problematic 😕

[wet arch]

LOVE what yall are building, but here r 2 UI changes im not crazy about...

1. having to go into a folder at the bottom of the page to "upload audio" is very clunky
2. whats the reason behind switching how the song populatde your create feed , it use to generate at the botttom of your feed, now it generates at the top. i dont like that, id love to at least have the option to choose top or bottom.

[wide fossil]

Suno. you should get an age verify as you stop adults like me from creating songs with swear words or sexual songs

[twilit hedge]

There's + button next to Create button, you can access Upload from there too. Still not one click as before but better than going through the hidden options.

[leaden marlin]

The video never updates when you edit song details

even if you regenerate

[twilit hedge]

report in #🪲┃bugs-web

[wet arch]

i dont see the + button, am i missing it ?

ohh down at the bottom "create" i see it now. thank you. that is at least a lil faster 🙂

[twilit hedge]

yeah

[long robin]

I would love to be able to use the Workspaces filter in my Library if that's ever going to be added I'll be stoked. Or have the library sitting on the same page as "create" so I can get to things for cover/extend without losing what page I'm already on.

[dusty relic]

Probably a low priority suggestion, but I think it would be cool if there was a credits page on the suno website showing us whose voice samples we are working with and maybe a song or two of them singing for real. They are very talented and deserve credit, in my opinion

[granite rose]

I found a really good one. You do have to pay but its not unreasonable. Id give you the name but Its on my machine at the office. But I was impressed. It actually pulled the auto-tune ring into a seperate stem 🙂

Im sure there are multiple layers to Suno's results. it may not be actual stem, but Im sure there is a seperate lyric generator (probably llm) . It would be nice to get access to the layers and edit them individually.

looking for what the service was... moises isnt bad but I think maybe I found a better one?

Okay, its moises. If your looking for stem separation, give it a try. I was impressed with how clean it was.

[dreamy depot]

I would like to use the editor to combine 2 songs. For example, take Female vocal from version 1, verse 1 and male vocal from version 2, verse 2 and make one song. So like a cut/paste section. I absolutely support adding sound effects and absolute driving specific singers per verse. If I could combine 2 personas in the same song....omg.

Bulk delete of all the stuff I don't want in my workspace.

[slow walrus]

click first, hold shift, click last

just like in your windows explorer

[tropic cypress]

When can I update the one-line encores😋

[royal lichen]

make a feature which can make clean version of Existing song, by changing What word we want. So we wont lose current good song

[gilded python]

Is posting a link for a game that features the music allowed?

[steel blaze]

I dont know if it exist or if i missed it. But it would be nice to be able to search the explore section. So if i want to know how "disco cloud rap" sounds i can have search bar instead of just the wheel turning.

[wintry perch]

Sometimes I have a song generate and the voice sounds almost identical to one of my personas but I didn't use that persona. I love it when that happens because it's like it knows my style. I really wish I could assign a persona to it after the song has been generated when that happens. That way I can like click on a persona and find all my songs by that persona (including the ones that I assigned the persona too) Right now you can't do that. I think this would be really really helpful!

[granite rose]

It would be really nice if, when I removed a section, the editor generated a very small bit of bridge to get from one section to the next cleanly. I've tried using replace section over a small area around the splice with varying success at pop removal.

Another thing that would help would be if I could over/under transitions.

[twilit hedge]

If the stitch wasn't smooth or clean on a small bit of music... then try Remaster, more often it'd smooth it out

[gray sparrow]

"Feature Mode" allows you to use multiple Personas on a song. You can prompt Suno which Persona is singing/rapping by entering the Persona's name [Allan]

"Adjust Mode" Allows you to tweak up to 10 words. In case of a misspelled word, Suno can make a quick fix

[frozen sparrow]

How about pinning an album we want others to know about on our profile like songs

[granite rose]

What if suno pronounces something wrong, is there a way to hint it? In one song it's pronunciation voice as vice to make a couplet riyme but I'm fine with a half ryme. In another case the word is me but it's singing my. Probably because the grammar is poetic.

[shut viper]

Multiple tracks in editor based on stems. So I can edit just an instrument or just the vocals of a section

Export of midi files so I can continue to edit in post through DAW

[hot rain]

Is there a way for others to work on a playlist together like adding each other’s music on a similar topic yet?? I have a the Backrooms playlist and would be sick if others created and put their songs on there too. I’m obsessed with this creative direction 😸🫶🏻

[crisp rock]

I would have not been difficult to have the idea that we might want to place a song in more than one playlist. But this means we have to do this for every single playlist separately. Nobody had the simple idea to add a checkbox in the add to play list menu so we can do that in a single move.

[hollow locust]

Please fix this.

[sullen glen]

it would be cool to ask for specific instruments for a song, like for example, if i wanted ONLY flutes and rain sounds, it'll be only flutes and rain sounds. also for a way for it to be a bit better when making ambient music, i try to hint towards the atmosphere i want, but it doesnt really listen

[sleek vigil]

I'd love to have multireplacements, like this, i often get 2 (or more) ideas from the replacement i like and would like to do something else first before I decide for one, but if i apply a different replacement i can't get the others anymore so some sort of multireplace and select several to all apply them in one go would be neat

[sleek vigil]

Has been suggested in various variations but here, mini persona to summon different voices
I'd personally use such a thing also to make a line whispered or shouted and would have different personas of my same singer
no idea if possible but i like the idea

[slow walrus]

yes being able to change the persona at least on an extend would be nice. I have a male and female singer, and one of them I always cant give them the persona

[haughty dome]

Hey! the right hand panel has gone! for Home Create and Library unless you're playing a song.

[twin bear]

⚠️ Allow Personas to be used on Extensions and Covers ⚠️

[wintry perch]

how are you highlihgting and replacing more than one section at once! thats something we can do? How? teach me oh great one!

[sleek vigil]

Paint

[wintry perch]

paint?

[sleek vigil]

paint.

it's just a mockup to bring the point across, the elements are just copied and pasted in paint

[twilit hedge]

Yeah, that was just a mock up screen that @sleek vigil suggested, but if it was implemented, I'd think hold click + shift + drag the selection would be the possibility.

[wintry perch]

Oh man I would love that so much. Devs pretty please this one Vishanka said. I'm saying it with my whole chest!

[fossil trail]

Yo, i think a "train style" option would be kinda fire. Bcs sometimes your songs have a unique sound to them that you wanna keep, or you might be doing a more niche genre, and this would help maintain a good sound imo

[tropic cypress]

We urgently need a one-line lyric redux function， one-line lyric remaster

👀

I think the editing functions should be made more powerful, such as repeating a single sentence, adding melody to a certain paragraph, and so on. These powerful editing functions can make the song closer to the effect we want.

[compact plume]

What about using persona for covers 🥺

[open lynx]

Distribution features please

[trim lantern]

I understand you maybe you are an extensive Suno user on mobile when you look at the website layout you don't find a hierarchy and consistency! But yeah its strange and makes users lose time, finding the correct way to interact with the stuf.

[wet arch]

yes but i mainly use it on desktop. still same problem. i just wish they had an option to have either or.

[manic zodiac]

This is not an efficient way to track feature requests. Does a github/jira/gitlab exist for feature requests and voting?

If not, this would be killer for everyone. Tracking votes, merging duplicate requests, helps prioritize your focus.

Yeah, I can't find any actual feature tracker.

Just my two cents, but tracking suggestions this way is literally impossible. Github is very easy to sign up for, allows users to "vote" for a feature request using emojis, and can be filtered and tracked using a project board.

You guys at Suno should really consider using this. I'm discouraged to suggest any features simply because I know the chances of it being seen and for other users to show support is non-existent.

[upbeat mantle]

Will there ever be a lyrcis correction with the grammar added to the lyrics box?

[hardy anvil]

I don't know if this suggestion has been made yet but Workspaces should have their own shortcut on the sidebar between "Create" and "Library"

[supple current]

1. do background checks on who gets to puppeteer woodhouse (in discord) and ensure they are mentally fit for purpose
2. add delete button to clips inside the editor to reduce clip pollution

[supple current]

yea a roadmap on trello or even free-er ones would be neat for community input

[granite rose]

i do love what you have done here, but there ae a few annoyances. One of them is that "replace a section" seems to love to extend the tie of the replacement. My guess its to give yourselves more room to do what neds to be done to make the old/new/old transition seamless but it means every edit I make seems to lengthen my song. Whats worse is it civers the lyrics by inserting repeats of existing lyrical passages.

I always give my replacements at leat a couplet before and a cuplet after as belnding. I wish there was a way to say "contrian this to the length of the replaced bit".

Related to that, most of my edits are polishing Suno's attempts to write lyrics. It would be nice if there was a way to ONLY replace lyrics while keeping the music track as is.

[median wave]

has anyone tried the "remove" function in the Editor? It appears to delete upto 5 seconds of audio...(just tried it to remove an intro that was too long...works! - song ends up with the "crop" status)

[median wave]

I do the same thing with the replacement editor, give it a liitle of text either side of the change, but not too much or it will rework things you don't want changed. If it adds more lyrics than what I wrote, I junk that section with a thumbs-down and retry. After 5 retries, I'll stop. One thing that happens is it can change the displayed song lyrics (incorrectly, you have to listen back the song). If this happens, re edit the displayed lyrics to match the audio, then try the editor again, since it is very sensitive to mismatch between textual and audio lyrics.

[granite rose]

yeah ive seen it lose a line of lyrics a couple of time

[granite rose]

yeah ive used it a few times. Suno has a bit of a fetish for long pauses follwoed by a repeat chourus that I find awkward and I generally tighten those up by deleting som of the pause or just the whole repeat.

In general V4 seems to be a cleaner sound. Which can be a problem. Ive tried remastering a raw hard rock peice. 3.5 really got the distortion and hardness of the vocals. V4.0 doesnt seem able to reproduce it

[granite rose]

So I sm at tht annoying place i mentioned. There is a transition near the end of the song Id like to make a bit smoother, but both optons its giving me blow out the end og the song by a good bit. I like the ending right where it is.

[rocky jay]

Anyone know if there are any plans to increase cover from the 3:10 max timeframe?

[marsh lichen]

Just create a transitional section with the Replace feature. 3 seconds to highlight are generally enough. If it's not enough for Suno, it will just make it longer.

[elder flint]

Super idea, probably I'm not the first one - if you can make "minus", split voice from **created **song will open new realm of karaoke (ie listen track with signing and record your voice over music)

I don't see general feedback section, so my applause here to the Team and new editor, it's very nice and much better than before. Awesome! 👏

[civic ether]

I have a pro plan for the last two months , generating Hindi songs is fun but there are lot of flaws in pronunciation, and edit option can't give any good word replacement, please look into it.

I would be happy if I see Indian language trained model is more powerful to create songs

[lost quartz]

We need 2 separate models. 1 model is for generating. The quality is pretty average but the note choice, melodies, and rhythm are awesome. V4 is pretty damn good at this already. Then a second model which all it does it remaster it, so that the bad guitar etc can be awesome. I want to uplaod a demo and get it to remaster it as if I had it to a band to do the song properly. Guitars actually sound like guitars and not the atrocious degrading we have now where they don't even sound like guitars. No shimmer. Don't try to make 1 model do all. One model for remastering and covers, one for initial generation. Like a Songwriter model, and a Session Band model. I want to upload my demo and then get it remastered, maybe even do pitch correction on vocals.

The other thing needed is for us to be able to do like Audacity, or a video editor, where we can put any cropped section of any song in any order, and then once we have it all lined up, we hit render and it joins all the parts together. Even if it's just 2 parts at a time. So rather than having the first generation, and part 2, and get whole song, we specify wan is the start section, and the part 2 section, and then click get whole song. The cruel thing is when an extend generation generates 2 brilliant lead solos and I want to use both but I can't. So I have to do it in Audacity which means I can't share it on Suno. If we can join then we can upload 2 parts of a demo and then join them so Suno can do a cover.

If you don't let us join any clips we choose, then at least let us do it in Audacity and then upload it in full back into Suno.

If we get those 2 things, Suno will be unbeatable.

[wary hazel]

hey everyone! what command (Style of Music) should I enter to have children's vocals? I would like to create rock/metal but have children sing?

[clear folio]

I agree with everything I red about the variation tool and join to the request: it could be a powerful instrument to make a song complete. Personally, I'm working hard alone to make a song with more voices, but I cannot take the same "characters" (at least one is lost). Second: I have one or two songs perfect as a basic, but just one section is repetitive and flats all the rest. I tried editing with the tool, but, despite my indications, the parts come out very similar most of the times. Then, I can't split a section into other parts, to add a little dynamics, I mean: I need to change one section to make a variation, but then I need that just the last 2 lines of it reprises the melody to launch the chorus. I can't separate that 2 lines, so I have the variation, but it doesn't fit anymore with the chorus (in case it's quite different). I don't know if it's clear, but it's very limitating. It's a pity when you got the "almost perfect" piece, but you have to give it up, cause it sounds really too flat in the middle or you have to take it like this, while it could be perfect with just some little adjustments.

Why not making the tool able to split the sections into others or add new ones (with or without lyrics) and why still not introducing the covers with your personas?

I see A LOT of people asking exspecially the last one.

[halcyon vale]

Suno should add a feature that allows users to upload their own audio—like original instrumentals or songs (excluding copyrighted material)—to be used as training data on the client side. This would let users expect what they want on Suno’s output based on their own style.

[granite rose]

This is an observation less than a suggestion, but please continue to keep older models available as they seem to have different limitations. Most of my 3.5 songs come out better remastered on 4. They are cleaner, the voices more distinct, etc.

But 4.0 seems to be unable to create hard, dirty rock and roll. I have one song where the edginess is absolutely necessary, and a half-dozen attempts to "remaster" have come out bright and cheery... which is not the intent.

[clear folio]

Agree also to this, I asked the same time ago. Given that Suno can recognize a copyright song from a few seconds, there should be no problem in inserting an original audio/text. It would be great for a real author to have a band arrangement for a real original piece. I have a lot of lyrics and basic music, but I just have my piano and I can't find a band, THIS would really be a way to help musicians. For example, I would have liked to use my music for the songs I mentioned before; ok, some of them went out beautiful, I can't complain, but it's not my music, anyway. = (

[karmic quiver]

Did you ever find a solution for this? I'm having the same issue with updating images on the website.

[fast fog]

No. Afraid not. It’s really frustrating

[karmic quiver]

Blah!

[dry rapids]

||e||

[fast fog]

My thoughts too

[karmic quiver]

My work around was to add it to Canva and use both the Canva and Suno iOS apps. 🤷🏻‍♂️

[fast fog]

Canva?

[steel blaze]

Jag jobbar på ambulansen och denna låt är för de som lider av depression och inte orkar längre. Mött en del under åren och ibland så slutar det otroligt tragiskt.https://suno.com/song/c2989e08-d6ba-46ea-b9b4-35aa4d2f120f?sh=o0eSh21Jmiu5mDqR

[karmic quiver]

Yeah.

[karmic quiver]

I downloaded it from Canva to my phone photos, then uploaded it to Suno

[fast fog]

What is Canva?

[karmic quiver]

It’s a super basic design app.

https://www.canva.com/free/

[opaque gull]

Can we sing a few lines and upload them to suno, so that suno uses our voice to sing? If not then this would be my next suggestion.

[lavish sleet]

Help: My issue is that when I run out of credits but still have two weeks left on my plan, my only options are to buy more credits or just wait it out. Why not let me renew my plan early? I’m ready to pay $30 right now, but I have to wait because of the subscription Terms of Service. I get that you make money off top-ups, but that’s a short-sighted way to treat loyal subscribers who keep supporting the platform. I see this as a bug, can it be fixed?

[granite rose]

I suspect its not that simple. I have no visibility into Suno, but if I had to guess, I would guess that the lyrics are generated in an LLM and then sent to a TTV to turn it into speach which then goes into something similar to an autotune to turn it into a sung line.

[torn pawn]

Devs - when you crop a song, the persona UI gets removed. Can this please change, ty 😦

[marsh lichen]

Not only when you crop but basically when you do anything with a Persona song, including replacing sections or extending.

[median wave]

I treat those as "alternate endings" and usually I can cut it off during the pause, and if there's any residual sound, fade it out.

[granite rose]

In some cases I cut it off. If I really like it I cut out the pause and then regeneraate over the seam a few times to get it to clanly transition

[granite rose]

I would guess thats because ther persona is already known for the song?

[marsh lichen]

The problem is that you don't see anymore that a persona was used to create the song.

[frigid owl]

What if you could add a weight slider option to a song cover? For example, let’s say I’ve uploaded a sample track and want to create a remix in a completely different genre. I could set the weight to 40%, meaning the generated track would retain 40% similarity to the original. Right now, it feels more like a lottery, can be nearly identical or completely off.

[modern summit]

i don't know why it's not a feature but it would be nice to allow songs attached to a persona to be permanently deleted. or at least maybe put a screen warning that the song is attached to a persona, with a confirmation to delete.

[safe latch]

Guys, the weird narrow serif font used in different parts of the site has got to go lol. It looks so bad!

[sick lion]

This is a suggestion site-wide:
Concerns arise over copyright and protection of individual songs, and anyone can extend any public song within the site, and post it to their account (thus claiming rights?). Similar to a "friend request" on (shudder) social media sites, I propose an "extend request". Once initiated by the third party, the original song owner will have control over whether or not they can "transfer rights" of their song over - dual rights for the same song cannot exist if the piece was independently created, so the extend would be considered a derivative.

Regardless of the murkiness of copyright as I put above, it could help people protect their works from being lifted without consent within the site. This is especially important for lyric writing - people have already taken steps to protect their songs by omitting the lyrics in the final version.

[wispy marlin]

@tulip sinew i use audio with all of my good generations as they ouput exsctly as intended.

My issue? Can’t put public!

https://suno.com/song/3ef2287a-476f-4b7a-8961-31cc6a9af58e?sh=wCiMmxopbgZ8zTA2

hear me out - can we get a acceptance moderated check for this or so?

Alsio: push this one public if you like it - im stoked by it 👍🏻👏🏻

[modern summit]

i don't know who suggested it so i'm following: being able to stitch together different parts in clips. like in one clip you have a good chorus and bad verse, and in another you have a good verse but bad chorus. if they come from the same base seed/base song, then we should be able to stitch them together, like we are adjusting parts and then we get a full version where the ia applies some polish to the clips transitions and stuff

because of this idea i found myself digging through older versions that didn't exactly work but had parts that i loved and i decided to crop them and keep them somewhere

[willow fog]

General feedback for all platforms Suno is on- can we get a shuffle/random play option? If it already exists, I don't see it so make it more obvious, if that's the case. Thanks!

[manic zodiac]

Please create a proper feature/suggestion tracker. I recommend GitHub Discussions. github.com/features/discussions

[jolly ore]

I was looking for this feature as well, attempting to upload audio split 50/50 and using the infill feature to bridge the gap only yielded a huge quality loss. It seems any uploaded audio is just compressed and miserable. It would be amazing if you could just "finish" a song seamlessly.

[latent ruin]

Your resume feature plays at max volume regardless of the current settings

[open rock]

@latent ruin can you provide reproduce steps

[latent ruin]

Let me attempt real quick

have song playing at volume below 100% -> F5 page -> click play button after the playing bar appears again

[open rock]

checking

weird, I can't reproduce

what is click play button after the playing bar appears

how to make the playbar disapear

[latent ruin]

Oh what the hell? I thought this was the bug report channel, sorry

[latent ruin]

by refreshing the page with f5

It loads last of everything, to my knowledge

[open rock]

right

so refresh will show the playbar right

[latent ruin]

yeah

[open rock]

can you point me what is wrong in this reproduce step

[latent ruin]

nothing, but I can reproduce it consistently

Might be a browser issue - I use Firefox

Wait, hold up. It's not the volume bar that's set to 100%

It's playing audio that is

It'll appear to be set to what it was, but the actual audio is maxed out

[mystic coyote]

What's the point of hiding the trending topic now?

But now nobody sees it anymore, they only see the top 5 at the beginning, esp it was something that gave visibility to people who are active daily, it seems to me a serious mistake on the part of suno

Before you saw the big thumbnail, now until you open the song you see it small

Obviously I don't watch the global, but I fight daily to stay on trend in Spanish.

How do you open that window?

[jagged chasm]

[Type: Suggestion]

When you click on a song in the "Library" or in "Create", a menu pops up. Please add an "Open in new tab" button.

It is quite frustrating always being redirected to the song you clicked. I have lost lyrics like this multiple times.

[tame fiber]

i suggest to remove the new homepage as fast as possible friends

[sage mirage]

I give the new homepage my seal of approval, who I follow is front and center, as it should be

[open rock]

issue is fixed

thank you for the report

[latent ruin]

Thanks!

[leaden marlin]

What would it take to let Suno use more than 3000 characters when making a song?
I tend to use a lot of characters when I describe how I want the song

[spare crown]

A lot of us have ex-boyfriends and ex-friends IRL that we want to write dis-tracks about, but don't want to restart drama. LOL. Musicians date musicians. Another consideration is that one of the people I now IRL stole a song I wrote, so I have everything hidden now. There are many good reasons to allow blocking of users and I see no downside. I would like them to see a "User not found" rather than a "You are Blocked" message also - to avoid the drama. In general i am in favor of having social networking abilities bc this community is amazing. If I could just block that one person, I would make songs public again.

[grim geyser]

I am an amateur singer, why can't I appeal to make my covers public? I use raw recordings of my voice to guide the melody, and I'm not covering known work to make covers, I'm covering my own work to improve on my own work.

anyway there's a term for that. I'm not familiar with it, but even though I'm using cover to make a better version of my song, it's not actually a cover.

the create menu needs an option to fill in accompaniment, and use the prompt to generate the music for the vocals.

I would like to submit covers for review so that I can go through an appeals process to make my "cover" public.

all you would need to see is a vocal stem that I clearly recorded myself, and that it isn't a known work.

[leaden marlin]

I would like to suggest adding a feature that allows users to download all MP3 or WAV files from a workspace in one click. This would greatly improve efficiency, especially for users managing multiple audio files.

Currently, downloading files individually can be time-consuming. A "Download All" option would streamline the process and enhance the user experience. If possible, allowing users to choose between MP3 and WAV formats would be even more convenient.

[crisp rock]

hi, where did the entry "Edit song details" (or whatever the name was) dissapear? It was there to change song title ! ??? How do I change now the title ?
They removed it! I can't believe they make always changes that make it worse, not better and make us loose even more time! And this without any notification!

[sleek vigil]

Just click on the song and then into the title panel and edit it

[hazy dock]

I need to know exactly which part of the lyrics are problematic and need changing to follow guidelines.

That will be nice Feature Update
And save me time

Right know dont know which part to change from the whole lyrics..

[opaque mortar]

I would like to suggest Esperanto (https://en.wikipedia.org/wiki/Esperanto) on #language-suggestions but can´t find that channel. Reasons for Esperanto: Conducive to more understanding and peace in the world. A very beautiful language to sing in (similarities with spanish and italian). Wikipedia offers an estimate of around 2 million speakers in the world. On reddit I found this comment to a post comment made 4 years ago: (https://www.reddit.com/r/Esperanto/comments/o78hfm/how_many_esperanto_speakers_exists_in_the_world/?rdt=41760 - see the comment by CodeWeaverCW

It is useful to know that 3.5mil+ people, and probably way more than that, have at least examined the language. If we go by learners and not speakers, we may see that 10 million figure that Zamenhof dreamed of, which would be a very big boon for Esperantujo and the internet's influence on it.

It's also worth looking at what well-known languages (possibly) have fewer speakers than Esperanto, to illustrate community size:
Basque: 660k
Breton: 250k
Irish Gaelic: 440k
Ossete: 550k
Welsh: 750k
Icelandic: 314k
Belarusian: 4mil (not necessarily fewer than Esperanto, but the gap is close)

[ionic kite]

I love the interface that riffusion have, hopefully suno could implement something similar for projects, and the ease for adding / removing.... also it would be highly lovable to have an option to "archive" songs or mockups you don't want to see in the sog library, but still searchable or a way to look into the archived mockups... maybe even added to each project

[granite rose]

And William Shatner made a (very strange) movie all in Esperanto.

[crisp rock]

yeah, thanks. I had figured out. But this takes more time and before! Worse and worse and worse all their "ideas"

[crisp rock]

hey lazy Suno, how would it be with making finally some of the many suggestions here ?
Or implementing also some basic functions for your poor display filter like "show only unlistened songs" or "show only whole songs which has not been edited at all" ?

[stone cloak]

Hey all. Doing songs by parts, trying to find the best parts working together and then assembling outside with a DAW. Then, I would like to use the remaster, but it is limited to 2min. If I remaster the same track by pieces of 2 min, got remastyer unconsistancies for each pieces: any way to solve that issue?

[knotty hearth]

I also want two different voices (father and son or mother and daughter)

[knotty hearth]

It should be priority #1 😄

[knotty hearth]

Yes, that's the format but it doesn't work all the time. Generate again until you get it.

[knotty hearth]

It would be nice if we know when and what features are coming.

[crisp rock]

to the Suno devs: did you ever test the conseqences of removing Edit song detail from the menu ? I don't think so!
If I chose liked songs in the Create window filter and go to a page and edit the name of a song with your new silly function, if I jump back after editing the title, it brings me to the first page and with deselected "liked." So if I want to get back to that song that I changed the title, I have to remember the page first, then press again the like button and then look again for that song. This takes forever ! Very stupid change for us, poor users !

[tepid trench]

Do you mean this?

[robust jewel]

According to the update posted today, "The end of shimmer marks the next phase for v4, which has been in beta since its November launch. While v4 is no longer in Beta, there’s still work to do. " If it isn't in Beta anymore, what is it in? Gamma? Because it sure isn't publicly available...

[snow remnant]

Hey Suno, why do you don't follow choices???

[woven fractal]

[[ Changing "Remove" bubble to something other than "crop". ]]

If I've noticed correctly, when we Crop something the resulting gen labels it a "Crop" in the Library.

When we Remove something, the resulting gen lables it a "Crop" also, instead of a "cut" or "cut out" or even a "remove" if you wish to nominalize it.

After removing something I go back and all my gens say "crop" and I'm confused which ones I removed something from, other than guessing it's the last gen I did.

[scarlet wigeon]

🇫🇷 👋 Hello everyone,
I must admit that I don't understand the new interface layout, and especially the one defining "Trending."
It may seem clearer at first glance, but when you use it, it's just annoying, and you move on.
The impression that remains is that everything is scattered and messy.

[hollow apex]

Ummm just realized I was posting in the wrong server 😅 sorry

[crisp rock]

yes.. this was much better than what is now ! Awful! I wonder who implements and approves such changes to worse...

[tulip fox]

Nest extends under the full song

[tepid trench]

Do you mean the option has gone completely? I took that screenshot last night after seeing your previous message and checked it on my paid and my free account to make sure it is the same on both. I think the position and name of the menu option changed. If the option is completely missing for you, that sounds like something in the cache isn't up-to-date. I'd see if clearing browser cache (possibly also cookies) helps with that - or test with a different browser.

[crisp rock]

It was not here anymore last nite! That is 100%. I even made a refreh. I will take a look now

Ahh, they put it somewhere else! Now it is in the first menu entry. I did not see it last nite !Thanks

But this kind of changes with NO NOTIFICATION despite DISCORD and "What's new" in the app ????

[proud parcel]

When creating a playlist it would be nice to have the ability to change or move around the position of a song that in the list. THis would be could for desktop. IOS and android

[haughty dome]

LIBRARY — When I navigate: Library, Playlists, {Select a Playlist} I'd like to see the Library header. Library questionnaire #📝┃user-research-feedback message .

[slow walrus]

Why you did remove the artist name from the song embeddings?

and when we speak about song embedding it should contain at least first 2 words of genre prompt. because that would make it easier to listen through some songs in showcase. Because with so many songs posted here it would help to know at least the genre without having to open the link. with this you can only guess by the cover image what it might be and you skip a lot of songs for your preferred genre this way

[marsh barn]

is there a solution to suno giving me the old "An error occurred while creating the persona."?

[cunning blaze]

Is there a way to choose a BPM tempo when generating songs? WIth vocals aswell?

Idk if asking the prompt to do it will work

[nova basin]

I would like to have a reversible sort order in the library and in the profiles that can be reversed. For example: Newest first and oldest first (also in the app)

[next ridge]

Look at all that room!! I can read more then one line of lyrics at a time

[desert whale]

Hi guys. Just wanted to make a request to ask you please don’t ever sunset V3. Still prefer it most of the time, especially for the genres I like working with 🙂

[next ridge]

Prompt: "My input dialogs seem to be made for ants, please fix this css." ❤️

[next ridge]

Also copy button here:

[stiff oar]

Hi, SUno is great, I have some tips for better organized UI (or maybe it's possible even now):

1. My library page: filter by workspaces
2. Rename song title

[delicate cedar]

Sorry this does not work at this time. Any written guide or HOWTO that claims otherwise was likely hallucinated by a GPT. Those who have made claims before have been unable to consistently back up their results with more results.

[lone breach]

suggestion-

Ello, this is my first suggestion. I remastered some awesome songs V3.5 > V4.
I noticed it sounding better but many things are still missing like WHERE IS THE DROP???
You know what I mean right?
To be honest, V4 remasters are good but not for drops and energetic parts, those energetic parts with mostly the drops in it are transformed into silent beats and I don't think many people like silent beats as the head-part of the song. Can't sent links in here so I'll so it in General.

[cloud saffron]

Feature request: Replace Vocal
Use the same exact instrumental track, but generate new vocals.

[haughty dome]

If you have Library suggestions, you are requested to take the survey ... #📝┃user-research-feedback message Rename song title in the header.

[haughty dome]

You can try something like this. Results are not consistent. Before lyrics put: [Tempo: 140 BPM (Upbeat, Quirky)] For Vocals try: [Vocals: man (deep, rough)]

[tulip sedge]

I'd like to see a way to convert a "crop" into a "full song". Or does anyone know how that can currently be accomplished?

[left igloo]

Id like to pay to own song that were made during a free plan

[wooden yacht]

When creating a folder, do you want to allow folders inside the folder to be placed inside the folder? The borders are too hard.

[ionic kite]

when using the edit, it's very hard to see exactly where you set the start and end of the replacement area, it should be thinnder and maybe even have some transparency to it so you can see the waveforms behind it, you can zoom, but still its hard to see it

and also there should be some kind of visualization to how long after the replacement which will become affected by the replacement

oh i guess this should been written in layout

[tulip fox]

It would be nice to change the number of tracks listed, so rather than sifting through pages of songs to organise them into workspaces, we could have a compact list of 100+ songs that are quickly selectable and then moved.

A compact list of playlists and workspaces, just to help manage them better

[crisp rock]

which genres is V3 good at?

[lean whale]

A function that convert a workspace to a playlist would be very useful.

I’m collaborating on a song and have a couple dozen versions that I’d like to share, but my only option is to manually add all the songs in a workspace to a playlist to share.

The other way the feature could work would be able to share a workspace.

[lean whale]

Also, as I’m creating songs in a workspace, I’d love song names to include an incremental number. I need my collaborators to be able to say, “I like version 011 and 023” of a specific song.

[west swan]

Hey! Just a suggestion—it would be awesome if there were an option to generate (and download) a subtitle file (SRT) from the songs created.

[west swan]

Unfortunately, it still looks just as small to me as always 😦

[abstract sun]

NEED HELP! When I create instrumental music and select the "Instrumental" option, every generated track includes promotional vocals like: "Please go to Suno.com slash ey-oh-essor Suno.com slash android to make music FOR FREE with the real Suno."
I’ve made 60 tracks and nearly all have this issue. I tried multiple fixes but it persists. This ruins my workflow. How do I disable these vocals?

[lean whale]

It would be a great feature to add where users could use credits to convert a song to obtain ownership. Does the remaster functionality count? Like does a user own the rights to a remastered version of a song they created while free?

[left igloo]

I dont know that what id like to know

[desert whale]

Older styles

[glass echo]

Has anyone seen when they will fix the Style of Music text box. The way it is now really makes difficult to type inside.

[midnight whale]

Yes, it's been 5 months since V4 was released and they haven't talked about V4.5.

[torn pawn]

devs. PLEASE. I'm losing too many credits generating covers that have been bugging the past few days

the covers end randomly before the song ends

its happening almost every time i generate a cover

sometimes it ends 20 secs before but i just had a cover end at the 1 minute mark

the og song:

[midnight whale]

V4 is out of beta OMG!!!

[willow violet]

Remove the water-mark when downloading videos from Suno.

[robust jewel]

Because it hasn't been five months since V4 released. It's been five months since the beta for it started. It still hasn't released, despite the update saying it's "out of beta." It isn't available to free users yet. It would be very strange to move on to V4.5 or V5 when V4 still isn't out.

[tepid trench]

It is out of beta according to this update: #🚀┃updates message
v4 no longer has "(beta)" in its description on the model selector.

[robust jewel]

Ok, so what does that mean? Because, previously, being in alpha or beta meant that it was only available to subscribers and leaving it meant it was publicly available. Saying "it's out of beta!" and changing nothing doesn't really mean much...

The statement when V4 came out was "v4 is now available in beta for Pro and Premier members!" Strange that it's "out of beta" and yet not.

[tepid trench]

During beta there would've been significant model changes. Historically, outside of beta the models did not change (v2/v3/v3.5 should be the same as when they left "beta".)
Whether it is available to free users or not can be independent of what stage of stability the model is in.

[next ridge]

That is my custom CSS I was making suggestion.

I will post it here tomorrow or next day, almost done with left and right panels

Then you can just paste it into chrome or firefox extension and have the fixes

[prisma crow]

How to get I to "Feature Mode"? I would very much like to be able to create a duet song using 2 of my Personas.

[desert whale]

Please please please let us use our own vocals as personas. I’m sure there are ways to safeguard against people abusing the system to avoid copyright issues.

[haughty dome]

LIBRARY - There's a need to link a song in Library (or Playlist or Trash) back to it's relevant Workspace.

[prisma geode]

https://suno.com/song/e8dc9f6f-10e0-40c6-bd6f-9c34fb78158d

[tepid trench]

This channel is for suggesting improvements to Suno.
It's not a place to promote songs.

[dull pendant]

Seriously, there has to be something done, because I keep wasting points for wanting to generate "female" voice, while I keep getting "male" instead.

[river dune]

Maybe the voice identifies as female 😛

[next ridge]

100%, also need to just fix how we pick workspaces and playlists, TAKES FOREVER to find the right one and half the time the one you want doesn't even load into the list until like 3rd try

[left igloo]

Lmao

[raw kraken]

At the moment, in the Style you can add female vocals, in Exclude Styles add male vocals, and the Lyrics you can weight Suno to specify the gender by adding [Verse - Female] [Pre-Chorus - Female [Chorus - Female] [Bridge - Female].

[desert whale]

Hi team 🙂 firstly, love all the updates so congrats on all the shipping. I just wanted to say though it’s starting to get a little overwhelming, I would personally love an option to go back to legacy mode for extensions (I.e how things were for v3 - 3.5)

[solar briar]

Replace with Upload covers anyone?

[dawn moat]

Suggestion: Let creators give the melody a vocal range ie LOW: E4 HIGH: B4

The use case is that myself (and other artists) use Suno for inspiration to build a song. I'll go through 15-20 iterations to find a melody that fits my vocal range. Suno favors very high voices as a norm.

The workflow I do all the time:

1. Create lyrics and build the flow of the song
2. I use another AI to help me draft keywords (Since I can't say "Morgan Wallen" or "Zach Bryan" as keywords) (The AI breaks down the style and instruments of the artist)
3. Generate 15-20 songs until I find one that has the right sound AND the right vocal range (Usually within 10 generations i can get a sound i like)
4. Separate tracks
5. Re-record the topline with my own vocals
6. Rebuild the instruments in a DAW to get higher quality sound (or sometimes I pay professional mixers to recreate the Suno music)

Suggestion: Let creators use their own voice for the vocals

Music artists and producers would be able to more quickly identify the potential of a song if it was generated with a pre-determined voice.

1. User uploads their voice sample (or records it directly in the app)
2. User Voice is saved permanently
3. Pick your voice when creating a song

[crisp umbra]

My improvement for Suno would be an option to tell the AI to change volume of the created songs. Recently i get, with 3.5 and 4, songs that are on the maximum volume and bass is unbearable to listen to.

[faint zephyr]

Yes, if you mean using the End of a track, like when suno begins another version near the end of the track and you really like it. Just use the Crop window to save the end and then use Extend. The cropped part becomes the Intro and then you can add lyrics into the extended part and keep building it up from there.

[rancid flicker]

#💡┃suggestions-web a button that allows use to exclude every single style only way for a style to be apparent in the music is to mention that genre.

[tepid trout]

Allow tracks to be max 6min to 6:30min and fix extending track extending tracks now messes up tracks really bad make these two changes and I would get a premium account.

[unkempt berry]

Out of Beta? Shimmer is back, quality loss and strange metal sounds and not following instructions. It was better a month ago. Also remasters after making a cover.

[wintry perch]

Can we please, I am begging you get a button that switches voices from female to male and male to female. I hate when I have done everything but a raindance, to make it a certain gender voice and its the wrong one. Even when it is absolutely perfect everything I want in vocals, cadence, timing, pitch, and the song style itself and then its the wrong dang voice. Ugh. It ruins things every time.

[oblique basalt]

Can we please have the option to keep the track and replace vocals. Many songs come out great but the vocals, more often than not, ruin it.

[radiant shell]

It would be awesome if we had the ability to extend from a song that was faded out. Assumed we could, but would never generate an extend from a fade out.

[hearty shoal]

I’ve had that message, it simply means that the browser didn’t refresh, the song will still be there

[hearty shoal]

Extend before the fade out.

I would like to be able to pick between several EQs before hitting Create to fix issues related to overpowered and/or unbalanced instru. Pop, Jazz, Rock, Hip-Hop, Electro, Lo-Fi

I would also like to see a lot of documentation on v4 and how it works to stop wasting credits exploring things that should be explained.

I would like to see an icon wi-fi like to show server status as it pertains to server overload so I can stop wasting credits if I know the server is so busy all I will get is trash gens.

I would like to see the addition of .gif support so we can have animated profile pictures.

The algorithm that features songs on the front page is broken, I’d like to see human curated lists, some of the songs that are at the top aren’t necessarily good, I’ve seen extremely good songs not getting enough attention because they’re just not being spammed everywhere.

We should have a dedicated spot on our profiles to link a Spotify, YT etc. as you know a lot of us are distributing our stuff elsewhere. In the same vein, some of us sell merch.

[radiant shell]

Oh nice, good call. Thanks!

[hearty shoal]

I’d like to see a Translation tool, one that can provide an automated translation in a target language and that the artist can edit in specific target languages that would be found between Lyrics and Comments

I’d like to be able to upload an image before hitting Create like it’s possible on Riffusion. That way it saves me the headache of Remastering a song so that the actual picture shows up when I share it, because Generate Video never works.

When I extend a song and use the Instrumental slider after having had some lyrics, and the lyrics box is empty, I’m expecting to get just instrumental and not random lyrics.

I would like to see more instruments, as it stands there is also a huge lack of FX, and triggering them if they do exist is a real pain. We need a list of them all, and a tutorial on how to trigger them. It should not be based on luck, it should work 100% of the time.

I would like to see a closed beta on a raffle based system to get access to unreleased models. That way a select few can provide you with feedback on what to improve on before a version is released like v4 beta and we’re all left to deal with something that doesn’t work properly.

I never gave any feedback but this is what I have to say. Been waiting since November to open my mouth and I hope it helps.

[hearty shoal]

A lot of us if not all of us are part of more than just the Suno Discord server, being able to link 1 server would be nice. Especially when you’re a creator in a niche sector looking for a place to create better content.

[slow walrus]

You can try cover function

[slow walrus]

The voice control leaves indeed much to be desired

[willow violet]

#💡┃suggestions-web message still hoping for this ^^

[lean whale]

Generation of unique stems. Instead of using an algorithm to split stems like it seems to function now, Suno would actually generate instramental versions that were stripped down along with a acappela version of the vocals. So like "Remaster", but it would "re-track" or "multi-track" parts as individual generations. I imagine a dialog that had checkboxes for which instruments would be grouped on tracks or totally separate. So if a song had a really great beat, that drums could just be generated using the re-track option with "drums" or "beat" selected.

[glass barn]

2 songs = 2 credits

[haughty dome]

I'd like to keep extending until I think the track is finished. I don't want it restricted to 6.30 mins.

[wintry perch]

isn't it four minutes?

[dense halo]

when are they gonna add the ability to leave out instruments. I'm a drummer and just wanna use AI to make rock/metal tracks but add my drums to it myself. Is it all baked in together somehow and that's why it hasn't been an option yet? or are they generated as separate stems?

[slow walrus]

currently the stems are only filtered from the final mix unfortunately

thats why they sound so bad like you know from audio editors. and thats why you still can hear the voice in the instrumental stem

[worn wigeon]

Feedback for training data It would be nice to be able to generate nature sounds. Right now, everything that's created is music, which is nice and obviously the initial goal. Expanding to nature sounds as a kind of ambience music would be a nice editition for meditations. Thank you!

[crisp rock]

any other music AI you can tell the automatic lyrics generator module to make also an intro or a solo. Suno's AI seem not to accept anything else except some info about the song content.
Not in the lyrics window, nor in the styles window.

**/////--/*-
Most changes from Suno are not improvements, just the opposite.

1. A very stupid idea to move the "Song detail" in that "Song editor" button with double function (Menu + Edit window opener).
   Many time I opened that audio editor when only wanted to change the song title.
   Before was better.
   Why does edit title has to be in the same menu as editing audio? It makes no sense.

2. moving the Song Title to the top of the window takes a long long way with the mouse from up there to down there where the Create button is.
   When creating many songs you do that a lot of time. Cause who wants all songs with the same name ?
   Besides that, you need one click more as before to make the string in the song title active.

3. Where is the automatic song title generation gone ? That was a partially good thing (perfect would be only with slight different names for any generation).

3 changes = 3 worse functions as before.

Bravo Suno! Working all the time against the users, the ergonomy, our time, our nerves.

[thick plaza]

I feel like a genre/sub-genre identifier tool would be of GREAT value. Something where you can plug in an artist or album (if the artists style has changed over the years) and Identify the genres/sub-genres

[winter hound]

Aren't there websites that have that info already? Discogs, for example.

[delicate cedar]

Rateyourmusic?

[astral sigil]

You can make nature sounds not sure full length but i make nature sounds intros and outros

[pulsar pagoda]

My Semi weekly request of letting us combine cover and persona

[slow walrus]

damnnn I really thought we can now change persona on an extend, because with the GUI update the persona field didnt disappear when choosing extend.
but when selecting a persona, the extend is removed, me was too early happy

[pulsar pagoda]

Yeah, using any of cover, extend or persona will remove the others. Hoping they update with that option soon.

[slow walrus]

it looked so hopeful, because before the gui update from a few minutes ago, choosing extend right of the bat did remove the persona field, but now it does stay there, and when selecting persona, the extend is removed instead. too bad

I need it so much, because I have 2 singers and its annoying when Lucy has always a new voice :/

[delicate cedar]

Meanwhile I'm sitting over here hoping that we can one day allow limited participation with our Personas instead of a binary choice of public or not.

[slow walrus]

Also cool to have would be:

• Language preference setting

Because currently it prioritizes english. So when I write german lyrics and the same word exists also in english, then suno sings it in english
The suggested setting would then prioritize in my case german and would sing it german then

[pulsar pagoda]

meanwhile i genrated a song in english and it gave some random part in what sounded like german,

[delicate cedar]

Now that is a good one to post in #💡┃suggestions-web for the Team to see. Exactly the way it is here. You explain the issue and give the reason it's an issue very clearly.

[rare obsidian]

✅ Warning logged for wojicamusic_60785. I couldn't DM them. || this channel is not a showcase.

[pulsar pagoda]

we need an [x] button for the big red box that says "moved to trash" to get it off my screen

[next ridge]

New interface is MUCH better, still needs some work imho, but compared to the other one GREAT

Still overscaled on 4k, but at least It's not working on a ship in a bottle anymore

[hollow steppe]

What about having feature parity on all devices / experience (desktop/mobile browsers, app…)? Crazy that it’s not consistent.

“Why is feature parity important?

Making sure your product or service offers a comparable set of features and functionalities across different platforms or versions is important for three main reasons:

It gives a cohesive user experience
It fosters trust and reliability in your brand
It improves accessibility.
Essentially, it’s about guaranteeing every user — no matter their choice of platform — gets the full benefits of what you offer. “

[empty spear]

Implement proper freeing of DOM objects on https://suno.com/feed
Once you scroll through about 20–30 pages, you can kiss your browser goodbye — it’ll hang itself.

[thick plaza]

While true, having it all in one to help with creative flow would be great.

[pallid tulip]

[SUGGESTION, FEEDBACK, All Plattforms] I often have such phenomenal results in the instrumental versions of some songs. But then the vocals ruin the whole song. [IDEA]: It would be great if there was a way to get the instrumental version STEM “CLEAN” afterwards, without audible lyric stems. --- Is it not possible to use the AI to determine which “recipe” was used for the song and then recreate the instrumental so that a nice-sounding version would be available that I could then edit further?

[crisp rock]

the mute AI music applications

All music AI apps and their developers seem to be mute. No sounds at all. They have forgot that sometimes an acoustic signal is more worth than 100 written messages. It would be useful to hear a "beep" when the queue is full for example. Otherwise you have to watch all the time on the screen .

[crisp rock]

The hour when the sub expires/is renewed is MISSING! Only the date is written. We need to know also the hour.

[unreal burrow]

i d like to finally see our saved favorite personas being able to sing along with our uploaded melody, right now we re forced to take random voice which AI randomly picks, and sometimes the song turns out really good yet voice is not the kind one really wanted

[cunning sable]

Why can't we download videos of our own songs with free accounts now?

[wise urchin]

A feature to remove all songs that are not organized in playlists. Playlists would be then the 'marker' to always retain them.

[slow violet]

Hello, why is it on the editor, when I replace a part of a song, the generated clip is not playing. It shows 'Loading' while the time is running.

When I go back on the 'Create' page, it can be played, but only on the editor, its not playing.

[dawn shuttle]

The new design in the Creator area is nice and offers a lot more options, but I can no longer access my workspaces, and the title entry is also gone. It's a pain, because I have to change the title later or use the first line in the lyrics area, which then results in the title being incorporated into the song.

[tepid trench]

Title can be changed by clicking on the "Untitled Song" text at the top where there's a pencil icon.

Workspaces are accessed by the back button (the circled < symbol) at the top, to the left of the current workspace name.

[crisp rock]

It is SOOOOOO difficult to add a keyboard command to ALL of these buttons?

[autumn mason]

Please Please make it so that you can get a dry main vocal out of sumo. As easy as just doing an instrumental. Not the AI separation but the same for music. When you are done with a track or have uploaded a part of your own. Just have the option to render a really great lead vocal and maybe vocal stems. That would elevate Suno to the stars. For making demos. Since you can today play a melody with cords and click track and upload and make Suno interpret the melody in to vocal singing. The possibilities would be endless. And also som sliders to have more control. Since Suno AI must have all the data for the vocals that should be doable.

[wild pumice]

I just saw anouncement by Riffusion. They made it generate stems separately. Not making stems afterwards but actually generate stems as separate instrumental and vocal audio. That is huge!

[autumn mason]

Hmm, I would have settled for just vocals but that is really interesting.

[wild pumice]

Yeah, I would like to have option to choose vocal only, like we now have instrumental only option.

[hybrid crane]

@celest aurora Hey — while we know you're skilled and probably had good intentions, sharing a tool that lets users download songs they didn’t create is a serious violation of Suno’s Terms of Service. It puts you and others at risk of account action. Please remove the tool from google and stop promoting it on this server.

[craggy vessel]

/

[worldly oxide]

Please improve the Burmese/Myanmar language for version 4. Pretty please. I'm a paying customer.

[slow walrus]

the paid plan doesnt guarantee exotic language support

so that "I'm a paying customer" is a bit wild here xD

[worldly oxide]

The language is not that exotic. I'm speaking on behalf of 55 million people who speak this language. Just wanted some representation. I love Suno. It is revolutionary. To be able to personalize music with a few clicks is just amazing. Forget about the "I'm paying customer" Trying to get attention, since you commented, that worked. Yeah, sadly, this population needs hope. Thanks bro.

[cyan tinsel]

Please remove the copywrite detection and vulgar language system as a big part of using Suno is recreating existing songs in different genres. If the system detects a copywrite let the user create the song just dont allow them to make it public. That way if they download it and repost it they are the one going against copywrite and getting striked.

[tepid trench]

It is against the Suno Terms of Service to even create a song by inputting lyrics or audio that you do not own the rights to, regardless of what you do with it afterward. Suno is intended for making original music. So they are unlikely to remove the copyright detection, and if they do it won't be any time soon.

[sweet nebula]

Will SUNO AI get a tool to separate vocals and instruments?

[quiet latch]

Would be great if the credits can be saved exemple trial gets 50 daily credits i decided use 20 credits today tomorow i will get outhers 50 Credits so in total since didnt used the outhers 30 total is 80 ...Becouse some times we can use the total 50 credits but we not fully satisfy with the result.

[brisk osprey]

https://suno.com/song/eb65f76a-1b90-47ce-bc55-a52097b81f88?sh=bvkYxlwhpdjUsI1l

[woven fractal]

@brisk osprey Looks like a nice song. Maybe personal as well.

It seems you've posted in the "suggestions- browser" channel which would likely get no listens here unless you reference what the post was for, or to do, with browser changes.

[sullen hawk]

Collaborations possibilities

Sometimes i want to collab with an other user, but we cant do it yet. If we do, we have 2 different songs.

There should be a ft. Option, that we can both upload the same song.

[heavy edge]

make personas attachable to songs as a seperate option

[cerulean girder]

help Login Error"You do not have permission to access this page"

[torn geyser]

Is there a way to control line wrapping for the lyrics when downloading the song as a video?

[slow walrus]

a small suggestion: Option for auto generate a song title on a instrumental song, ideally fitting to its music

[pulsar estuary]

search function for workspaces when moving songs to workspaces

[pliant tendon]

Maybe v5 can introduce some new [Syntax] too.

[Verse] [Interlude] [Pre-Chorus] [Chorus] [Bridge] [Final Chorus] [Outro] [End] [Intro] [Instrumental]

These are the ones I'm aware of. There might be more but what if they introduced new ones? Ones that modify the vocals? One that demands a very specific note to be sung at a specific spot? i.e. And then the [Note: G4]stars caress my [Note: F4]heart
Would be lovely tbh.

[steep crag]

PLEASE make option for users to BLOCK someone here. I am having problems with another user. Email sent to Support, it seems they are very busy from reading other threads. One guy was waiting 3months to get help. I cannot wait for 3 months to stop this creep. It's a serious issue, I am female, he is HARASSING me, with repeated "LIKES" of my songs which he took and turned into crappy things. @dire frost you can check with SUPPORT please.

[left crystal]

You think they will add duets in the future? Maybe add a block for descriptions on how the music goes, because SUNO is like a teenager in its rebellion phase that still ignores the simplest things.

[midnight whale]

https://www.reddit.com/r/SunoAI/s/qYjrklJEzy

We need Suno's team to read that

[woven fractal]

It may be simple to ask a dog to speak, and she'll try, just remember that when she barks, that's all you get. So for now be one with the dog. After some evolutions though, you may get a speaking dog. Be hopeful. 🧘‍♂️

[jagged chasm]

[Type: suggestion]
[Concerning: reproducibility]

Hello Suno team! Can you add a "Random Seed" option when you create a song?

Why: help to create reproducible styles/songs.

[ember bolt]

Any chance we might get midi

Idk how the generation works so unsure if that’s in the cards

[austere fog]

bring back the old ui, new one take up way to much ui space on browser

[pale glen]

When trying to skip ahead in a song feels like playing modern minesweeper.

[frank tartan]

It would be great to have a way to clean up older or imperfect tracks by removing unwanted audio artifacts like background noise, harsh spikes, or overly loud bass—without altering the original performance or turning it into a remix or cover. This would make it possible to preserve and polish raw recordings that would otherwise be unusable, while keeping the integrity of the original song intact.

[gleaming olive]

In "Explore" Browser version I need a Volume Control Slider Please !!

[frank tartan]

With so many musical styles out there, it’s not always easy to know which direction to take—especially when you're not familiar with all the genre nuances. It would be incredibly helpful if AI could analyze the lyrics and suggest suitable styles or arrangement ideas, explaining the reasoning behind each suggestion. This could serve as both inspiration and guidance, helping creators find the right musical identity for their song.

[brisk robin]

Why is there no option, during replacement, to ONLY replace the lyrics? That would be incredibly helpful. I just want it to say the right words and pronounce the words correctly, leaving everything else exactly as it is. Please add this feature.

[glass loom]

It would be awesome if you gave an option to create and download a Chord chart of the songs so musicians could play them. I Chart the ones I pick but would make it a ton faster if it was something the program did?

[slow walrus]

there are a lot more working

Of course also [female vocals] / [male vocals]

[slow walrus]

duets happen its just super hard to trigger them, because its already not guaranteed to get the right gender, its always dice roll if the tag is used

[slow walrus]

and if the replace section happens to be longer, please DONT overwrite the overlapping part - just do the obvious: move the original part to the right, so that the replacement section FITS, instead of overwriting original data due to this

or would also be nice if the individual timeline elements could be moved around by us too. But for now we see the replacement even at the timeline being overlapped if longer than original and we can nothing do about it

[woven fractal]

Add to that, it doesn't appear you can scrub the audio either. Just indicate where you wish to hear, but no execution.

[indigo ginkgo]

Is there any chance, like any chance at all, that at some point we can at least publish audio upload songs on our profile, they don't need to go on trending or anything like that. But I have been playing with audio uploads and the stuff you can make with it is insane, I love singing and playing in samples so I can experiment and if I can't put it on my profile, no one will ever see or hear it, which is such a shame. You can even permaban anyone who gets reported for using copyrighted samples, I just want to be able to work with my own...

[steel knoll]

Can we get API's so we can integrate Suno into our apps? ❤️

[dry summit]

Update v3.5 and v4 Song timeings to 8min for Longer songs so i can finish my Shadows and Shadderd Album...

[worthy yacht]

This should go for all versions, really:
We need a means to actually denote chord lines as sometimes chordsheet imports implode without this feature.

[woven fractal]

Update: You've been answered in Technical channel. #📚┃suno-school message

[upper isle]

This suggestion will probably fall on deaf ears, but PLEASE add another payment method. Link sucks; it doesn't accept my debit card or most of my credit cards. I have to use a Capital One 360 account just to be able to pay you guys, and they take forever to post my deposits. If you could add PayPal, or even something proprietary instead of relying on Link, that would be awesome. I want to give you money, help me be able to do that lol

[heavy edge]

make it so suno can differentiate between male and female vocals in EVERY song not just some x.x spent so many credits trying to get it right

[median wave]

i have many occurrences of replacement clips being exactly 2:04 long. All of these were unusable and cost credits. Hope this is fixed soon.

[rare obsidian]

✅ Warning logged for starland_15113. I couldn't DM them. || Please don't needlessly ping members of the team (as you have just done by replying to one of them.) You replied to a completely unrelated message that was sent a year ago.

[trail acorn]

telling me what was moderated in my prompt.

[bright patrol]

I have a suggestion , is it possible to update our Suno account using Paysafecards? Will it be possible in the future? 🙂

[vast dawn]

Instead of using "My Song Title" for songs with an unspecified , use [username][song number]. That way all songs will have a unique name.

[primal topaz]

Add the ability to reorder songs in Playlists.

[supple magnet]

Oi

Songs in multiple languages with the same instrumental – It would be amazing to have an option to create versions of the same song in different languages, keeping the same vocals, instrumental, and rhythm. Basically, it would be a translation of the song without changing its essence.

Creating instrumentals from different sound forms – Suno could generate instrumentals from a written score, an image of sheet music, or even sounds made with the mouth, mimicking instruments. Additionally, any kind of beat or noise could be transformed into an instrumental, expanding creative possibilities.

Creating instrumentals from different sound forms – Suno could generate instrumentals from a written score, an image of sheet music, or even sounds made with the mouth, mimicking instruments. Additionally, any kind of beat or noise could be transformed into an instrumental, expanding creative possibilities.

[obsidian temple]

Click the "Get Stems" button..

[obsidian temple]

Been suggesting this for ages now. having 100 pages of the same title is maddening and so counterproductive.

[weak sphinx]

Suno Music Bashing

What is Music Bashing?:
In music bashing, also known as a "pick one" game, two songs are put head-to-head and you have to choose which one you prefer. It's a fun way to compare music tastes and spark friendly debates.

Reasoning: there is too many small creators who need recognition so this will help more learn about small creators also theres a algorithm so the more rap songs you vote for the more rap songs in your bashet algorithm/suno algorithm not only boosting small creators but helping them be seen by you a bit more

Personally that is a win-win

[plain kiln]

I have the idea that Suno gets a nice navigation, where you don't have to click awkwardly.
I work with several tabs. To get a new tab, I sometimes need to do SIX actions with the mouse, it's annoying.

1. you are inside the creation section and want to get to a comment section from your profile. (right click home, open link in new tab, click the tab, click picture, click profile, click the song) =6 clicks
2. you are on your profile and want to open 3 tabs of personas ((right click home, open link in new tab)x2, click personas, click the 1st persona, (click tab , click picture, click profile, click personas, click the persona)x2) = 14 clicks)

So, while 2) isn't really needed, it demonstrates how bad the navigation is, unless you have a better idea how to navigate.
-> But imagine you are on a persona page, and you want to see another one...
How would you do it? Try it. You will be confused.

[ionic kite]

please make a nicer and a more streamlined UI, workspace is clogged, too many sections, partial songs, scrap songs makes it hard to know what you work on 😍😍😍

[pliant canyon]

It's great to have personas now in cover mode, now we need a cover strenght slider where it keeps the melody and stucture from the original, so you can recover a song with a different singer till - complete something different .. If I cover a song now it's like I just copy the songtext into a new song template and choose persona.

[marsh lichen]

Yes, please give us a possibility to add Personas to Covers.

[pliant canyon]

You can use Persona now in cover mode, but it feels like it just copy the songtext to a new song template and doesn't use the melody of the original song

[marsh lichen]

No, I can't. When I choose Cover and then choose a Persona, the Cover button is removed.

[pliant canyon]

a i see... you're right, the cover disappears... it's a bit confusing because in advanced option you can add persona but cover vanishes.. sorry my fault

[still abyss]

Good afternoon.
You may be interested in implementing the following functionality:

1. If a song is commercial, it should be visually marked with a "golden" circle
2. On a paid subscription, introduce the possibility of purchasing commercial use for your old songs (created on a free basis) - by purchasing them for points (for example, for 50 or 100 points)

[plain kiln]

You can do it;
Not a very well implemented feature. But possible (though no one tells you):
"Just" drag and drop the songs.
If a song is further down the line, you should drop it maybe several times 😆 .
Then, after all is done, you can press F5 to see the updated list.
Very long lists are laggy. (over 80 I think?)

[haughty dome]

Have you ever tried creating a New WorkSpace? (top right)

[ionic kite]

yes ofcourse. but then you get overwhelmed by workspaces... I do have ADHD and messy in nature, but how does people here keep it neat ?

I wish there will be a way that any project you make in the future will have their own workspace and is easy to handle, workspace in suno disappears, then coming back, and pretty much makes it hard to have control. I tried to do it, but it just keeps messing up, i make new projects and im in the wrong workspace and yeah haha :D... it would probably help to have some kind of folder structure feature like normal folders like if you want to make a structure that you have everything from 2025 in one, then month and under there per project, also sorted alphabethically, or you rearrange as needed

just a feature suggestion

my folders on the computer are neat btw

but sunos workspace system is too flat

[ionic kite]

I know youre not suppsed to talk about other platforms here, but another i use has a proper way to arrange projects, you have projects, and easily can just check them into one another, simple, but it also has a semi flat folder structure, but its mote stable, since m a programmer I¨'d be happy to mockup a suggestion

[haughty dome]

That makes your suggestion a lot clearer. Like a Windows folder structure is how I'd like to see it, too. I'm labelling my workspaces as "YYMMDD - xxxxxxx " where the xxxxxx = a memory jogger, like the name of a Song I was working on, or a Style, or a Persona.

[ionic kite]

yeah. but even naming them will not help much, cause it jumps all over the place, and sometimes the workspace is gone alltogether. you have to refresh and try again many times

the vanishing stuff is really annoying, on mobile sometimes it says i have no songs and no playlsits 🤣have to terminate the app and restart it

[thorny dragon]

Bring Cantonese as an option

[heavy edge]

using multiple personas in one song like the option to just tag a persona in the song and it will shift to that persona

[trim lantern]

New music just a slow mode https://suno.com/song/88970382-abf4-4114-a1f2-8637d11579f3

[gloomy rivet]

Im not sure how other people use persona but I would like to use it as a character/AI artist and keep the voice consistent every time I use that persona. I would like to be able to add like traits like not just genre

[woven fractal]

I have a nice "conceptual" duet of two women singers that do pretty well on sounding similar on all songs written for them.

I originally created a duet in a song. Made a persona. Created songs with that persona, copped out the individual vocals, and made personas of those. So I have a persona for one or the other as well as have done some duets or good simulations of both voices in one song.

Your mileage may vary, but it's certainly possible for a one vocal artist singing new material.

So, as for a suggestion. It already exists as part of using personas.

[leaden marlin]

please enable personas based on audio clips I just want to make a persona based on my own voice. There's already copyright protection in place as in you can't upload a copywrited song (I even tested this with obscure artists and songs just to see and it detected them and wouldn't allow me) So i'm not exactly sure why the feature is disabled. I specifically bought the premium subscription to be able to use this feature and I was unplesently surprised when I found out I couldn't.

[slow walrus]

Update 3: Model testing/listening
OK, so there’s been lots of noise about noise in the v4 threads, and much of it is focused around heavy music. To show you some of the progress we’re making with new model testing, I'm including a clip I made earlier today! Let this be your warning: it’s pretty heavy. 🤘

That sounds nice indeed, especially for such a heavy song, sounds promising ❤️
My biggest wish is that the creativity and treble from v4 doesnt suffer from this though. I would dislike a return of a muffled 3.5 at metal music which is/was my biggest worry with the attempts to fix it. But this one has kept its treble, nice job 🙂

Promising.

I also wish so much:

• Voice Control - when I say female at a verse, then give me a female always, and not random
• fix often broken pronounciation/cut off words when using all capital letters words (I like to use them because they trigger quite reliable scream vocals). When writing in regular letters the pronounciation is amazing though. so weird that this makes a difference
• Let us be able to use a different persona on an extend. Its so frustrating when my singer couple has always one of them a new voice because I can only choose 1 persona
• For the current model especially it would absolutely help if we could extend at e.g. 2:30 minute, because even if the song is 5min already, extending at 2:30 you would get the quality from 2:30. So it would be SOO GOOD if we could say - append that 2:30 extend at the end of the song rather than at 2:30 again and removing the rest. it would be such a nice workaround to retain the quality. Also being able to add any extend from any song would be nice - so if we still cant change the persona at extend I could at least make a coverversion of my song and choose then the Lucy of my singer couple, being able to use extends of her then on my main song would help

• Replace should have an option to use the same seed, so that it would only change the lyrics and not also the melody and singing
• When Replace section is longer than the original - then move the original section to the right, so that the replacement section fits instead of overwriting the overlapped part ...
  when we even see the overlapping now on the timeline it looks even more wild that it gets not adjusted and not even the user cant move the timeline elements around so that we could fix it ourself
  being able to move the timeline elements around would of course be another possibility to have extensions everywhere we want and not always just at the end of the song applied

[woven fractal]

Suno is not all knowing, so it likely can't distinguish/discern between my voice and Elvis, though mine is better. So it has to limit ANY song with a vocal in order to stop those of us, nefarious enough to try, from uploading any artists vocals which are not our own, inevitably, therefore filtering our own vocals from being let through as well, in the process.

This will then not need to limit to copyright material but also trademarked voices such as Mickey Mouse, Scooby Doo, and Alvin and the Chipmunks.

[twin hinge]

Add a number icon to the 'thumbs up' icon so you can see how many likes you have on that song.

[upbeat mantle]

There Needs To be an improvement on the Remaster Function that gives the user a More controlled way to look on what gets improved in the song that way we can avoid those gens that either get a overpowerd vocal or instrumental part of the song which can get very annoying depending on how many gens your creating.

[sleek vigil]

It's interesting that you say that, I found the music example attached rather disappointing, I don't know what it sounded like before but it still is just shrill and hollow instead of rich and heavy

[waxen maple]

BULK DOWNLOAD -

I already mentioned this almost a year ago. In the hope that there will be a bulk download with a future update. Unfortunately, this has not yet happened.

It would make my workflow and certainly that of many SUNO users so much easier. The time savings would be immense, so that you could concentrate more on the creative creation of music than on downloading songs.

Please please please dear SUNO team, please implement a mass download in the next update.

Where I simply mark all the songs I want to download beforehand and then to the right of any marked track under the three dots just go to download and which format the download should have. And then all the selected songs are downloaded one after the other,.... that would be so fantastic

a 98 % happy customer, I would be 100 % happy if there was a mass download.

[vast dawn]

I can't name the song when I create it as the "Details" section seems to be subscription only. So I try to edit song details, and it hangs trying to load some sort of editor. Why? Are you using "the big dummy's guide on how to confuse users and drive them away" perchance? It's working.

[leaden marlin]

Subject: Suggestion to Improve Stems and Track Generation
Dear Suno Team,
I’m a user from Ukraine who loves your platform but struggles with some limitations, especially when working with Ukrainian lyrics. The new Stems feature is a step forward, but the quality of separation is still poor, and editing vocals (like fixing accents) often ruins the music. Also, generating two tracks at once feels wasteful since one is often unusable due to glitches or bad quality.
I’d like to suggest generating vocals and instrumentals as separate layers from the start. This would:
Allow users to edit vocals (e.g., fix Ukrainian accents) without affecting the music.

Make the second track useful (e.g., an alternative vocal style or melody) instead of throwing it away.

Improve stem quality since they’d be created independently, not extracted from a finished track.

This could make Suno much more flexible and enjoyable, especially for non-English users. I’d love to hear your thoughts or if something like this is planned!
Thanks for your amazing work!

[slow walrus]

My impression is it is an improvement to what we have currently. You think its not better at all?

[sleek vigil]

I don't have a good "before" example since I'm on V3.5, but the "after" sounds like plastic, hollow and unauthentic and wouldn't invite me to try it out again

[slow walrus]

to me it has much less distortion like usually in v4 and doesnt have to remove all the treble to reach that like 3.5

[sleek vigil]

The overall sound isn't pleasant, I rather deal with the robotic doubling on v3.5 because the overall sound is way more authentic instead of hollow, shrill and artificial

[slow walrus]

for my music v4 works much better than 3.5
3.5 never has treble and the creativitiy is much lower

But I like 3.5 at aggrotech
But rock & metal I absolutely prefer v4

and the voices are so much better too

[sleek vigil]

I disagree strongly, I've yet to hear a single song of V4 that I liked

[slow walrus]

really? weird

[sleek vigil]

The creativity of v3.5 is insane, V4 has been very disappointing on my end

[slow walrus]

nah v4 much more creative to me

[sleek vigil]

I will not engage into a match of taste so I accept your verdict

[slow walrus]

https://suno.com/song/ce985d7d-426f-4ee7-93dc-7e38a1d78a95

Alone this song its soo good ❤️
especially if you understand german 😄

[sleek vigil]

I'll listen to it in 5 minutes and yes I'm German so won't be an issue

[slow walrus]

ah nice then it will be interesting to hear your feedback 😄

and enjoy the amazing guitar solo at 3rd minute ❤️

[cunning blaze]

What if you could use RVC models with SUNO

[rigid badge]

I have a suggestion - a filter to sort songs so that finding the oldest stuff isn't a tap fest on a mouse 👍 especially when you can't remember how many pages you have to click through to find (and remember) what was done and when. (oldest - newest = perfect!!)

[latent canopy]

I think a cool suggestion would be to have a section where you can watch it make the music and you can nudge it or how ever you would call it and have it change to what you like instead of having to constantly recreating a song to get the one you like, I also think it would be cool if you could make uploaded audio into persona's!

[maiden bison]

when will V4 be free?

[midnight whale]

In iOS app is free

[maiden bison]

really? when will it be on the website cuz i used laptop

[lucid light]

What's next, an AI-generated comment to paste for when you're not sure what to say but need to say something?😂🤣

[wary plinth]

can we have time management in our song for specific instruments to come in and out?

e.g: like if I want a drum solo starting at 2:04 into my song, it should start there

[trim lantern]

@wanton compass . I'm sorry to contact you in this way, first of all, but i don't have too much time. The voice of Portuguese of Portugal needs to be better represented. We are an important country in the world, we have done a lot, and we are seeing strange things happening online and offline. I'm asking for an improvement in the native Portuguese voice adjusting grammar and tone (i know you have a lot of work, but as a local native citizen, i wanna keep Portuguese culture alive and its very difficult to generate things in Portuguese from Portugal with our grammar and tone i understand this is hard.). Thank you for your time and attention spent on this message, and sorry for the way of contact and sorry for my errors in English if they exist. Thank you all.

[woven fractal]

That or better.... an AI moderator that checks your work before you click Enter.

"Hey Dasvon...I've read you're comment and even I am confused. Do you really want to say this.....?"

"Dasvon... while your 1000 word dissertation of a response is nice, i think we can shorten it to just 20 words...."

[pliant tendon]

So, this is something that will be difficult to implement, but a separate field/lyrics syntax where you can define your language (+ perhaps the dialect).

[LANG: EN-US]
[LANG: BG]

This is in continuation of @trim lantern 's suggestion about improving Portuguese of Portugal. When a particular language has a dialect that's more popular (like Brazilian Portuguese), it's hard to force the model to sing in another one. The same applies for Hebrew, Bulgarian, Arabic... And I guess even French because Québécois singers do pronounce their words slightly differently.

[leaden marlin]

Subject: Suggestion to Improve Stems and Track Generation
Dear Suno Team,
I’m a user from Ukraine who loves your platform but struggles with some limitations, especially when working with Ukrainian lyrics. The new Stems feature is a step forward, but the quality of separation is still poor, and editing vocals (like fixing accents) often ruins the music. Also, generating two tracks at once feels wasteful since one is often unusable due to glitches or bad quality.
I’d like to suggest generating vocals and instrumentals as separate layers from the start. This would:
Allow users to edit vocals (e.g., fix Ukrainian accents) without affecting the music.

Make the second track useful (e.g., an alternative vocal style or melody) instead of throwing it away.

Improve stem quality since they’d be created independently, not extracted from a finished track.

This could make Suno much more flexible and enjoyable, especially for non-English users. I’d love to hear your thoughts or if something like this is planned!
Thanks for your amazing work.

[lunar lotus]

Suggestion: Please make it possible to crop and fade out in one step. I want to be able to select the end point for my fade-out instead of being locked to the end of the song, so that I can fine-tune the fade. Currently, I have to crop the song to set the endpoint where I want it, then do the fade-out, and there's a high likelihood that I'd have to do both of these steps multiple times to get it just right (each operation costing another 5 credits). This would make me want to just download the file and edit it myself in another tool.

[deep fiber]

Suggestion: If you can and have the time it would be great if suno had a PC feature where you could generate a music video for your songs like on the phones!

[marsh lichen]

Actually, cropping and fadeout don't cost credits. Or did something change? I don't use them quite often.

[lunar lotus]

It looked to me like it did. But even if it doesn't, having the functions combined would be a real quality of life improvement.

[queen inlet]

Can you give a option for language selection so the the AI can know which language song its singing. I am from south India and its very difficult to get proper Kannada song ,It feels like some foreigners is singing song, Some tongue twisting word are flat singed by AI .

[distant kettle]

Hello, I have a problem. I created music in Suno. I tried to download it as a video, not as an mp3. I tried but I get an error when downloading. Please fix it.

[twilit hedge]

This doesn't sound like a suggestion... post it in #🪲┃bugs-web

[slow walrus]

they dont count credit

[tidal ridge]

PLEASE, create a way to change a voice stem for some music generated in SUNO for another voice (also from SUNO) from another music, like a voice-only AI to AI cover. I think it will be good for improving the stability and coherence across many musics of the same persona. (Here, persona voices vary A LOT). The AI models to do this are already available. Please, just incorporate this in suno pipeline, and if you are afraid of copyright infringement, limit this function just to Suno native AI voices for kinds of music inside Suno System.

[slow walrus]

avoid gender tags, especially in lyrics. It will trigger new voice
Sometimes you have to put in style box female vocals if the persona is a female. that usually turns out still working. But if you add [female vocals] in lyrics as well, then you will get a new woman. The best result is if you can keep gender tag away everywhere

[wise pasture]

make a setting where you can do an instrumental with 1 instrument, like a Solo Play. (guitar. Drums. Shamisen. Ect.)

[frail leaf]

Won't generate because lyrics contain producer tag "you're dead" ? It's just a normal phrase man, I am sure it appears in many songs. Please take this off

[rigid summit]

Here is a suggestion: Can we get a new suno version, with like... more singers, better prompt adherence, an advanced mode with seperate boxes for singers/instruments/music style/music mood/etc that way it knows exactly what each category we want... rather than the million token giveaway... I'd rather keep my subscription and make better stuff...

[twilit hedge]

How about spell it a bit different ... like "you're ded" ?

[soft plover]

It seems like the moderator on the web site is a bit too strict. A prompt like "A song about Norman Bates" gets flagged as inappropriate (presumably because of the word "Bates") and so does "A song about Alfred Hitchcock's Psycho" (prsumably because of the word "Psycho") and so does "A song about Strangers on a Train exchanging murders" (presumably because of the word "murder"...but that's exactly what the story is about.) I understand the need for moderation, but sometimes songs need to be dark.

Now it's just getting silly "A song about Marion Crane" marked as inappropriate. What?

[feral aspen]

Not having access to CROP makes entering contests into a headache, you just have to hope and pray Suno ends the song at the right time. Free Users could really benefit from the CROP feature.

[spring girder]

Can someone help me? I want to create something inspired by the song 'Tacones Rojos' by Sebastián Yatra, but I'm having trouble pinpointing the style. I've tried Latin pop/retro pop, finger picking, live piano, funk soft melodies, dance, urban, etc. and nothing seems to fit. If someone could help me, I'd really appreciate it

[woven fractal]

Can free accounts download videos anyway? Are you free or paid subscription?

[woven fractal]

I doubt really it allows much in the way of using proper names of movie titles, songs or artists, anyway. So "a song about...xxxxx.." Or "in the style of...xxxxx. " won't pass, I'd think.

[feral aspen]

Try channels #general or #📚┃suno-school for prompting advice. (try repeating style words 2x or 3x to make stronger, if you get one lucky song that is closer to what you want, turn that song into a Persona, then use that Persona)

[robust jewel]

Are V3.5 songs really going to be winning, if only they can trim off the ending, over V4 songs? The contest isn't exactly on a level playing field anyway...

[cloud zephyr]

I've just came back from a break before v3.5 I've had some success with 3.5 vocals. I've got a really good intro to a song (i think so anyway) But I'm struggling to get a vocalist along with it

[haughty dome]

That is a thing already. For Pro.

[manic locust]

outline shadow for the play button in the thumbnail. (Some images camouflage which song is playing sometimes and would be less stressful)

[jade pendant]

is extend feature buggy right now? i've tried extending 4 different songs, and all of them dont follow the lyrics, and keep repeating and starting from places way before or after i specify

[fluid vessel]

I don't know if this is a thing already but I would like that the filters that I apply on my workspace folders didn't reset every time I go to the website. The "hide disliked" filter is applied by default and a few days ago someone went to my profile and disliked all my public songs, which is not a problem itself, each have their own tastes but toggling the filters on and off every time I go to the website gets boring after a while.
Or even better make it so that it only hides on the workspace folder if I dislike them myself if possible.

[safe valley]

I would like to be able to upload a song and find the correct version in my list of 50k+ songs, in those cases where ive downloaded the mp3 a few weeks ago, and find it and realise its perfect except there needs a little bit of fixing. This could be easily done without much overhead by just comparing hashes of the audio data, or by comparing the images perhaps from the meta of the mp3. Would save me a lot of headaches.

Simply have a feature where i can press "Find song from file" upload file, show me the song matching the exact file, so i can extend replace etc on the correct version without manually having to scroll through thousands of songs.

(You could even generate hash from the audio file locally on device, and send that hash to do a search on the server, which would prevent any potential copyright worries, and make it use no storage space at all)

[amber violet]

Feature suggestion: "Clean" especially with V4 noise, just have it go through a cleaning that removes the noise without changing the melody, music, beat, etc.

[random siren]

"ima steal yo chain then dump you in the rain, leave you where i left you, without a chain, you a lil boy" rap style

"ima steal yo chain then dump you in the rain, leave you where i left you, without a chain, you a lil boy" rap style

[simple saddle]

I would like to have the ability to move multiple songs at once to a new workspace or add to a group.

[arctic python]

Also to delete to the trash can, I'd like to select a bunch at a time rather than one by one

[modest mirage]

Feature request: STEM separation is terrible. When generating a track, I would like to have each instrument rendered to its own track instead of suffering through a horrible STEM separation. No STEM separation is that good, but SUNO's is really terrible. Having actual instrument tracks would make a lot of our lives much easier when trying to work on a track in a DAW.

[limpid otter]

plz make PC text windows open up into text editor like on mobile its really hard to see the whole song when all you can see is 5 lines of it all at once.

[prisma aurora]

I know its been brought up before but we need to be able to use our own voice clone as a model. I would pay 5x easy for that feature

[pulsar bone]

I had success with "yorded"