💡┃suggestions-web

[wraith osprey]

please enable "Crop" on the mobile browser. sometimes I want to complete a song and I don't have access to a PC but can't finish it until I'm on desktop because neither the app or the mobile browser supports it. I'd imagine it's just a matter of whipping up a mobile interface for it, all the hard work is done already.

[weak heron]

As a note, you can use desktop mode (desktop website) in a mobile browser and use crop and replace, but it's a bit tricky. Don't zoom in and most things will work.

[gleaming warren]

Hi firstly thanks for creating such an awesome program. I am a singer songwriter and recently was diagnosed with terminal brain cancer but it started in my throat and the surgery damaged the right side of my vocal cords. The app has allowed me to continue to creates music which has been my greatest joy. However I have a question. Is there a way that I could use stems to seperate my vocal from recordings I have of my voice before surgery and use this to teach the AI how to sing with my voice as it used to be. I’m currently writing a lot of songs to leave for my family my mum and my kids etc and it would be great if I could do that in my voice as they know it ?

[leaden lynx]

Can we get a empty all or delete all ability in the trashcan please or a ability to select more then one at a time to delete bad copies

[buoyant elk]

Man someone has to fix the volume control
Well mostly at the mid part of each song and The end part

by the way i got another thing, When i Extend songs for my last tiny part of the end song the Ai start's to Blast the volume up and the Energy just gets way to hardcore

i Recommend Fixing this in future
i know how to get around it, but wish it wasn't there,

[covert cobalt]

I would love love love if the custom lyrics editor had an option to go overlay "fullscreen" similar to the new "Full song" feature pictured here, easier on smaller screens to see. (people with windows tablets and smaller laptops)

[prime carbon]

Can you please do something so we can know exactly what we need to remove in the text , everytime i spend 30 minutes to search what is the problem , i loose a lot of time cause of this , please do something .

[royal hull]

I hate to criticize every time I post here, but the new 'write lyrics by line' feature leaves me puzzled. Was there anyone who actually asked for it? What value does it offer over the free version of ChatGPT? Wouldn't it be better to focus on features you already announced and labeled as 'beta'? There are plenty of them.

[haughty dome]

CROP - change to Green for Keep.

[autumn ferry]

Your text maybe contain a violates word, or your style contain existing artis, or the whole song feels explicit.

My experience

SUGGEST: I need, delete all button in trash menu, so no need to delete one by one.

[tidal ridge]

I would like to suggest a way to upload a sample of any music and the bot listen and offer me the prompt for something like that, not copy, but the mix of ideas to generate something like. In midjourney we use a /describe code and it looks any photo and suggest a prompt to describe. The music drescibe may be similar.

[pulsar estuary]

pretty sure u can alrdy make personas with uploaded samples of a voice. i think the problem is your songs are going to be in Version 3.5 (which has very low quality vocals imo). trying to remaster or change to a different version will change the voice. you're prolly better off letting your family be more aware of AI tech so that they can do something w/ your recordings in the future

[thin ermine]

Unfortunately Suno's official stance is that personas can't be made from uploads.

[pulsar estuary]

oh really?

:/

[gloomy steppe]

https://youtu.be/0sizyI30j8Q

maybe this mans opinion helps

[pulsar estuary]

thought it was Ryan Reynolds for a second

[haughty dome]

It could get very busy! 😉

[wraith osprey]

thanks but I tried that before posting and it didn't work. If I had a mouse connected it wouldn't be an issue, but there's no equivalent to click and drag in the mobile interface. Tap and Drag doesn't seem to work

[weak heron]

For me it works without a mouse with Google Chrome and Ecosia browser when you mark it as a desktop website.

[wraith osprey]

what's the trick? I have it in crop mode, i can see the waveform but when I tap with my finger it doesn't give me the selection interface.

[weak heron]

You can put the starting and ending time into the field below the waves. Moving the selection doesn't work for me either.

[wraith osprey]

oh I didn't consider that. I could see that working, I'll give it a shot. thanks (it worked)

[weak heron]

You're welcome.

[sacred sluice]

I thought i could use persona's in the cover option in one of the former versions but it seems disappeared after introducing the workspaces? i hope it will come back

[eager crown]

can I suggest that cropping from middle(removing middle)?

[gleaming warren]

I am using version 4

[pulsar estuary]

all uploads are currently recognized as v3.5

so to continue working w/ the upload, it'll be in 3.5 unless you remaster (version 4) the upload

don't know about u but i personally dislike 99% of how the remaster turns out

unless it's only instrumentals

[tepid trench]

Pretty sure you get a choice of which version to use when working with an upload, without having to remaster first.

(Initially, subscribers could only use v4, but they changed it so you could use v3.5 if you wanted to.)

[pulsar estuary]

3.5

[tepid trench]

is the default.

[pulsar estuary]

this sounds like v3.5 even tho it says v4

so yea it's v3.5 quality

they all sound bad

[native meteor]

When moving songs to trash! instead of manually deleting one by one possible to have a Select all button? and then a warning like "Slowdown You are about to delete some songs continue anyway/ Take me back!"?

[karmic sigil]

There should be a way for Remaster to detect what Persona was used originally on a base song before any other operations (Crop, Extend, Cover, etc.) were used and then have a toggle to reapply that Persona to the entire song if wanted during the remaster process.

[rancid zephyr]

I have a couple of suggestions that might have already been mentioned, but I'm going to share them anyway.

1. Allow 3 or 4 singers in one song.

2. Include effects like gunshots, thunder, or wind.

3. Enable editing the music without changing the lyrics.

4. Offer a choice between female or male voice options.

5. Provide a dropdown list of music styles and moods.

6. It would be nice if we could communicate with the AI to request changes.

[haughty dome]

2. Have you tried to use these sound effects? https://suno-ai.notion.site/Using-Metatags-34944efe09ec41d693e314f13af44695

4. is a bit hit and miss. Although [male] and [female] usually works. https://suno.com/song/18d3a2d4-ca04-4d1e-97e5-270311dc8666

5. There are so many Suno would probably say "no". There's always "Search" to find a genre. or the "Explore" options.

[queen wadi]

please remove SUNO from the videos for paid subscribers. Thanks!

[jade pendant]

maybe there needs to be an option to remove the watermark from paid users

[shut dagger]

lemme do covers with personas please <3

[prime carbon]

where you find dat word ? i dont see any list ?

[fast fjord]

I want to be able to concatenate crops.

[thin ermine]

It'd definitely be nice to be able to more easily pull SFX from the training data. There ARE a lot of sound effects in there, but tapping into most of them consistently is difficult. I think beeps and gunshots are two of the easiest. Rain, Thunder, Birds, and other nature sound are easier to get in v2 and v3. Here's a playlist with a few: https://suno.com/playlist/49a40042-a09a-421b-9df1-3f6d976221f9

[pulsar estuary]

assign codes or a personalized signature to each Persona so that they can be recalled in the Lyrics

[Persona Zx120op]
insert lyrics

[Persona Ux19y35]
insert lyrics```

kinda getting tired of finding a persona I've found by wasting more credits just to find it again
currently personas are stuck to their pre-existing style, but what I want is the voice

[pulsar estuary]

hmm think i might have found a way to consistently search for personas now :)

[golden lichen]

Please turn off human verification for paid users--

[weak heron]

The most important thing right now: Follow my instructions to create a song!
Suno is doing everything it wants, but not what I'm telling.
One example: I say [Male vocal] inside the lyrics, Suno does most times female.

[runic pier]

I have a theory about this, besides the fact that it's learning from multiple sources and thinks that your instruction "wouldn't make sense" within the context of the song, there might also be the added contradiction from normal Chat GPT prompting. (Suno uses chat GPT to interpret) Square brackets are used to note something that the AI can choose to do, for example "give me a picture of a hill landscape with a [house and white fence and a] river running through it."
You might see the contradiction, in music square brackets are also used for labels, instructions and extra info, so building the music layer on top of the previous layer might be what is causing instructions to be ignored more often.
I'm not sure this is 100% the reason but I feel strongly enough about it to have started using <these brackets> instead of square ones in most cases. This does, however have the added problem that the AI will sometimes read what is in the <angled brackets> so far though, this hasn't happened too much (but I have only written one song since coming up with this theory so.. 😬 try different things I guess..🤷‍♂️ )

[weak heron]

All i have found was that you can use [] for giving instructions inside the lyrics. I am missing an official instruction about this.

[golden lichen]

Are you sure chatgpt is involved?

[runic pier]

no just a user

suno user*

[golden lichen]

ah

[runic pier]

there isn't any instructions anywhere really but if you use any other AI prompting tools then you might find there is some crossover like using curly brackets to specify that ideas are linked or part of the same thing Example: {Sombre trumpet, Melancholic feel} -not too useful within Suno but I still use that every now and then, purely because sometimes instructions are ignored. I think it makes it more likely to include all the elements if things are grouped. Again though, I'm not an expert on Suno I could be wrong on this.

[runic pier]

sorry I misread that, I'm pretty sure like.. 70% sure that when you type in that box you're typing to Chat GPT. Maybe it isn't that model but it's very likely to use the same rules.

[eager knot]

I have a suggestion for Suno, when you are going to make an audio record I really want you to be able to create a voice persona, since you were the one who recorded it, if not, there is no point in you recording it and not being able to use it to monetize or etc.

[runic pier]

ooh, I have a question. Are you saying that you can't add uploaded sounds to an existing persona? I haven't used Personas much at all but was thinking I might be able to upload a bunch of clips, and add them to the same persona to give a vocal imprint and then make a song from there with the voice I want without the song being restricted to being private. Would that work? can you (or I) do that? (is that even how personas work??)

[vital prairie]

There are so many features. But one thing I find is that the song still just cuts off. Cant you fix that instead of worrying about sorting the trash by recency?

[jolly inlet]

I still think a slider called like "chaos" or something like that for covers would be a neat feature.

0 Chaos would just create a normal cover, and the closer you get to the max setting the more it deviates away from the original.

[eager knot]

That's not exactly it, I want to record my voice and make it become a persona.

[runic pier]

Sorry, I wasn't very clear. I know you meant that, I just mean do you have any knowledge about personas that would suggest that you can't Imprint a voice on to one from uploaded audio? Because that also does seem like an option unless you know something I don't.

[eager knot]

As far as I know, there is no way to make a persona from an audio that has already been uploaded, but if you record your own audio, theoretically it should work, yes, you can make the persona, since you didn't upload the audio, but rather recorded your own audio.

[runic pier]

Surely you don't need the ability to make a persona from a song, Because adding a file to that persona's song list is the same thing, no?

[eager knot]

Yes, that would be the same thing, so why was the persona made? To replicate the voice in other styles, right? The same thing would be your voice as a persona.

[runic pier]

I think Personas copy over more than just the voice, they replicate genres and style as well in cases where no style tags are given.

[eager knot]

So you can record your persona with that style and replicate it in other styles.

[runic pier]

again, really not an expert but I saw a playlist once in which the theme was "a Cappella only" -that was the only style tag and style influence allowed but Personas were allowed. And it made for a lot of variation between people's songs. Some were not A cappella at all.

so.. maybee??

[woven pilot]

it would be pretty damn nice if everything wasnt 4 mins smdh

[twin grove]

A couple (seemingly) easy-to-implement ideas:

1. Let us use Filters in Library
2. Add "Disliked" as a filter option so we can clean up workspaces a little more easily

[willow granite]

Pro Tip: get your API up fast, the new music gens hitting the market this year will have APIs, and are gonna be onpar if not better than your offering, and you guys have 0 retention due to not offering an API for people to build services around your product.

[thin ermine]

I don't believe that's quite the case. I believe the content moderation is handled by ChatGPT, but I'd be very surprised if it's actually modifying our prompts.

[sacred sluice]

You can click the top entry, then hold the shift button and scroll to the bottom entry to click (while holding the shift button) the last entry, you can now remove everything on the page. If you want to remove or move selective entries, hold the control key instead and select the specific entries you want to move/remove

Add a "create stems" button next to the "create" button to generate instrumental track and vocal track instead of generating two different songs. The goal is to generate clean audio and not splitted audio stems that contains eachother's harmonics.

[autumn basin]

you should REALLY change that if you continue a song in V4 it sticks with the sound at all and dont add a totally random melody, singing or rhythm... V4 is just a surprise box and it needs to be fixed!

[wise ore]

Hello, I've already sent an invitation to 4 of my friends to get 250 points, and I only received the points for one friend, even though all 4 of them did it the same way. What should I do in this situation?
(My Username on suno is Akgeri)

[upper crown]

Image from Me About UI/UX
Description from ChatGPT
In my opinion, the idea of using a more “compact” display (showing the first few user avatars plus a “+X” badge) is a better UX choice for several reasons:

Reduced Visual Clutter: Listing many names and text (“Mar Rod and 5 other liked…”) can overload the interface and make it harder to quickly grasp the information.

Easy Access to Details: If someone wants to see exactly who liked the post, they can simply click on the “+4” (for example) and view the full list in a popup. This approach saves space in the main interface while still providing detailed information on demand.

Clear Visual Hierarchy: Avatars immediately link the action to specific individuals—not just by their names, but by their images as well. When the notification says “X and Y liked your song,” we already understand it’s a group action. Showing a small badge with the remaining number of users indicates there are more participants without overwhelming the notification.

‘Don’t Overload’ Principle: The main purpose of notifications is to briefly inform users about an event. Additional details (the complete list of names) should be optional, accessible with a click or tap. This keeps the interface clean and prevents users from getting lost in too much information.

Overall, a concise notification with a popup list of users strikes the right balance for both UI and UX. The second option—avatars plus a “+number” indicator leading to a popup—provides a cleaner, more streamlined experience.

[hoary kite]

I use browser on Android (Desktop site OFF)
Here's some suggestions:

• add a way to rearrange the order of songs in a playlist
  (You can do it with Desktop site ON, but it's inconvenient, and extremely hard to control with a touchscreen because it's trying to scroll and move the song at the same time, and it's also extremely laggy for me)
• add a way to remove songs from a playlist
  (You can remove them with desktop site ON in browser, but it's very inconvenient)
• add a way to change the cover picture for the playlist
  (Would be very helpful, especially if the cover it put, is tona song you removed from the playlist)
• add a way to view the video while on the site
  (Currently I only can view the video by downloading it, or sending myself the link to it, and that only sometimes works)
• make it so when you change the cover of the song, it changes the picture in the video too
  (Currently it shows the new pic as a cover, but once you play the video, it shows the old AI generated image, that usually doesn't fit with what I want the picture to be)

[flat maple]

This simmer is killing me the sound of a be quiet but higher is destroying my songs and with no money back it is taking the piss. I love Suno but it is not enjoyable at the moment when you’re not happy with the outcome

[haughty dome]

skip/forward 10 seconds on 'Play Sound' ? Thumbs up if you think that could be useful.

[uneven plank]

Suggestion: A way to delete multiple or empty the trash bin. 😮 Instead of 1 at a time. 🙂

[knotty dust]

If you add mass deletion capability, would highly recommend that you make sure to also include a warning / disclaimer popup before executing such a command, to prevent accidents.

🫶

Little details like the animations when generating new tracks are nice and all, but IMO it would be super helpful (specially for users with larger libraries) and overall make for a nice quality of life update if Library searching were a bit more robust.

In particular (Listed in order from most to least important) I'd love to have access to the following specific features:

1. Ability to search our Library by both text string AND Liked 👍 at the same time.
2. A "Sort by" feature, for sorting the list of results by things such as track length or date created, ascending or descending.
3. Ability to alter the amount of results shown per page.
4. Ability to add negative search terms alongside positive search text, to exclude certain terms from the results.

Doubt I'm alone here, so please click 👍 if you agree with this post.

[uneven plank]

Of course that too. 😄

[native meteor]

https://tenor.com/view/meme-pikachu-pokemon-shocked-gif-16985133

[void phoenix]

I have a real good suggestion.

1. Get Stems

• This features is really good but there is a drawback, sound quality will never be the same, some glitches sound etc will be there. You guys have mentioned that as of now it’s in testing or still in developing process.

But I would suggest that to give options for the user to generate while generating a new song an option for generating Vocal and music separately, such that when they play together it’s a song. When this option is use 3 audio file should be generated one is the music alone other is vocal alone and last obviously the combined version. In this way the audio or music quality will not change and there won’t be glitches sound too.

Some people, including myself love to play or tweak with the generated music or vocals.

I am sure people will love this feature more than anything else, it okay to not give it in free version, this can or be the game changer for you guys.

I hope you guys read my suggestions in good health, hope SUNO AI will be at TOP of all in this Genre.

Please like if you think this suggestion is good.

[acoustic heron]

workspace or playlist bulk download

[violet sierra]

Exact search results on top! Approximate search results after the exact results tab is over.

[woeful mirage]

My guess is that they aren't generated seperatly. Your suggestion might be the ultimate goal for many, but they'd have to generate things differently i guess

I would love to see a filter for Workspaces in the Library !

[blissful sail]

Would love a fast way to filter, and select a large bulk of my old songs to move to trash and delete. For example songs without 👍 or songs with 👎

[slow walrus]

And how is remaster done? That clearly regenerates the voice and instruments seperate.

Because they change without quality issues there

[arctic whale]

Would it be possible to change the selection view in the profile?

For the moment you can pick between Top and New. The new seems to refer to the day of creation, so that songs that have never been published but lay a while in the library (for whatever reasons) do not show up as newly published although they are even if they have been created some time in the past.
I didn't want to publish old 3.5ers that I remastered in bulk, so I do it one at a time. Then I noticed that they didn't get much attention. Now I know why.

[rain kestrel]

I finally did it!!
I made the realtime Suno Lyric Video Player✌️

https://www.youtube.com/watch?v=Y611AoXZ1P4

[full badger]

I was wondering if ya'll would be down to make it to where you can publish a song but not the prompt! I wanna cook a meal for someone without having to give away the recipe, ya know?

[somber dove]

(This is for across the browser and iOS/android) I think it would be good if u could make it so that credits reset based on the user’s local time zone like for example someone in CST would have their credits reset at 12 AM CST while someone in EST would still have their credits reset at 12 AM but at 12 AM EST

[signal parcel]

I’d love to see multimodal support in Suno v4.5 or v5, where a generated melody seed guides both the instrumentals and vocals, ensuring a cohesive pattern. Additionally, having vocals and instrumentals generated as separate tracks—with stems naturally separated from the start—would allow for greater flexibility in modification and enhancement.

[rancid zephyr]

sorry for late replay how do we use this tho ?

[plain kiln]

Make a system with "secret smileys"
What do I mean and why could it be good?

• I mean that Suno users can click smileys on a song, maybe good or bad.

• but the name is displayed nowhere.

• the user gets no bonus for doing that and the smiley cannot be connected

• smiley already can tell a lot
  -It could help improve songs. As you get an honest feedback.

• of course, you can only compare to yourself, as no one has your fans

[stone steppe]

Feature Request: Precision Timing for Extensions

hello again,

i have a suggestion to improve the extend feature. it would be useful to let users extend songs by tenths of a second instead of just whole seconds or minutes. this would be a replacement for the stopped feature of fully integrated clipped extensions.

why this is important

• adjusting by tenths of a second helps align tracks better and makes changes smoother.
• this removes the need to crop using other tools, saving time and keeping your creative process smooth within the suno platform.
• better control of timestamps improves track quality, whether you're adding extra parts or changing the ends.

proposed solutions

1. Let cropped extensions still be identified as part of the original track, keeping them connected even after cropping.
2. allow users to extend tracks in smaller increments, like 00:00.1, for more exact timing.

Thank you.

[void phoenix]

It can also be done in single track but the generation limit has to pass more than 4 minutes. It would be around 8 minutes. First part can either be only vocal or track and the other part left version. In this way it can be in one go. Like when mixing in DJ vocals and music if it’s separately than we have full control over the output music. 😊

[haughty dome]

If I want to keep the recipe private it stays in a .txt file on my PC. Surely that's enough?

[karmic sigil]

Notifications don't notify if someone likes a song you shared that isn't public (or can't be due to upload vocal). They should.

[high forge]

Maybe the ability to extend a song from the moment it cut off at the 4 minute mark, sending the whole song again but a finished version, making it only generate the missing part of the first version, but returning the full song

[stone steppe]

that’s a different request from what i’m talking about. i’m asking for precise extension control at the tenth of a second level, not just a way to resume after the 4 minute cutoff

[high forge]

Right, both things would be connected to the "extend" feature 😅

[stone steppe]

they both use the extend feature but they tackle different issues. both can help though

[tepid trench]

Minor UI suggestion: Label what this is for (e.g. "Page"), at least for non-mobile browsers where there is plenty of space available. Either to the left of these controls, or just before the page number.

And/or somehow highlight that this is how you access your older songs, the first time a user has more songs than would display on page 1.

[north jasper]

A way to cut out parts, or condense parts without having to do a "Replace Section"

[livid vigil]

Please add a song length defining input. Thanks! Love this program.

When using instrumental mode, I have tried to get some vocalizations like "oohs" and "ahhs" but the prompt doesn't seem to be recognized. If I take it out of instrumental mode and ask for no lyrics it doesn't recognize that either, so there doesn't seem to be a way to achieve instrumental with vocalizations and no lyrics. Would appreciate any suggestions that will help me achieve this, but also suggesting that it be a feature. Maybe a toggle button for non-lyrical vocalizations, or training the AI model to recognize what that means.

Asking for monologue doesn't work well either.

[leaden marlin]

Grid display of our library please! So we can browse 4-6 pages of creations on one page, title pops up on mouse hover

Grid, and list (w/o thumbnails),, just like in windows browser.

[harsh remnant]

Please for the love of god a larger upload size. 2 minutes is not enough, I need at least 4 to make a half decent cover.

Charge the credits for it or whatever, I don't care, just give that extra time

[iron eagle]

can i suggest to have longer upload time or no limit at all?

im a bedroom songwriter. i would love to upload my full song here that i sang on my own and use the cover feature to polish everything & re-sing my song. alot of my song will have birdge right towards the end where the melody & even chord progression is different. with just 2 minute limit, it wont be in the upload so the ai will just sing my bridge normally or not the way i want it to

pls make it happen~~

[languid ore]

Bulk library download please

[iron eagle]

did not realize some one alrdy writing this while im writing my own suggestion about this too lol

[harsh remnant]

I'm the exact same position as you. Suno is sleeping on their biggest legitimate use case

[summer spindle]

This update really sucks.. nothing more to say...

[autumn mason]

Please implement better stem separation. The model must be able to output better quality? Just the lead vocals. As for doing the same for instrumentals. Now I have to first write the lyrics. Do a promt to get a melody fitting the genre and feel. Then do a stem separation. To be able to cover the stem vocals to get a better quality I just crop the whole vocal stem (marking the whole song and crop). Then I can do a cover entering something like studio acapella, lead vocal etc. It would be great just to have a slide button saying acapella like the instrumental. Then the AI renders an exact copy of the melody that has been generated before but totally dry and no harmonies.

[stiff heart]

Quick suggestion and not sure if anyone else as mentioned this, but it be great when cropping to have a tick box to add fade out so you dont get a dead cut off. I know this is easy to resolve manually, export use an edit tool re export, but you can lose quality each time you export but just a thought

[vestal geyser]

How about a feature where you can download audio and lyric videos from other users similar to how you can do it with your own?

[pulsar estuary]

there's a srs extension bug where the extension can degrade the full song and needs to be addressed asap

also the fact that extensions can change the beginning and/or end parts that's outside of the selected waveform section

[pulsar estuary]

isn't the notification capped at 16? just add a Like Visibility dropdown for each song (and visible for all users)

[long robin]

This could go anywhere - a channel or way for people to explore/find public personas to use & play with 🙂

[fair turret]

Pls give us the ability to add in missing elements of a track reproducing the track exactly but with the new element added in. So for instance I have a track a and it is missing a kickdrum I can recreate track a and i will get the same track with a kickdrum added. Same with vocals record scratching etc

[prisma kayak]

i want to be able to take a part from one generation and be able to splice it in another genration from the same song or any song, also being able to upload one mp3 as a main song and another song for the ability to reference an uploaded song or generated one

[runic pier]

Ok SO. hear me out.

New model.
It's Entire purpose is to clean up audio on previous Gens.

You train it against shimmer and static before release.

Users will then select "clean up" in the model select drop down and "cover" the song they want to clean up.

They'll get a song back exactly (enough) as it was but without all the noise.

Suno devs are championed for actually being helpful for once!

This tool could also be utilised in the "edit" page to clean up smaller, specific sections of a track similar to "replace section" except it would actually be worth using credits on.

You know what we ACTUALLY need though?? -An "Add to queue" button on songs.

For those times when you're listening to a song and browsing and see something that piques your interest. ya know “ I don't necessarily want to stop listening to this one, I just want to play this other one next. Like a temporary playlist.

[weak sphinx]

A way to fix song lyrics, sometimes I accidentally misspell a lyric which ruins the song. And I can’t fix the direct song with it without using credit to pull a different beat and vocals that don’t hit the same

[pulsar estuary]

only solution is finding a similar syllable

and it'll rewrite just fine

[weak sphinx]

That wont work 😭💀

Even if so i added some words cause grammar.

[pulsar estuary]

well idk what ur mistake is exactly so

[weak sphinx]

Listen: I misspelled and/or forgot to add a word to a line in the song. I am purposing a idea to be able to edit a song with all the same vocal style everything just will
A. The extra lyric
B. The spelled correct lyric

If you cannot understand idk what to tell you

[pulsar estuary]

yea and I said finding a similar syllable works just fine for me if that happened to me

[weak sphinx]

Why would we work harder then just get something simple a way to edit lyrics?

[pulsar estuary]

im giving u a solution for now, you're not getting that update any time soon lol

[weak sphinx]

Says who

Tf

[pulsar estuary]

tell me how long have Extending other ppl's songs have existed

and that'll answer ur question

[weak sphinx]

Is a purposed idea just meant to be submitted right in

[weak sphinx]

Tell me how past a bill has been signed into law

From the legislative branch to the president

[pulsar estuary]

what are you projecting?

[weak sphinx]

When did i say that it will happen in a quick manner

[pulsar estuary]

i asked u for the exact mistake and u regurgitated words i've alrdy seen

and started to act aggressively

blocked

[weak sphinx]

Oh im blocked is that meant to hurt my feelings

Because you
A. Didn’t elaborate
B. To me your saying something not important

You said you didn’t know my mistake

I then reexplain it in baby terms

So not my fault you can’t read, not my fault we cherrypick information

[kindred hull]

Hi! Several suggestions, from the very precise, needed and sensible to the most awkward...

1. The Extend feature tends to generate clips with lower sound quality (with a "metallic" feeling as if the sound were sampled at 22.1kHz instead of 44.1kHz). This is particularly annoying when the clip generated is musically awesome, but technically poor. I've tried the Remaster feature to undo these issues, but it doesn't fix them.

2. While generating a song, the personas used to create the songs in the list shown do not appear. Sometimes they appear when the song is fully generated, and sometimes you have to wait to the end of the generation AND refresh the page to see the personas. Please note, that the persona used to create the currently generated song does note appear.

3. The behavior of the playlist is somehow awkward. If you are on page 3 of a playlist and create a song, the songs created appear on top of the list, but the playlist does not return to the first page. I don't know if this is intended, but seems a bit weird.

4. The Reuse prompt feature copies all the attribute of the song (lyrics, style, etc.) BUT the Persona used to create the song. Again, this is not a real bug, but I think that the most correct way of reusing a prompt is copying ALL attributes (including the persona).

5. Now we get to the realm of weirdness. The Generate video is SO limited. It would be awesome to have some "Create custom visuals" feature. It would be OK if this feature would also use some credits. I'm not talking about really creating AI videos (like Runway or Kling-AI), but to have some presets with some parameters to add variation. Even simply having visuals similar to the old Winamp-style visualizations (psychodelic landscapes, colors following to the rythm, animated photos, etc.) would be great. This feature could also be a really useful tool for making Suno even more popular, as most of the platforms where users share content are video based (TikTok, Instagram, etc.), so it would foster people sharing their great songs.

[teal shore]

find a way to make "regenerate video" regenerate the video first time every time

[trim arrow]

Add a section to control more precisely a song that you want to create. So can control the strength of the prompts and if you want a short or a long song.
The strength of the instruments and if you want a more or less clear voice. It can have some sort of switch like the exclude prompt thing if you want to dig in those parameters or not. I think it will be easier to create something that you want more easily!

[rancid yew]

We need a much more easier approach, to generating, specific, and unique vocals for a character. To me personally I don't really feel like Persona does that. Persona just specifies a consistent type of song style. We need a feature where you can like choose, age, role (like tenor, bass, alto, etc.), and gender. Then click on generate voice or something and boom you have, Specific Character Vocals. Then you just add the name of your character like [Voice: Bob], anywhere in the song you want that specific voice to sing... Because it's kind of annoying when I do my best to specify a female vocalist and it keeps giving me a male vocalist instead

[slow walrus]

yeah at least it should follow gender wishes better

voice control improved would be huge improvement

also if the persona song has a female in the song, please save her too

and not just the lead singer

But what I would really, really love is:

Being able to use any extend to append on any song at any timestamp. oh god this would be so HUGE

because the AI can do better transitions than my human hand, and the full song could then be actually at suno instead only youtube

even if the ai would do no transition work on it I would gladly take it. would be better than nothing

[pulsar estuary]

ive found that the AI determines which gender is best suited for the lyrics. so if your lyrics has a masculine feel, then it will choose a male vocalist and vice versa if it sounds more feminine.
i don't put male or female vocals in my prompt most of the time actually turns out that i use mostly female vocals in the prompts for certain songs and those songs had its lyrics readjusted to be more "feminine"

[runic pier]

I was thinking pretty much exactly this earlier! Adding voice mixing would be really cool. you could choose the qualities you want and balance them to your tastes with different weights for pitch (high to low), things like smooth to gravely, nasal quality if you want to do something less mainstream. Then save that specific vocal profile to be used across songs.
You can Kind of do this already, It's implicitly within the AI if you consistently put [bob] and settle on the same for for him throughout multiple songs (and reinforce it with likes) then you can get very similar vocals on different tracks. but

but really it's just a shame this is the "Suno Fanfiction" section of the discord.

cus lbh, none of this is happening. 😅

[slow walrus]

ive found that the AI determines which gender is best suited for the lyrics.

I have a male lead singer, and a female (lucy)

Me and lucy then left the car
[...]

was forever sung as female xD

[pulsar estuary]

i meant in the music style, not in the lyrics

[slow walrus]

if I need just 1 voice, I just add male or female vocals then there

and put it as the first tag, first one shows highest interest

[pulsar estuary]

pretty sure suno has trouble recognizing that parameter, that's why it fails sometimes

or idk if it's related to how songs degrade after X minute

done it several times where the male vocals turn into female even tho i wanted female, had to replace section

was towards the end of the song

[slow walrus]

it reads it
it also reads
[female voice]
above an verse
But its always a suggestion, not a command, and thats what I dont like for genders

But if you need just 1 voice, then you can put it in stylebox. that one is followed better. And you decide the priority.
so if your highest priority is e.g. a female voice, then use it as the first tag

[pulsar estuary]

yep

[plucky yoke]

I'd love an api endpoint to generate music. Can't be that tricky right...?

[fair turret]

REMOVE SECTION TOOL -- Pls give us a tool to remove a secion of song between points a and b (where a and b are in the middle of song

[slow walrus]

👆 + copy and paste a section from one song to another
And of course what I mentioned earlier would be absolute huge
#💡┃suggestions-web message

Another thing:

Add possibility to zoom in and out in the waveform of the crop/replace

I mean it works, because accuracy is only .2 steps anyway, but you would have a tiny bit easier time to find the correct spot

[frosty ferry]

You know what would be nice? An option for PRO users to upgrade songs made that they generated from the 'Free' License to a PRO license, allowing them to keep songs they made when not a pro, and upgrade them to 'PRO' songs. This is very handy if a person made a song they like, but want to monitize it, but they werent PRO at that time.

[leaden marlin]

just now, i checked out someone else's song. you guys have some awesome pictures along with them. so i just found one i love. i won't use it. respect copyright.

make a DM system, please.

i want to ask this person what they use for their images.

i want to compliment people on their work. some of the music just blows my mind!

i want to, you know, make friends. this is a community after all and just discord doesn't do it for me.

i'd love to chat with others on the platform. my guess is that many more people would love that.

[zenith cosmos]

i noticed that "anime opening" was deleted as genre. it doesnt make an anime opening songs now. also it doesnt work if i write j-rock.

[ornate osprey]

Creating one song should cost 5 credits as it is now but remastering should cost 1 credit

[slow walrus]

remastering is doing a full length generation of the song. Understandable it costs the same
but replace should be lower price

[leaden marlin]

a little tight on the censorship here guys. i was listening to a song of mind and said out loud: "this is f*cking beautiful!". so i wanted to make a persona out of it. based on my exclamation. wouldn't do it. the ugly colored no can do thing popped up from below. so i had to give it a straightforward, clean name. suno is about art. about music the f-word is practically intertwined with it. especially in the minds of many artists. it's also one of the most expressive words there is.

[slow walrus]

and speaking about replace:

The most common bug is clearly these 2min generations.

Why we dont get a refund for these? This should be easily doable because:

A selection cant be longer than 30 seconds, so why you cant make a check like this:

If generation is longer than 1min (or 1:30min or whatever)
and if [x] Make same length as selection is selected
then refund

Because you would never ever use them. Because it adds more verses into the 2min than you did select, which would come on top of the regular. No one is gonna use that.

[deep monolith]

I want Suno to create their own Streaming platform and allow the users from the Suno community to submit their best songs to get played. We could have our own yearly award show rewarding the best songs. It’s the future Suno should be going in to survive and grow. I am waiting even if we have to pay another tier of subscription it’s a good future and move.

[wooden lance]

please suno people add a collab feature, so just like on streaming platforms like on spotify, if 2 or more artists collab for a song or even an album, the song appears on the profile of all the artists who worked together, so it also boosts the stats of everyone who made the song

[sterile stream]

Thank you, devs, for including a "comments" section in the songs. Great new feature!

[sterile stream]

That is next level. I'm dreaming of that. This would support those creators who aim to promote their tracks on the long-term

[slow walrus]

comments ❤️

thank you ❤️

[vestal geyser]

So... either it's implemented or it's a stupid idea. I'm going with the latter. Anyway, how about alongside a song, it also gives you the instrumental version of it?

[fair turret]

PERSONAS - PLEASE turn off the default public setting as there is no good reason for this. If we want it public we can choose to do so I shouldn't have to be on guard that it goes public by accident. Thanks to Bill Gates and lengthy TOS prompts it is quite easy to accidentally click through annoying popups

[pastel mauve]

Quick suggestion. If you guys want to incorporate a comment section onto Suno. Please put a block button feature as well. So that way I don't have to deal with people arguing with themselves in my comment section.

[marsh lichen]

Actually, you're wrong, I use those long fragments (by the way, they can last up to 4 minutes) when I want to make an intro or a middle section longer in my instrumental tracks.

[pastel mauve]

My case and point right here. I don't want to have to deal with this nonsense. If it was me, I would have clowned this man, blocked his low IQ Unga Bunga Self and kept it moving. So please give us a block button.

[slow walrus]

you use them by selecting a 10 second section and then you even press the:

[x] Make same length as selection

and keep the 2min track? Thats what you do?

Without the checkmark it would make a little bit sense because then you want a longer part.
But WITH it? come on thats not what you do

[kind temple]

There should be page-lists for albums so that albums with longer playlists can be properly shown in a more organized way

[hard wagon]

Heard that there was a comment section that you are working on. I'd just like to input a little here, I personally think that just commenting on songs doesn't really help with the issue of people not even listening to the songs at all. If you're going to release commenting in a plain form, maybe make it so that you earn the right to comment after listening to a song for x amount of time. It shows investment, that way it reduces spam and unwanted comments. Make them listen to gain the token to comment on a song. Commenting randomly should be avoided. If you make incentives for users to do thinks it will create a positive experience, even if you are requiring people to listen to songs in order to comment. TLDR; I want people who comment on my music to be invested

[slow walrus]

the requirement that you must have listened to the song is quite nice idea, that should prevent bot spam

maybe not the full song, but for example 30 seconds

30 seconds is also the requirement for spotify, tidal etc a user must have listened to, until the song is paid out

[marsh lichen]

Oh, you're right, it's definitely not fair that 'Make same length as selection' feature almost never works. But sometimes I'm even glad that it made a larger section.

[uneven plank]

Anyone else on browser figure a way to move or turn off the song bar or minimize? sometimes it covers up a song on a list for me.

[next barn]

Can SUNO go back to the Square art, than this thing? SUNO doesn't even look like a 'Website' anymore, it looks like a phone app.. Why can't they make the app, and site different from one another????

[alpine turret]

Can't create because 403 error!

[normal bolt]

When I ask SUNO to create lyrics for me, instead of sometimes telling me that it won't allow copyrighted material (hey, I didn't write the lyrics, SUNO did), just keep creating new lyrics until you have one that doesn't infringe any copyrights.
Also, don't let it create words that SUNO refuses to sing/rap anyway, like the n-word, for example. That comes up a lot when I ask SUNO to create rap lyrics.

[alpine turret]

Is it possible that my account does not support create?

[alpine turret]

When generating, it always prompts that the system has an error, 403 error

[true umbra]

Often, Suno just repeats the same track in Covers, Personas, and general song creations. However, when I use the Report Bug Feature, there is no option for Repeated Track or any way of adding the details of 'Error.' As this is a common occurrence, maybe this should be added as an option in Report Bug.

[gray yarrow]

Can we please have the option to sort the create track listing so new tracks appear at the bottom? This way new tracks are played as they are generating instead of me having to manually click up the list to play the new tracks.

I also agree with @true umbra I think I have wasted many credits on cover tracks that are basically identical to the original

[bold wraith]

Make an exception to non-Custom and add Persona there please.

Add more options to Custom such as: Topic, Key, Tempo, Vibe, Male or Female Voice, Vocal Style, etc

[subtle sun]

Please make the AI ​​ONLY SING THE WORDS INDICATED AND NOT ANYTHING ELSE.
Because, if I make the text "too short", the AI ​​will repeat a random verse and that's annoying, I don't want that.

[arctic crest]

It would be great to have a selection to download songs. More than one at time. Whole workspace?

[queen crow]

Would be awesome to have 32bit/48khz wav file downloads generated by actual high quality audio going all the way back into the Suno neural generation chain

[runic pier]

just a quick question as you put an idea in my brain. Have you ever tried adding a lyrical 'style suggestion' more than once above a tab to make it more consistent? I haven't yet and don't have any projects rn but do you think it would work the same way as duplicating a style tag would? for example:

"[female lyrics] [female lyrics] [female lyrics]
blah blah blah,
this is my song bit"

would this make it more likely to accept the command, do you think?

New Suggestion: Allow users to alter the amount of songs genned at one time.

2 at a time is great! but sometime both gens are good when extending and you may want to use both. this solves the problem in an easily implementable way.

[slow walrus]

no idea. but good idea. I should try it.

[lean coral]

while I love the new bits and pieces the team is putting out. I'd really prefer quality of life updates and fixes for actual production

• Individual stems - Drums, guitar, etc, all seperate
• stems that work. Lately the stems are either very muddled or missing key instruments. or there's bleed from the vocals
• Improved sound quality - there's still shimmer, there's hollow reverb in the songs, there's noise after the 2 min mark.

Comments in songs are cool. but that's icing. I need a better cake.

[slow walrus]

stems that work. Lately the stems are either very muddled or missing key instruments. or there's bleed from the vocals

of course they have if you just use center channel extraction from the full mix rather than having this for real seperate
Would be awesome if stems could somehow catch it before theyre mixed together (but maybe the ai outputs them only mixed together from the first step on idk

[lean coral]

That's my point. they aren't real "stems" well...stems in the meaning I know them for.

[slow walrus]

I just worry a bit about spam bots
Maybe you have to listen at least 30 seconds of the song until you can comment to avoid spam bots. They will come. we know there are bad people in the internet everywhere

[jagged chasm]

Feature Request: Automatically move downvoted songs to "Downvoted" workspace.

I keep downvoting songs which I made and don't like and it really clutters my main workspace. I have to manually delete each and every song.

My suggestion: Add a "Downvoted" workspace that will contain all of the downvoted songs; it also must have a "Empty to trash" button which automatically deletes all of the songs in the workspace.

Why?: It will help other workspaces be decluttered and instead of searching thousands of songs for those that you have downvoted (to delete them) you can easily do that with the click of a button.

[pulsar estuary]

yea... i think spam bots on comment section is a much lower priority than that extend feature if we're talking about bad ppl

they said they're "working on it" in an email so we'll c

[slow walrus]

how common is extend abuse though?

I dont even have the desire to do that, I want own songs

[pulsar estuary]

imma delete that msg so ppl don't get the idea

[slow walrus]

its indeed problematic that its possible, fully agree

[pulsar estuary]

not only that

im also deleting this soon

[static salmon]

@grave salmon what do you think about having the left side of the interface to be toggle bar? we dont really need to see it all the time and takes up valuable space 🤔

[slow walrus]

Playlist needs a search function (especially the liked songs playlist )

[near frost]

SImple suggestion to implement (I think.......?) A way to regenerate a share link to get past discord caching and using an old version of the song/video

[slow walrus]

Put a / at end of the link and discord will load it again because the link changed

[near frost]

Tried that and sadly it doesnt work

[slow walrus]

hmm @worthy dust gave the tip

[worthy dust]

you hate to edit your link .. just remove the last letter ... save ... then edit again and put the last letter in again ... save .

you have to wait 10 minutes after "regenerate video" or changing the cover.
someting like this: ht tps : //so.cm/song/8146e34a-9c7e-47ab-b8f9-3ebf60 <-- this is your posted link...
Edit your post with your link to: ht tps : //so.cm/song/8146e34a-9c7e-47ab-b8f9-3ebf6 (remove the 0) ... save your message.
Your embedded Video will disapear.
Edit it again to -> ht tps : //so.cm/song/8146e34a-9c7e-47ab-b8f9-3ebf60 (add the 0 again) ... save.
Then Discord will refresh the link with your new cover and new video.
If this wont work directly ... the waiting time between "cover/video change" and edit your link .. was not long enough.

🙂 but this way it will work everytime 🙂 you force DC to recache your link

[near frost]

Tried that. I regenerated the video days ago. ANy link I post, edited or not, still shows the old video from when I first posted it

[leaden marlin]

add gifting subscriptions

[vagrant oriole]

The comment feature needs an option to edit lol

[delicate cedar]

Or at least a way to tip people with a few credits.

[leaden marlin]

yes 🙏

[vagrant oriole]

Oooooh now that is an AMAZING idea!! pls submit multiple times until they implement it lol .

Like a "subscribe" type of feature to pay the artist with money too would be cool

[delicate cedar]

We've been asking for it for over a year now. The idea was this would be a way to encourage people who for whatever reasons could not afford a pro account but still made great music. Or a way to say 'this? do more of that!' with an exclaimation mark.

[vagrant oriole]

Agreed! I'd have to know this would be made an option soon. It seems they are truly listening, It's just gonna take some time to get to everything

[delicate cedar]

I said we've been asking for it, not that it had been approved.

It seems like a simple thing to us but it's likely much more complicated for Suno's part. Like if a song was made with a tipped credit, does the original tipper get ownership? These are the types of things that need to be thought about and documented. And I'm sure there are considerations that we haven't thought of that they have.

[languid ore]

But if you do approve the bulk Playlist download feature it be nice if it's only our own Playlists that we're able to download. In case people don't want their Playlists bulk downloaded by others

[fair turret]

Folder view -- please enable this in the library so we dont have a zillion pages and cant find anything there

[unkempt zealot]

SunoCleanup is a mock/proposal project designed to tackle a real issue for Suno users. Currently, deleting unwanted songs from the trash bin one at a time interrupts the creative flow, reportedly taking up to 15% of users’ creative time. This project demonstrates how automating the cleanup process could streamline song management, freeing up more time for creativity and innovation. It's a proof of concept aimed at sparking feature improvements that enhance productivity and user experience on the Suno platform. PS I HATE THE CURRENT SYSTEM HAVE I EVER MENTIONED HOW MUCH I HATE IT

Link to the project page to see the readme and proposal. https://replit.com/@TackiNO/SunoCleanup#README.md
Feel free to remix it / fork it if you want.

Link to test it -- will stay open for about an hour. (I have no clue I'm not developer / coder this is a first time for me. )
https://d1c3b60e-e0ef-40c6-92cb-acfa627c3ade-00-2811dps613q98.riker.replit.dev/

I just want to bring attention to better trash bin functionality

Let me know what you all think.
@prisma iron @void prairie @smoky dome @still parcel <--- sorry for the pings you all were randomly selected

[fair turret]

SELECTIVE STEM REPLACEMENT -- Please give us the ability to take an existing clip/track then give it directives like: Improve Vocals Adusjt Vocals (Deeper higher male female etc etc) Replace/Improve Instrument X Add Instrument X Add Backgrounds or Add Libs etc etc

[fair turret]

Also to download/Search for ALL full songs

[languid ore]

Yes or download ALL liked songs created by the creator of them

[glossy spoke]

I suggest releasing an official API

[leaden marlin]

refresh v4 tries

[tulip fox]

Added a bunch of songs to a workspace, but miss-clicked and a little note came up and said "Added to workspace", but I have no idea which workspace and I have a lot...

It would be great if when you open a song it showed what workspaces and playlists its in - only visible to the creator.

[scarlet shell]

be able to download a playlist

[hidden laurel]

Has anyone suggested having a method where you could change how similar the song was to the uploaded piece? Sometimes I'd like the variation to be a little more than what it is, but sometimes i prefer as close as possible.

[uneven plank]

A way to view followers and see songs created by the people you follow on browser.

[distant sigil]

They already have that for the web browser https://suno.com/me/following for people you follow and your followers https://suno.com/me/followers

[uneven plank]

When I go onto my profile, I can't see followers or the people I follow, only the amount. :/

And Now I see its in Library... sort of needs better way to view it from profile too?

[distant sigil]

Seems fine to me having it that way

[distant sigil]

In the library page then those two along with the playlists page is part of it

[vivid haven]

Make it possible to search for songs you liked from other creators

It would be great if the videos created would support a higher resolution and if it was also possible to have them in landscape mode so they can be published on YouTube.

[haughty dome]

Make comment boxes extendable like the Create-Lyrics box. You'd need to up the number of permitted characters, too. I've been thinking I could use the Comments as a notepad to help me tweak lyrics. Copy them into a comment field, edit as required when I notice where the song went wrong, and re-copy and paste to Create. It would help us to read/write any long comments, too.

[north shoal]

are all comments viewable by all users .. ??

might be a place where your greatest ideas end up somewhere you're not ready for

[prime carbon]

broooo its too much do something and explain why he dont want , give the list of what he dont accept exactly with word by word and all of that we loose too much time

[glass barn]

More accurate sounds and vibes to match the custom uploaded Sample

[prime carbon]

brooooo

[crisp ember]

Because there is a button for the new comment function, this button should work as a show / noshow button.

It is really annoying, that there is no way to hide the comment box.

The button could even be a tripple-state button, to show, that there actually are comments.

And the default for own songs should be to not allow comments...

[visual brook]

user story:
i listened to a song that goes into my history (me/history page)

i don't want the fact that i listened to it to influence any algorithm; it would be good to have the option to remove the song from my History items entirely (youtube has this feature, for example)

i also don't want to downvote the song, as downvoting is a different use case

[plain kiln]

Dear Suno team!
Great updates in the past days.
We will see, how the comments will play out. So far so good.

Now, 20 Persons liked my song ->WHO ARE THEY?<- ??
I can 3 persons, the other 17 have to be guessed. I am not good in that game.
I understand that your system solved a problem. But it created another.

It is difficult to untangle.
-> Maybe if the pics that you clicked on vanish... <-
I guess you get many happy people with this little fix!

[strong swan]

People cant have nice things. They will misuse it with for example Like spam. But i dont have to tell ya.

[clever briar]

Does this mean we will finally have Covers available to everyone for free, or am I reading this wrong? Have my suggestions finally been answered?

[median crater]

Can’t see how like spam will affect this more than it already does, at least with @plain kiln suggestion you can identity who they are … as it is now when it stacks the profile pics you have no clue as to who like your songs except 3 of them … those that use the new comment feature can be identified but not the rest 🙄

[plain kiln]

So you can spam if you know who liked your songs?
How so?

[strong swan]

To return the favor of liking your song by liking their newest for example. I usually get like flooded if people release a song to signal me to like theirs, also when my songs to through new they also get like bombed by the same accounts. Its a pretty sophisticated system of user made Bots and manual organized like bombing. And this has to stop. Just stop.

[pulsar estuary]

need a proper End Song Feature so I don't have to extend the song just to crop the extension leaving X minute behind to Replace the section to end it

[plain kiln]

Okay, but at the moment the user with bots are on the win here.
Well, if they react immediately.
Of course, you can check every minute.
But he bots could do that without problems (should they be sophisticated as you say)- if there really are bots

[strong swan]

If you could see the stats of thousands of likes from a single user in an hour, it would make more sense to you why i speak of "sophisticated" ... And why it needs measures like rate limits or grouping of notifications.

[pulsar estuary]

ive seen users w/ a couple of followers, no posts on suno but has 2k+ followers on YT where as people with a significant following on suno but no followers on YT
i don't think botting is winning anything on suno

prioritize issues that costs credits instead

[strong swan]

Just to get us all on the same page here. Yes, you can see who is giving likes and receiving them. You can see who is releasing songs, You can see who is in the trends. And if you log that you get extremely interesting graphs.

https://tenor.com/view/look-at-this-graph-nickelback-gif-12683147

[pulsar estuary]

simplest solution - Like visibility, see who's liking the songs, who's following, etc then it becomes clear who's botting

[median crater]

Well, that’s the problem … you can’t
If you get 17 likes on a song within a certain timeframe you can only identify 3 of them

[strong swan]

the user doesnt need to see this data. they can if they have a really deep need, but its not a thing that a casual user has any advantage of. its the platforms job to get rid of bots.

[pulsar estuary]

just add a drop down..

[median crater]

Beg to differ, I’ve heard many that like to see who like their songs

Start a survey and I’ll bet you’ll be surprised how many would like to see just who likes their stuff 😉

[strong swan]

if you really want you still can. Just not where you usually look. I will not go into any more details, because that probably was already too much help for potential abusers.

[median crater]

Cate to share that in a DM then?

[strong swan]

of course not.... under 100 percent gurantee not. never. and thats no joke.

Because... graphs.

[median crater]

Eh … okay?

That doesn’t make sense at all

[pulsar estuary]

lol

[median crater]

Well … whatever, have a nice weekend 🥳

[strong swan]

The message is : If you really want you still can. ... And its not : I will show you how to cheat. You can do that tiny bit of research by yourself... thats the "really" part. Dont take it personal i dont trust anybody, neither online nor offline. But thats a me thing.

[median crater]

I’d rather spend my time making a new song over taking a degree in computer engineering 😉

[strong swan]

In theory thats what should happen. Spend more time into making quality content, rather then pumping and pushing out medium songs with forced topics... Doesnt mean do you that, its just a observation of the current overall situation.

[haughty dome]

People were gaming the system. I had someone like 8 of my songs inside one minute. They were only doing it to pique my interest so that I would take a look at what songs they made. And then you wonder how come they have songs that get thousands of views/likes in less than a day when the song is nothing special at all? It was all to get on Suno's Home Page and therefore push others off the Home Page.

[pulsar estuary]

im aware, cause botting is an ancient problem. im saying add Like Visibility so everyone can see who's liking who and who's following who like every social platform

combined with allowing Likes after X minutes seconds of listening, it'll deter gamers/botters

[surreal elk]

Please let us sort whoever added our accounts as "fans" by the oldest / most recent, because I've got so many notifications that I can't see who just followed me, and I kinda like following people back. Thanks ~

[spiral escarp]

I'm new and I like to see who likes my stuff so I can check out theirs and see if I would like theirs. similar tastes and all

[wind solar]

Is there a way to make just one at a time?

[cobalt ibex]

Would be nice to have a "sort" for the versions of songs...so we can see just the version were working with at the time..

[robust jewel]

V3 alpha 2-22, full release 3-21 (29 days). V3.5 alpha 5-24, full release 5-30 (6 days). V4 alpha 11-19, full release still unavailable (80 days and counting). As such, I'd like to ask that the banner ads for v4 at least remember they've been closed.

[tidal star]

could we please have a pin song option on our profile 🙏

one that moves a pinned song right to the front of our profile so we can show a song we're proud of right in front

[fair turret]

END SONG FEATURE -- Please add feature to end song as quickly and cleanly as possible. Maybe with a coiuple options taper sound. big finish etc. Having to spend hundreds and hundreds to try and end a song is not right. This happens wayyyy to much thanks

[next ridge]

Is their an actual support for Premium subscribers, I seriously get soooo many captchas. It's getting to the point it would be hard to use my credits because of it.

I'm in US, static IP, using the website normally....

Also how in the world is building not the correct answer for this?

[next ridge]

Seems like should be pretty easy to vet premium accounts for bots, so paying people don't have annoying experience.

Or at least just verify me per session or something

Anyway I love suno just feedback, if you need any info happy to provide logs, or sys info etc, I'm developer, ad block off, nothing fishy about my setup, no proxies, just a plain browser window and hitting the button.

setFocusedKey causes a re-render and the re-render causes the collection to update, then the updated collection triggers the effect I think.

Finally maybe not bug but site 404's attempting to get: https://suno.com/auras/Aura-05.jpg

[earnest pewter]

Hi all can we add a page index to library have 200+ pages and take me long time to get to the first song I create

[next ridge]

being able to access workspaces outside of the create screen would also be nice (plus re-ordering in the workspace)

[stark onyx]

Is it possible to divide the prompt in 2 one day?

1 Prompt for voice and the other for instrumental.

Because I have the impression that there's a "default" voice that comes up all the time and frankly there's not enough space in the prompt to choose a voice + an instrument...
This avoids the need to use artist names to copy them, as there's more space to make a promte dedicated to the voice.

Hopefully my suggestion will make it to the top, thanks !

[rapid tinsel]

Hi everyone. Who can I speak to about billing? My card was charged twice!! Reached out to the billing and support emails I’ve gotten no reply! Just an automated response each and Everytime. I love suno and all but I’m sorry the customer service is ridiculously terrible. Please someone help!

[cobalt ibex]

Mentioned this yesterday.. but really need a "sort" feature for models. Would help for those that are "remastering" songs from an earlier version model...

[sweet flicker]

The option to Pin comments.
Many of us, myself included often immediately check the comment section, having the option to pin comments as the creator of the song would be a nice to have feature ^^

[kindred hull]

Don't know if this is technically feasible, but I have notice that some created with personas can not be used to generate another persona. However, sometimes the new songs are distinct enough that they deserve to create a brand new persona.

[vague zinc]

Hey. Suggestion: Fuse Extension.

Once an extension of a song is made, click to make it fuse with the original (limited to it, since its the data father of the extension child), a Smart Fuse function so people do not need to use external musical or sound loop programs to extend the song in a way that you have to cut, paste etc. The function needs to have a timeframe you put, or recognize where the extension of the father song was extended from, and thus have that data to refere to when fused.

And hello.

[tepid trench]

Not sure if this is what you mean but, on the extension part, you can click on "..." and go to Create > Get Whole Song.

(If you have multiple extensions, you can do it on the most recent one and it will join all of the previous ones together in one go.)

[vague zinc]

Will try, Didn't know. Will check it out shortly.

This i Exactly what I wanted. Cool @tepid trench

Thanks

[tepid trench]

Great 🙂 glad I could help!

[haughty dome]

I mentioned this a while back.

[haughty dome]

Reach out to @grave salmon

[leaden marlin]

I'd love to be able to search on the home page instead of trying to scroll through all of the genres, simple search bar, [search: "gothic rock"] [search: 'Heavy Metal"] for example. bring up songs that tags match what your looking for

search in the workspaces would be great too

[random fulcrum]

I feel like Suno has gotten way more censored. Can we really not use the word “Kill” in our lyrics??

[marsh lichen]

I really think that Suno must have the feature where we can turn on/off any censoring. Come on, more than half of rock music is NSFW in their lyrics.

[delicate cedar]

[Removed links]

https://tenor.com/view/anime-girl-gif-17974732

We...can't?

Lyrics by Remi so forgive how horrible they are.

[slow walrus]

had never issues with that

But I write german lyrics

[random fulcrum]

I tried making a song based off the “Taken” monologue and it wouldn’t work unless I removed “and I will kill you” line

[slow walrus]

weird

maybe the issue is the person you want to kill in the lyrics? otherwise idk

[delicate cedar]

This monologue here: https://genius.com/Liam-neeson-taken-phone-monologue-annotated ?

Yeah, that's copyrighted. You're not supposed to be using it.

[languid spoke]

This may have allredy been requested but I would love to be able to take a recording of my voice and use it as a bass sample for the ai to work off

[humble grail]

Next Contest could include other countries too, and, in case the problem is related to $ transfer, you guys could provide "free credits" for like an year or so.

[acoustic creek]

I would appreciate a feature where it highlights where vocals are detected in an uploaded recording. It sometimes detects vocals where they aren't and I'm curious if there is a pattern.

[humble grail]

Did you upload a humming or a whistling?

[acoustic creek]

Nope. Just a backwards playing guitar.

[humble grail]

hmm.. that's new to me.

[acoustic creek]

I'll show you

https://suno.com/song/1046b966-1512-4db9-afec-4edfb4cdd36a

I think its the guitars, because the blizzard sound wasn't causing a problem before.

[humble grail]

Yeah, it probably detected something.
It detected something in my whistle months ago.
I worked on a song for hours. Many credits.
Couldn't make it public. Such a nice track! =/
https://suno.com/song/a7f24bde-0c76-4109-873d-6a64848049a7

[acoustic creek]

Aaaaaaaa.

[knotty rain]

Personally I'd just love to be able to get sheet music from the songs -- lead sheets, Vocal/Guitar/Piano, whatever.

[crude river]

If you could merge two songs with the same genure as the first audio that would be dreamy

[midnight basin]

I know that ending songs can be a problem sometimes, especially instrumental tracks, so I was thinking a toggle that creates a track or an extension that is 30 seconds or less. Been having a lot of trouble with an instrumental track refusing to end when i want it to, and it doesn't help that i want it under 1:45. Cropping is nice but can lead to a song abruptly ending on a off note, so a 30 seconds or less toggle could fix that i think.

[novel wren]

i hope iam on correct room for it... can we get at "instrumental" at all versions (much) more signs in text box? actual is instrumetal only gambling for me and not really a good thing because normal without instrumental is so hard bugging at generatiing ..

[crude river]

Cropping songs or putting in a time for the song like 30 second music or 11 minute music

[vale sphinx]

A "female, both, male voice" button, especially for cover

[tulip fox]

Users are stealing music from others on Suno. Please disable text selection for lyrics to help prevent this. Additionally, consider using methods like rendering lyrics as images or on a canvas, encrypting the text, or dynamically loading lyrics to make them harder to extract from the source code.

[slow walrus]

any phone can read text from images

probably also google lens

if you want to prevent it, you need to remove the lyrics from your song

[knotty rain]

I think you might try cropping it a bit further, and then Extend it with a new ending, you might get better results.

[inland bolt]

I've been thinking about this a lot lately and I think that many users would appreciate having their own voice characteristics to use for AI generation. A way to train the AI using your own voice and create a personalized vocal character. Not quite like the persona, that tends to be buggy right now and create ghost voices and lyrics that you don't input... But something that only influences vocal quality. You would have to put safeguards in place to keep people from using other people's voices besides their own, maybe make it so there's a certain phrase or song you have to sing, and if it doesn't match up, it doesn't work to create the vocal characteristics profile. This would also give more variation to the vocals produced by SUNO because currently, a lot of the songs sound like they come from the same singer.

[languid spoke]

Ye

[midnight basin]

I had tried that, but the shortest extension I was able to get was still far too long, being around 1:30 long, when I want the entire track to be no longer than 1:45. I really like what I do have, but the best place to crop for an ending still caused it to end abruptly. I did try to do an extension in 3.0 which does end a lot better than other versions, however getting the full song caused it to sound duller with certain key sounds not sounding the same afterwards. I am assuming this is a product of it converting the base 4.0 version into 3.0 due to the extension being 3.0. I have also tried dozens of meta tags, but I have yet to find one that can really shorten the length of a 4.0 extension. This is why I propose a 30 seconds or less toggle, as this may be an easier solution to impliment for now than fixing the issue with the model as a whole right now. Another user after also mentioned that they would like a custom length option, where you can set a min/max variable if i understand correctly, which I believe is also a good idea.

[neon hamlet]

Hi, may we get an ability to send push notifications to our followers when a new song is published? Maybe limit to one or two songs a day or so?

[inland bolt]

I thought this already happens? I get notifications that the person I'm following released a new song.

[tepid trench]

This was added recently, along with comments. For example I got a notification that Mr.Tree had released a new song, as I follow him:

[short notch]

Dear Suno AI Team,

First of all, I want to say that I truly appreciate the amazing work you're doing with AI-generated music. Your platform is incredible, and it opens up so many creative possibilities!

I wanted to suggest a feature that I believe many musicians would find extremely useful: the ability to export sheet music or MIDI files from the generated songs. As someone who plays an electronic drum kit at home, I would love the option to access the drum notation or MIDI files so I can play along with the AI-generated tracks.

This feature could benefit a wide range of musicians who want to practice, remix, or perform the music created with Suno AI. It would make the platform even more versatile and engaging for instrumentalists.

I’d love to hear your thoughts on this! Keep up the great work, and thanks for considering this suggestion.

Best regards,
[Notas Sintéticas]

[tardy bear]

Masters, help me
I want to make a pure inst, but I don't know the command to use. Can you help me?

[cobalt ibex]

Maybe put a limit of 30 days for junk to stay in trash. but no way to perm. delete playlists, personas...or clips created from covers or personas..is definately something that need attention !!!!!

[grim sinew]

Hello how to apply brandambasador

[prime rock]

At this late date why does v4 not know what [end] means? Surely it’s possible to program it to stop when someone asks? Spending so much dang money on a song that stopped in the middle of the final note at 2:38 when it could have just finished it, trying to get it to finish that one note and not just ramp up again into another song. Come on guys, do better. I can’t even trim these and get something useful because it never shuts up so there’s no place to trim it.

Trying to get this to shut the hell up, I offended the gods of Suno and it started spouting random numbers at me https://suno.com/song/5c616206-2b8a-4f68-81f3-331792fbc220

This looks like a straight up bug where it’s spewing programming instead of words

[distant sable]

please please please let me extend more precisely... like 0.1 seconds precise.. sometimes i cant extend before a certain drop

[trail wharf]

Mixing

and ai Music video generater

[leaden marlin]

I suggest Google Chrome

[wicked ravine]

come check out my server

[trim lantern]

There’s always room for improvement in the field of accessibility. While it might seem trivial to some, web accessibility is something we can continuously enhance. When listening to music on Suno, especially when lyrics are included, it would be really cool to highlight the exact timing of the lyrics in sync with the rhythm. This simple addition could significantly improve the experience for people with various disabilities and provide a better way for them to connect and express themselves through music. Music has always been a powerful bridge for humanity, and making this small change would show respect for everyone, enhancing how people experience and engage with music.

[pulsar estuary]

not surprising, but what's stopping them from typing out what they see or pasting a screenshot into chatgpt to have the bot type it all out?

[trim lantern]

Another fun feature that Suno could consider is a space for karaoke! 😄 This could really help connect the community in a lighthearted way, giving everyone the chance to enjoy some funny moments together whether singing badly or surprisingly well! It would create an engaging and entertaining atmosphere, turning ordinary days or nights into memorable experiences. Who wouldn’t love a little karaoke session to break the ice and bring people together?

[fair turret]

LIBRARY FEATURE: Please add a label showing which workspace an clip or song is in the library if unassigned please show that this would be majorly helpful while organizing, Allow to sort by unassigned (by time and date as well so they will group as normal in library)

[tulip fox]

Nothing, but let's make it harder for them.

[tulip fox]

Also, these people who steal lyrics are obviously lazy and I would say they are too lazy to type the lyrics out.

So maybe there needs to be a script that scans the Suno database when people hit create?

I write mine from scratch too... but dont hide my lyrics, but you've got me thinking.

[pulsar estuary]

not every1 is this smart tho so im deleting my msgs k

[delicate cedar]

Please don't advertise here.

[stone steppe]

repost in hopes it gets added one day
#💡┃suggestions-web message

[royal venture]

Id like more specifics on how things work in suno, from meta tags, when hyphens are needed, if at all. What things can be controlled specifically with guarantee through tags in your lyrics, etc. I'd like specifics on what absolutely doesn't work.

(If you are actually training your model correctly, tagging specific things about the music you are training the AI with meticulously, then you do know what works and what doesn't, because the training is based off of your direct input. If you are not doing this, then this will ultimately be the worst in the long term)

I find it kind of crazy you can spend all this money on Suno, and never be able to create exactly what you want. You can waste so much time and money meticulously testing things just to see if it works, and ultimately be left with nothing in the end.

Without significantly better information and control, there are competitors who will eventually improve past what is provided here. Especially when their taking existing feedback about suno, like the garbage image sizes, and implementing better solutions.

I've been here almost a year, and the amount of feedback that is never listened too is astonishing. Not even an acknowledgment that you are listening and aware of these numerous issues, making an honest effort to improve them, and so on.

[green pasture]

I just don't like being tested with the bot check every time I turn around when I'm in a sesh...

[tulip fox]

I want to be able to tag songs in my workspace with colours, like red (bad), orange (ok), green (good)... Blue (relisten) ... or similar.

[tulip fox]

Wouldn't be good for the colourblind though.

[fiery wave]

Lil ✅ next to 🎶 that was created while subbed 💶 🤝

[feral grove]

A feature for fade out music, additionally to the cropping feature would be great.

[tired vault]

please make the music generator algorythm learn from the songs you are deleting/reporting/disliking

I hate spending 900+ credits per song then wait until next month to get them all back just cuz the algorythm doesn't stop repeating the same error again and again even after i replace section or cover the song its just so damn stupid Ugh

If i want to change the lyrics of a song, i need to create millions of covers until it works out, but when i successfully get the lyrics right on a cover it automatically ruins all of the structure, vocals or instrumentals, it's so annoying and crazy how i am spending real money to get nothing but bad quality glitched music with bad structure

do the devs ever read this channel or

cuz i'm not the only one saying this

O

[untold grove]

I would love two new features...... a fade option to fade the last 5 secodns of a track... and I would just love to have the ability to Cover with a persona... just wanted to toss that out. 🙂

[pseudo eagle]

Three suggestions:

1. Intro Extend, the same way extend works to extend and edit the end of a song, we need a way to edit and extend the beginning of a song. So in extend it starts at a select point and edits everything after that point in the song. We need a feature that does the opposite, something that allows you to edit everything before a certain point in the song. This way we can correct bad intros of songs that are good but just have a bad start.

2. Spot correction, choose a area of the song to effectively mix something completely new without effecting the unselected potions of the track. Make this support very small timeframes such as 2-3 seconds minimum to up to 1 min maximum. This allows for fine tuning, removing strange noise or samples, and fixing bad lyrics without the need of replacing good sections as well.

3. Accents & Languages, official support of languages though a language selection option and even accents would be game changer. Imagine being able to incorporate two or more languages into one song. Accents can add some authenticity to the vocals. This could be achieved with tags like [Irish-English Accent], or German-French Accent] for example. Accents already happen randomly in generation, this would give more control to the users.

[queen ivy]

hy guys. is there a way to create extension for only specific duration (i want to extend it for only 1 minute and 15 seconds)

[delicate cedar]

No.

Or at least not yet. The best you can do is try to adjust your lyrics prompt and roll the dice.

[queen ivy]

thank you. is every extenstion is the same 2 minutes?

[delicate cedar]

It can be up to two minutes.

[worn lotus]

Really wish the video let us put in pictures for each part of the song and had the video wide like youtube and words larger so it was more like this: https://www.youtube.com/watch?v=BwQ0QLurPCQ

[wintry kraken]

Suno desperately needs PERSONA delete function!!!!!!!! Its missing in the trash option. Please!!!!

[haughty dome]

Explain more.

[rancid pasture]

Is this going to be made possible soon?

[haughty dome]

Maximum 2 minutes Then Create -> Get Whole Song -> Then Edit Crop to keep the good bit. Cropping doesn't cost you. It generates a separate file.

[haughty dome]

I've learnt that if something isn't working, it's never going to work. Best to change genre, push out some likeable stuff, and feel better about life.

[wintry kraken]

After generating a cover song it is somehow generated with Persona ( but never used it ) and after moving audio file to trash i can not longer delete the file.

" Clip cannot be deleted, due to associated persona ".

There some personas in the trash but there is no option to delete Persona from trash then to delete a file.

[haughty dome]

Ok, hopefully a dev understands.

[untold grove]

V4 does not want to end a song.. maybe a option - [End:10] to cue the AI that you want the extension to end in 10 seconds... so it has a reference that you want to cull the song? I almost never get a song extend that is less then 2 minutes..

[trail wharf]

song mixing

[dusk lynx]

i would like a feature where you upload a instrumental song, and suno would apply what lyrics you type on to it

[lunar violet]

Browser login issue

[delicate cedar]

Are you logging in with the correct account?

[lunar violet]

Yes I login correct account, sometime login button not work, sometime OTP not received,This is browser problem

One another suggestion , Your respected website homepage overloaded and consume havi data. Please make your homepage signup or login page lightweight

[fair turret]

BACKUP -- Please give us the ability to download all our files to back up locally. Also by works spaces batches if possibl (prefer to keep any organization done)

[tawny relic]

That would be dope as hell, but the current upload time limit is bad for that. This new feature, if it supported enough input length, would need to be able to take additional markdown so we could timestamp per block, per line or if so desired, even per syllable, for the spots that suno just won't be able to get right. completely omitting timestamp tags would then be like a "full auto" mode. but how would suno know where to actually put the notes, if you were to upload an instrumental track that has no melody? (~edit: maybe it could optionally take an additional midi track that can act as a temporary guide)

[fair turret]

Would like to expand this to uploading full songs (not just instrumental) so Suno can redo them but keeping the basic song (make it better)

[tawny relic]

Yes, basically the musical variant of image upscaling

[fair turret]

Also gettin g with the option to change vocalist etc would be really useful

[tawny relic]

Could offer two modes. One that operates on the signal level and another that analyzes and then tries to completely "re-render" the input.
Lemme elaborate on the first mode... what I mean is it should operate like you would manually when enhancing a mix. Do stem splitting (drums, voice, bass, and as much single instrument types as possible, not sure what pretrained models are available right now. piano and guitar maybe?), then do an "intelligent" software based remixing of the stems to place them more optimally in the stereo and spectral fields through panning and EQ'ing each stem, to make an overall more full, rich mix with stems, that do not drown each other out, especially the vocals.

The second mode would works similar to the existing cover+extend functions of Suno 3.5/4 (or the v4-remix feature). Not sure HOW exactly those work, but I imagine a mode that tries to re-render each stem or instrument seperately while staying as close as possible to the uploaded one (maybe not just improving the signal quality of those instruments by re-rendering them completely, but maybe even fixing timing errors of each note to a degree. Maybe offer some adjustable setting for each sub-feature? (deactivate or adjust amount of timing and pitch quantization, skip single stems/instruments, etc etc)

Oh wait wait wait - in the context of the suggestion of @dusk lynx , generating lyrics over instrumentals, the above re-rendering thingy should offer ~~replacement ~~ correction of single words of an existing upload or Suno generation: Offer to correct spelling as well as timing and pitch, so we can edit the melody of Suno generations afterwards to a degree to enable users to keep good generations that just have messed up melody.

[tawny relic]

Oh, finally my suggestion that I previously came here for 😂
Display the licensing status of a generated Suno song for the user, because who remembers when they had premium active and when not after 1000 generations and 12 months later, right? 😅 That could include an option to 'buy' single generations into the licenced state.

[tawny relic]

I know you're probably talking about changing it on the platform itself so the result is publicly presentable on Suno itself (or even about some sort of additional "voice style" input in the create prompt, that would coax the model to follow that more closely), but if in doubt, there's always the option of replacing the voice with RVC afterwards. This... could ALSO be a future part of the Suno platform. While there are already established ones offering this and are centered around it, the purpose would just be to make it more convenient and offer a more rounded workflow experience on the Suno platform itself.

[distant sigil]

My suggestion is create a page that detailed shows what users are allowed and can’t create. I think it’ll be good to have on the site and call it Community Standards for it. Similar to this one ai platform but have it music themed.

[clear folio]

I know others have asked for this feature, so I join: it would be great having the possibility of creating covers with other personas in your list. I would like to make duets or more version for the same "concept album" and I need the same singers. I would like to experiment the same song in many versions, with the styles and voices I chose, too.

[gloomy python]

https://tenor.com/view/love-i-love-you-i-love-you-so-much-i-love-you-very-much-i-love-you-this-much-gif-14062478292165783442

[END ALREADY]
We need SoMeThInG v4 will obey to end a song in 5-20 seconds and not generate another 2:00 that may or may not end. I'm still stuck switching to v3 go generate a proper ending even on a v4 song. That's kinda sad. Maybe [End It All] or [Kill AI] or [IXNEY ONGSAY] I'm not picky. But it sure would be helpful. Here. I know you guys have been busy making awesome, so this heart is my appreciation. I just need to end a song well. thx.

[wise blade]

An Arrangement Editor would be cool. Kind of like this as it would enable us to do multiple replaces within one song.

A tempo slider (a pitch slider also) and drop down menus as an easier way to change instruments would be great as well

[inner forge]

You can edit the lyrics, and remove them. I put little blurbs on all my public stuff, since I have my lyrics saved elsewhere, if I need them,, but it would be really handy, if we could toggle lyrics/covers on/off. The lyrics are needed for making changes, and creating from the same prompt, and it's a little tedious to have to replace them, every time I want to do that. It is possible, though. You just copy/paste them back in.

[rare obsidian]

The /chirp command is no longer working

Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com

[gaunt wave]

I would like to see a button to disable the 'extend' button and the prompt.
Until then, all my music stays private, just like many others.

[atomic bloom]

I was thinking and I know i am not the first lol. But a feature request to be able to download the Midi Files as well.

[wise blade]

I don't think Suno uses Midi or VST. I really can't explain how the inner workings of Suno are but I'm pretty sure only wavs and mp3s can be downloaded.

[atomic bloom]

Correct. But I believe they can create a simple midi writing script in the midst of the ai code?

which will tell the ai to also create the midi file?

[wise blade]

Some AI platforms do use midis and stuff that can be exported into DAWs and trackers though. I believe Loopmagic does that. The only thing is, Loopmagic is pretty expensive. The good part about Suno is that its reasonably priced.

[atomic bloom]

i dont see why not? still data or digital.

[wise blade]

I hear what you're saying though

[atomic bloom]

which is why i love suno hehe

[wise blade]

There are a lot of features that Boomy and Loopmagic have that if Suno put in those features it could give us creators more control over what we create. So I'm all for that. The thing is people will likely tell you here to just use Meta Tags, DAWs and jump through hoops to get things done. Its not very intuitive. So I suppose my overall suggestion would be to somehow design Suno to incorporate a lot of those features without people having to come up with wonky methods to do it.

[tribal forge]

How can I ensure that my original voice remains as unchanged as possible when uploading vocals to Suno? Are there any specific settings or techniques that help preserve vocal characteristics?

[buoyant token]

hello.

I'm doing well with suno.

Is there any way to download all the music in a workspace at once?

If not, are there any plans to open one?

[pulsar estuary]

add a block user function so i can block bots randomly following me when im not even posting songs

also make sure block prevents the user from accessing the other person's page

[pulsar estuary]

it's better if u just crop the entire song to get a cropped version and edit the cropped version instead of messing w/ the original. cropping doesnt cost credits

[hidden ibex]

Could there be a Random song button in the playlist sector?

Could there be a way to rearrange the songs in a playlist to change the order?

[frozen sparrow]

if there was a way to upload our song image as we are genning. This saves a little time and then we wouldn't have to regen videos after changing lyrics or the image.

[untold flax]

I have generated 4,500 songs since version 1.0, and I really like the evolution this AI has had, but there are still many improvements that could be made—or at least that I’ve found necessary based on how I use the AI. I’m a paying subscriber.

Here are my suggestions:

Allow users to select the tempo of songs instead of relying on a random factor or requiring an uploaded sample audio.

Provide a list of musical instruments to choose from, rather than having to include them in the style box and leaving it up to the AI’s randomness to decide whether to use them or not.

Enable the creation of a song with two lyrics: one for the main voice and another for a secondary voice or backing vocals.

Let users choose which voice will sing the lyrics (or multiple lyrics) from a list of options. I’m tired of specifying in the style box that I want male voices, only to get female voices instead.

Add the ability to select two previously generated songs and turn their mix into a mashup without having to download the audio files, combine them, adjust them to 120 seconds, and cross your fingers that the AI understands what you’re trying to achieve.

Allow continuing a song in one style (e.g., rock) with a different style (e.g., opera). Currently, if you try, it either throws an error or the musical style doesn’t change.

I believe these changes could greatly improve the user experience, especially for paying subscribers.

If you could also fix the issue where a song cuts off lyrics halfway through, that would be fantastic.

Thanks 🙂

[pearl walrus]

I would have liked to be able to upload my songs to spotify.🙏🏿

[marsh lichen]

Subscribe to a music distributor that accepts AI (DistroKid, Too Lost, LANDR) and have them upload your music to all major platforms, including Spotify.

[small cypress]

We should be able to choose how we want the AI’s voice to sound like, like harsh aggressive or sweet or stuff like that as i had a song i wrote the lyrics i needed the voice to be harsh i loved the beat of the 1st one but it wasn’t harsh enough and the 2nd wasn’t a fan of the beat and was no where near harsh enough

[warm python]

Hello, dear developers. I would like to leave you my feedback. I am a user of your product. After version 4.0 I was very satisfied with the quality and variety of voices that the new version provided. But in recent weeks, there has been some degradation, I try to create tracks in different styles, but I always get popsovy the same voice, I've already tried everything, all the tracks are obtained with the same vocal and no matter what style of music it is. Similar situation was last weeks with version 3.5.
Please tell me if there will be updates of new versions of your neural network because to make quality tracks has become impossible. To make one good track now you need to send 20-30 tracks to the trash.
I really hope that you will soon make it possible to train my voice and I will be able to make tracks in my performance.

Thanks

[distant sigil]

Don’t like this feature especially if people can’t listen to the songs that’s there. I vote no for that one 👎

[trail wharf]

song mixing

[surreal elk]

If possible, I wish I could see who these other people are, in order to check their content as well.. Thanks ~

[stiff trout]

Can you maybe add a feature that shows what is the problem with explicit lyrics? I mean like which words are not okay cuz its a struggle to find them sometimes

[untold flax]

can you explain with more words that suggestion? are you talking about a mashup with two songs? or a remix with only one? actually you can made "remixes" with the cover tool 😉

[pulsar estuary]

that's the point... i don't want the botter to listen to my songs...

if you're botting, you're cheating and have malicious intents

nty

[distant sigil]

I thought you meant other Suno creators i hadn’t thought of the botters

[untold flax]

It's totally random and meaningless. Sometimes a word is inappropriate and the next time it's approved. And there are words as random as "god" that can be accepted or rejected. Also, if a phrase sounds like a famous song, it can be rejected as inappropriate... or accepted. It's totally random and meaningless.

[pulsar estuary]

i mean the block feature would easily be bypassed anyway, but a block is a block

[untold flax]

In the more than 4,500 songs I have generated, I have had all kinds of setbacks and problems, which are solved with tricks, not with AI tools. I think this is something that should change. It would be good if something were put in place to mark content for adults, and that this content was only accessible to users of legal age (paid subscribers, for example, having a payment card is a way to verify that you are of legal age).

But it detects restricted content mostly in English... in other languages ​​it is more relaxed. Or it doesn't check at all. 😇

[arctic python]

Hi, There are many issues that have eaten up literally hundreds of my credits.
-Songs cutting off at the end despite very clear proper prompts to fade out and end a song
-With the crop song feature can there be an update where you can crop a section within the song, as in when having to extend a song because it cut off without completing it (which invariably gives up to 2 mins just to finish the song off which is too long) and you end up with a very long song with additional bits in the mid section that are not needed and it would be great if we could crop the middle bits not just the beginning or just the end of the song. In my case I have the ending I wanted now using an extend but it gave me all this superfluous music to make up a full 2 minute extend, I wanted like 30 seconds, and using prompts to achieve this does not work, it just ignored the prompt. I don't want a 5 minute 20 song just 4 mins is long enough

[pulsar estuary]

from my experience, AI is reading [end] but if the waveform is too thick (not thin enough), AI won't end it

solution is to extend it w/ [end] and then hoping there's a thinner waveform to crop

if your waveform at the end looks like this:

it's harder to end
basically consume credits to trim until it can be ended and then end up with a Full Song that's distorted by a bug rn

[delicate cedar]

I'd like to propose making Personas shareable by invite only. So if I had a persona and wanted to share it with my friend Joe, I can share it and Joe can use it but no one else can. That way my Personas can remain exclusive to me and anyone I trust enough to share it with.

[fallen kernel]

I have been getting a lot of background static and distortion lately in my music. Everything I've made today is unusable. Can someone help me with returning my credits?

[hidden ibex]

I'd like to be able to buy songs from others.

[mossy cipher]

I've found that sometimes if you "remaster" the song, it can clean up a lot of that distortion. But it's kinda hit-and-miss, so don't take my word as a guarantee.

[languid ore]

Do the Playlists have to be public to download them? All my songs are private

[languid ore]

I need to download in bulk the 2432 liked songs I generated in wav format. It's going to take forever!!!!

Should I keep hacking away by downloading 1 by 1 or just wait until suno adds this feature. The word "hack" is a verb that means to cut, shape, or finish something roughly with a tool like an axe.

[crystal timber]

woud be awesome if there is a stem feature for multiple tracks and not just two tracks (vocals, music). Stem should also be in .Wav forrmat.

[grim geyser]

could train your model on the basics of a synthesizer, so it can understand what a moog ladder filter, or even a digital filter, such as ones found on any free synthesizer plugin does to the sound, in band pass high pass and lowpass forms, so that when it hears it happening in a track it is making a cover for, it can accurately reproduce it, it suffers greatly wile trying to do filter sweeps in general, whether it's on a synthetic sound or otherwise.

[wind solar]

Hello Suno Representative!
I am currently using the Premier Plan.
Since this morning, strangely, the V4 version's sound quality has significantly deteriorated, and the music duration has also shortened.
I generated music using the same settings for genre and lyrics as the original music.
Please listen and compare - there is a huge difference.
Original music:
https://suno.com/song/3b4bfbc9-7a68-43a1-a640-538920d398eb
Music I generated:
https://suno.com/song/4bd7e249-4c95-44ea-a53f-5ac3b34b6871
Additionally, the music playback time has decreased from 3 minutes - 3 minutes 30 seconds to 1 minute 30 seconds - 2 minutes.
What is the reason for this? How can I restore it back to normal?
Please provide a response.

[haughty dome]

Thank you very much. Worked well. Workspaces would be good too, but what did you have in mind? If it just downloaded the pinned posts that would be good. I'm in the habit of downloading the WAV file, but I'm guessing you can't do that?

[median flame]

"Persona"

[celest aurora]

Not sure I can downloading all the clips in a zip for workspaces.

Pinned ones, it's very doable. wav too

But all the clips on a workspace, I don't think I can, because the API endpoint is protected

[indigo mauve]

One thing that i think it would be rad its just like the frame to frame feature in visual motion ai videos.

Image that we can take two diff musics like a metal song and a eletronic song and tell the ai to join this music and creating a transition between them?

we could mix so many genres of songs

Also suno should have a feature that let us generate backwards, like sometimes i just generate a song that its rad but has no intro, its like it created the song in the middle of it, so if we could generate backwards i could pick up so many projects left to dust..

[hidden ibex]

I'd like to be able to make my playlists into records somehow

[marsh lichen]

You can sort of merge songs with Replace Section if you prepare a file with the end of one song and the beginning of another song and put a silence between them. Then you ask to replace the silence. I know, too time consuming but it gives interesting results.

[marsh lichen]

The same procedure. Replace Section. Download a song, put a silence in the beginning, ask Suno to replace the silence. Or you just can ask to replace the first 10 seconds if the length is not the same. It can generate a longer intro.

[languid ore]

How to download Playlist in wav please

@celest aurora I was able to get mp3s not waves

[languid ore]

By waves I mean full size .wav files, not compressed, is this possible, please

[small bobcat]

I really hope an official API is in the works

[rare obsidian]

Beware: there is NO official Suno API

🚨 Suno does not have a public API. Any claims otherwise are unauthorized.

[somber dove]

What if you get a credit or two for each like you get on a song

[smoky apex]

I would like to make an instrumental song that I end up loving and then would like to create lyrics in Suno and update my original creation

Adding BPM to songs would be great

[distant sigil]

I have a lot of instrumentals in various styles that you can use for inspiration of want https://suno.com/playlist/9f107a32-3aea-427b-a4c7-f68207eec439

[short terrace]

I'm new to suno pro. I have been a pro user for nearly 2 months because I wanted to try V4. I say it has some issues but I like it. It sounds very human like, However for a suggestion to make it better. One of the things I noticed about Suno's Persona it only goes up to 3:20. That could use an update to 4 minutes. Its just a suggestion. Before v4 possibly leaves beta mode for free users, I think it this has potential to have a 6,000 character limit and maybe make it 6 minutes. This is a suggestion how it could be better

[indigo mauve]

I did not thought of that and bless you for that

[marsh lichen]

You're welcome.

[narrow rune]

I love Suno. This is probably said before, but an "empty trash" function that lets you delete everything in your trash would be nice. Thank you. This program/service is awesome.

[royal venture]

whats with this

[steady river]

you gotta click the seagull

[tidal star]

I think its the cat

[granite smelt]

One suggestion: How come an American company can't generate the lyrics to the patriotic song "America the Beautiful?" To patriots like myself, this seems like a knife to the back, and it's quite frankly very unAmerican. Y'all better be ashamed of yourselves.

[teal knot]

Hello I think I have found a new limitation with suno generations. It can not do a trumpet bubble mute or plunger mute, in a way that it makes almost sounds like a growl. Instead it makes a continous tone, rather than a fluttering one.
I have litterally uploaded a sample of this and tried to have suno recreate it either through cover or even remaster. Neither can do it.

[haughty dome]

When in Create, it's a bit fiddly to scroll down to get to the 'Create a Workspace' button, and even then it pops out in a new window. So, can the create new folder button be put higher up to save scrolling?

[candid knot]

Guys, if you add more letters in the style of music bar there's going to be magic happening. Please next update or V5, i don't mind. you have to add at least 500 letters! I BEG U. I CAN FEEL THE CONTROL OF EVERY INSTRUMENT. IT'S CALLING OUT TO ME, U NEED TO ADD THIS.

[latent egret]

I am trying to make some ambient space music that has the quality of composer Hans Zimmer. Can anyone give me any suggestions? When I try a this prompt: Composed by Hans Zimmer, inspired space exploration, Ambient, Atmospheric, Peaceful, Meditative, Relaxing, Laid-Back, Ethereal, Natural, dark, Instrumental, Cinematic, Soundtrack, 50BPM

[zenith needle]

WHY hasnt the developers came up with a way to get the same beat from a song with no lyrics.

[shadow phoenix]

it would be cool if suno had a medley option that let you meld your songs together

[pulsar estuary]

to prevent other ppl getting the same generation as you i assume

unless you're talking about remastering

[ocean bolt]

Hey guys, this might have already been suggested before, but could you allow the download of the tracks with the best quality if those have been created during the premium membership, thank you?

[runic pier]

need longer max upload times. 3 minutes would be better than 2.
and does anyone know the reason speech in uploaded audio disallows Gens from public visibility? That seems pointless given there are always gonna be ways to get around language restrictions.

[runic pier]

gotta always think about bots and abuse. plus free credits are really never on the cards

[sullen gate]

That would be a good perk for them to add to the premier Subscription.

[hard perch]

I can't afford the current plan. You should add a $5 plan with half the credits of the $10 plan. This would help increase your user base for the premium plan. I want to use your service, but $10 is too much for me right now. I don't need that many credits at the moment, but I would love a $5 option. Maybe in the future, I’ll subscribe to the $10 plan, but for now, I prefer the $5 plan."

[runic pier]

That box is not like a check list that the AI goes through. More characters wouldn't help you. But it doesn't matter anyway because you can use "{instructions: Arpeggiated chords, solo male vocals. {Style: warm, fun, upbeat},]" for extra tag space at the top of the lyric box anyway. <These> brackets might work better as [these] are sometimes taken as a suggestion, not an instruction. {Curly} ones are used for marking separated items to be treated as one (or as connected.) There is already more control than simply adding more characters would give anyone given that you can do these prompts at any point throughout the song.

[sly delta]

Requres downloading and mastering. There are 1-off record pressing operations if you truly want vinyl. You would just download the Wav files and then send to one of these...

[sly delta]

I keep telling them that we need 5-6 minutes uploads. (But... They won't do this until the court case is settled since it introduces more risk for them...)

The public visibility is because any vocal is probably either a stem from a copyrighted track or something you are working on privately for yourself.

[runic pier]

Does suno not want you to upload your own previously written songs, or is that just collateral because of possible copyright disputes?
Uploading audio is the only way to tell the AI how to sing the melody, it's kind of what I thought the point of uploading was, until learned of the limitation.

[sly delta]

You HAVE to swear with an ESIGN that you own ALL THE RIGHTS to the recording and the lyrics. It's a legal issue. I'm 90% certain that the current limitation is legal. They let it lap and I uploaded a 4.5min clip about two months ago but then they clamped it again.

[runic pier]

damn.. lucky.😅 Doesn't look like there is much hope for a quick solution to my woes then. thanks for the reply 🙂

[honest bridge]

Suggestion: Allow ability to filter out song extensions and clips to only see the full songs

[humble grail]

Category: Most playlisted songs on Home
Aaaaannnd: Badges!

[fallen kernel]

THERE NEEDS TO BE A FUNCTION OF EXTEND THAT MAKES ENDINGS. So many of my songs cut off at the end. SO many extentions just extend for exactly 2:00 and cut off. We need to be able to make ENDING EXTENDS. Thank you for coming to my Ted Talk.

[steady river]

Especially for EDM which is so hard to end!

[steady river]

1 off record pressing (cutting) doesn't sound the same as a pressed vinyl though. So the only way to really get your music on Vinyl is to get a small 1000 record pressing 🙂

[patent glen]

A dedicated channel for communication of possible/ongoing issues, such as potential server lag/gen issues as well as a proper way to report them. It's annoying being constantly left in the dark when issues arise being a paying user.

I love all the new features/website updates especially the comments feature. But when things are constantly breaking such as persona's, being forced to open a new tab/reload the page for gens to appear it becomes quite annoying.

Thank you!

[stray birch]

BUMP

Honestly I would love a roadmap of things that are being addressed, or need to be addressed. The lack of communication is frustrating to no end, and it feels like a lot of stuff are being overlooked or not worked on (even if they are), because we're always left in the dark

[analog zephyr]

I’ve been using your software for a while. It would be nice to have a random music player. And a playlist set up so I can use my music in my twitch or YouTube channel so I don’t have to keep loading the songs.

[native grove]

Remastering Extend Parts, and allowing them to be part of the whole song.

[humble grail]

You can create a radio out of one song.

[vital pebble]

you can also make playlists in the suno app itself!

[humble grail]

.
.
.
Create > Song Radio

[humble grail]

I think he means like in soundcloud.
He can also go to the homepage > Trending Playlists

[crisp ember]

Search boxes, meanig at least Search and Library need some wildcards / advanced search characters.

Especially while non alphabetical chars are completely ignored atm....

There should be at least the following wirldcars:

? = alpha
(#) = number (the number sign...)
(*) = multiple chars (the mult sign)
& = AND
| = OR
() or "" = phrase

[fiery wave]

Allow user to remove certain notifications

[hidden ibex]

In the Replace Section page I'd like to be able to highlight Lyrics on the Right Lyrics panel and see where that'd be on the waveform pattern.

[jaunty saddle]

I would suggest making the feature of changing playlist cover art a free feature. Currently the free to play users find themselves deleting a playlist and then recreating it with the new cover art. This would add a lot of bandwith as is.

I would also recommend making the cut feature free for another reason, a ton of songs are created but only about ten-twenty percent of them get made public. Because a song might finish in one and a half minute, but then turn into another song for a total of four minutes. This can't be shared. It turns a great song into something subpar.

On the free side, it would also add incentive to pay, as if you cut something, you're more likely to extend a cut song, leading to you spending your points far faster. This can in some contexts drive people to paying so that they can continue hammering on their experiment.

More quality songs also means more things to lure new people in. With these features in place, I'm sure most of them will stay.

[versed quiver]

it would be really nice if there was a link to our workspaces like there is a link to our library on the left side bar of the desktop webrowser version of the site. right now, in order to find my workspaces, i can only find the workspace by starting a new song and then using the back arrow by the workspace name i'm in by default to get to the list of workspaces and pick the one i want to actually be in. which is fine and all if i want to create - but when i just want to access the songs IN a workspace that i previously created, that's an awful lot of hoops to have to go through

[pulsar estuary]

add a block function so i can block users from following

it's like they're having a midlife crisis just discovering scripting

most likely the same person with alt accounts actually

[amber violet]

oh! I thought it was me -- I felt like I was missing the workspace link on the menu.

[versed quiver]

nope. dunno if it's there on mobile or not, i do't create with my phone, i just use my playlists as a radio with my phone. but it's not anywhere on desktop, at all, unless you create a new song. it's not even there when you edit an existing song.

makes it very hard to actually use the workspaces

[keen burrow]

• I would pay extra for a feature where I could use my own voice clone as a 'persona' for future song generations

[versed quiver]

yeah

[granite smelt]

Suno AI, is refusing to generate songs like The Star-Spangled Banner, America the Beautiful, and God Save the King!

How is banning America’s own national anthem not completely un-American? These songs are part of our history, culture, and national identity. What possible justification is there for censoring patriotic music?

[stray birch]

THIS

[scarlet socket]

Could you make it so that the artist's blocking is based on context and not just the word? You can't make songs about every character in the Mortal Kombat franchise due to this. An example is Goro.

[jaunty saddle]

Copyright. I would assume. I don't think it's censorship.

[teal knot]

It can if you use Cover.

[hollow locust]

Make Suno not screwing up Duets a focal point for v4.5

[split magnet]

"CREATE SONG COVER", "CREATE SONG VIDEO", "CREATE/RANDOMIZE INSTRUMENT", "CREATE/RANDOMIZE/CHANGE STYLE", "CREATE/RANDOMIZE EMOTION", "CREATE/RANDOMIZE/CHANGE VOCAL", "CREATE FX", "CREATE AMBIENCE", "CREATE SPEECH","ADD INTRO", "ADD OUTRO", "ADD CHOIR".... competitors might do these.. Also, "AI RATED SONG" and listening or liking a good AI RATED song should cost micro credits to listener so that "CREATORS CAN EARN". perhaps 1 to 10 Credit based on rating. Pro's do not have to pay. Highly Rated songs can be published on your Live Stream Channels including SUNO Youtube. Also, If i create a song, the network should have access to my songs with API and I should make money from usage. Also, imagine if we can talk to SUNO Song Agent, the agents collects instructions and builds the song, music, or other sound elements. While playing, the user can still talk and adjust every element in the timeline. Like realtime DJ. The agent can listen to user to get sound clues and melodies or emotional clues etc. The user can say 'make it more ....(realtime choices)'.. call me if you want.

[little relic]

Would it be possible to have an editing option where we can remove segments from the song?

I know we have replace / crop, but it would be good to remove parts entirely

[fallen kernel]

I reached out to customer service about this over a week ago and received no response

[jolly inlet]

I still wish there was feature to regen a song just slightly different, like a variation system

[tidal isle]

Why I have to waste 400 till 700 Credits just to get an end for a song. It would be so easy to program it that [End] gives a end.

[karmic quarry]

No it doesn't. Wasted thousands of credits for this. Suno just adds extra nonsense lyrics if you are trying to Cover for a literal instrumental. It sucks at that part. Dunno how the devs will fix this

[karmic quarry]

This should be a thing. Hope the new version considers this. Would make music production more efficient for us poor and at-home artists covering/enhancing our own originals without relying on that broken EXTEND feature. Thanks for listening Team

[raven portal]

I thought it would be cool to have a part of the site where you could leave some saved texts, such as "Lyrics", "Tags" or some type of note.

[wintry socket]

Once again I am asking for an ENDING function. At least give us the option of selecting a portion for fade out. Otherwise, the Extend feature keeps producing unending pieces that just gobble up our credits as we keep hoping for the one to finally cone up that we can crop. As I have said before, I love Suno, but I am not recommending anyone pay for membership until this change happens because it's a needless drain on the credits we pay for.

[warm sand]

i am using the premier package and it is very buggy, the sound is broken, the song structure is terrible and absolutely cannot make any song longer than 2 minutes 59 seconds, i thought it was confused with the old model and sent an email but almost a week has passed and i still have not received any response. really disappointed with the premium user package

[sly delta]

Try v3.5 for a while it is more unredictable than v4.0. Audio quality suffers but you can fix that with v4.0

[inland bolt]

Can we please get a reduced cost for 'REPLACE SECTION?" I've had to replace sections of song where the AI bugged out and inserted vocals that weren't written into the lyrics and even though it's to fix a mistake that the AI made generating the song and not just because I wanted a different sound. Regardless, when we get charged for each attempt as if we generated a full song, for only 10 seconds worth of time? It's crazy. And it makes the feature almost unusable. Make it so that the full credit is taken when we accept the revision and create the full song from it, or limit the number of replacements you can try in one go before getting charged? Please, be reasonable.

I also second the request for the end function with a fade out option when we extend. This would be really helpful.

[sly delta]

Not likely to hapen since the matching is more compultationally intensive. I tried for months on this even arguing with support.

[sly delta]

Use the lyric tag [fade out]

[inland bolt]

Thanks for the info, I'll give that a try.

[teal knot]

A lot of times it does the opposite, where it ends when you want it to continue longer. Its annoying, there should be a toggle, or a condition, if it fails regarding the continuation or ending in an extention, you shouldn't get charged.

[terse nymph]

It's funny when I tried Suno for free it seemed to work and I thought hey I'm going to give them my money I really like this program. I then found out once you pay you only get part of songs, out of 50 songs I might get one or two that complete. Riffusion seems to be the best free solution.

[haughty dome]

putting this code after the lyrics closes out songs most of the time for me: [fade to end][end]

[autumn basin]

it would be great if V4 follows the singing style, melody, singer...well song at all....V4 is awesome in making voices but creating a new song every damn verse!!!!

[autumn basin]

im 500 credits in and need to take the one with a little error in it....you need to fix this!

[wintry socket]

Sadly, I don't find that it works nearly often enough. I've used lots of different prompts to get it to end where I'd like. Nothing works consistently for me.

[fiery wave]

Option to hide original piece that was extended so it won't show up under title

[terse nymph]

I really HATE Suno

[craggy vessel]

Add search 🔍 tab in Library idiot developers 🤬🤬🤬🤬🤬🤬

Rettaaard Developer 😤

[terse nymph]

Riffusion is free and it at least completes my songs

[craggy vessel]

Work for real problems rather then dropong stupid features like (You can comment now) fck offf.

Suno team not sorted previous issues yet and they focusing on nee craps shtup

[terse nymph]

https://www.riffusion.com

If I cancel my subscription today do I lose my credits and my song library? Not that the credits mean anything if it wont complete any songs.

[rare obsidian]

✅ Warning logged for singhires. I couldn't DM them. || That kind of language is not acceptable. Please be respectful. @craggy vessel

[terse nymph]

I just made 6 songs on Riffusion and all six completed, I made about 20 on Suno and not one completed or is close to usable.

[tepid trench]

You mentioned that it was fine before subscribing. Are you using the cover/persona features? If so, these tend to cut off around 3:20 ish. You can continue the song using the "extend" function if it cuts off.

[terse nymph]

@tepid trench no using create from scratch. I've burned through so many credits and nothing is usable. I've tried all kinds of prompts. Nothing works

[tepid trench]

Hmm... How long are the songs coming out at?

[terse nymph]

00:01:30 to around 00:02:30 I make the exact same song or Riffussion and they all are about 00:03:17 and complete

[tepid trench]

Might be an odd question but do any of the song artworks have the word "SUNO" over them since you subscribed, and are you using any kind of VPN? Certain things can flag accounts as using an app/website that isn't the real Suno, so there's a chance that might have incorrectly happened (if so, I can give you a link to a form to fill out but just want to check this info first.)

[terse nymph]

No I'm not using the VPN while on the app, I did get a message a few days back I was using a "Fake App" but I just put the web address in again and it worked

[tepid trench]

Ah, the message you got, did that turn up in the lyrics?

[terse nymph]

yes it did...it sang that I was using a fake app

Don't get me wrong I really did like Suno I like the sound it produces and I really want to keep it.

[tepid trench]

Yep, that'll be what the problem is then. If the site thinks your account is associated with a fake app you'll get all sorts of weird and annoying things like this happening.

Fill this out: https://sunomusic.typeform.com/to/zGsJSzsf

It may take a few days to hear back from Suno but they should put it right.

[terse nymph]

So when I go back and look at my older songs most are at 00:04:00

[tepid trench]

If it started going wrong after you subscribed it's probably something to do with the payment that has triggered this (can be an error in billing info for example.)

[terse nymph]

No some of the songs worked I think it got messed up when it was telling me it was a fake app

[tepid trench]

Ok could be something else (I can only guess based on the kind of things I've heard about happening before 😅). Hopefully it'll get working for you again soon once they process the form.

[teal shore]

if u want songs to end on suno you have to put [Outro] at the end of your lyrics and it'll fade out slowly 👍

[terse nymph]

@teal shore I did

and the song never makes it that far anymore

[tepid trench]

Yeah that's not going to help in this situation

[teal shore]

i made 3 songs, (30 creds total, v4) all faded out with outros, maybe you're using too many styles confusing the ai

[tepid trench]

No if it's giving short songs cutting off around 2 minutes and sometimes making songs about "fake apps" it's an account issue that the team need to address.

[terse nymph]

same promt on each platform: https://suno.com/song/f183768a-6ab2-4687-b57f-fc8091d2e79d

well it won't let me post the other link

[teal shore]

if you only write [Outro], no end no 15s, it'll work. these apps dont share the same tags

[terse nymph]

I did it without the 15s also but I'll try agian

[teal shore]

its like 3/4 chance, and the duration depends on the structure of your lyrics that you can extend further by adding instrumental parts to the song

[tepid trench]

It's likely that you'll continue getting short songs (and/or other symptoms) until the form has been processed, regardless of prompting.

[terse nymph]

most of the time it was ending in the middle of the song so I added the times to help along

[teal shore]

i make dnb and house so most songs are between 2-3 mins

[terse nymph]

yeah I filled out the form so we will see what happens

[ionic harness]

SUGGESTION: Add a block button on the "..." profile menu

[terse nymph]

See it ends in the middle the second one went a littel furthur https://suno.com/song/2781d79a-408a-4131-97f3-abdc6d426d0e

See I was getting complete songs in the begining https://suno.com/song/4f04b19d-e1a8-4245-9fa5-9a1d53f680c8

I think maybe I did get flagged I'll see what happens after they recieve the form

[teal shore]

i think if you put something like [Vinyl Scratching] after the chorus it would be closer to 4mins

[fiery wave]

Wiuh.. Take a chill 💊 & in a corner

[fiery wave]

I've had many 🎶 s to end as they seem el nice Maybe a temporary glitch in your case.

[teal shore]

i'm sure if you cancel your sub it just doesn't reoccur

[austere night]

will they ever fix it so the cover and persona function doesn't stop at the same time?

[hidden ibex]

You should streamline Rocksmith song add ons
https://www.ubisoft.com/en-au/game/rocksmith/plus/rocksmith-workshop/tutorials/additional-features

Cause it'd be neat

[distant sigil]

On the Suno profile page can you make the bio/about section a little longer? I would like to see that if possible.

[median wave]

good idea! though I would like to download each individual instrument to its own track, so it can be mixed in a DAW.

would like to see the suno player, optionally display lyrics, rather than having to download the mp4

[simple sonnet]

#💬┃general-chat message

I wish I could buy things with my Suno credits other than Song generations maybe like avatar badges or dumb things like flowers or GIF or just giveaways of credits to people who have really good songs and you want to give them credit more than just a thumbs up 👍

[sturdy glen]

When you upload an audio and you end up with a long title, browser spits out 500 error and vague "error" UI display when title is too long and a remaster is asked (because remaster appends "(Remastered)" at the end and therefore cannot do a new track because the title is too long). This likely also happens with Cover. Error message should be more specific for the end user to know what to do (manually edit the title) rather than random error occurred message. Do the devs even check this chat? I've posted about a bunch of bugs and never got a reply

When trying to upload FLAC, file upload "starts" but hangs forever and never finishes (site likely doesn't accept FLAC at all). No error is ever displayed and user waits forever not knowing what happened. Terrible UI (for this I've learned to convert to wav and upload that instead). Maybe tell the user FLACs aren't accepted? Or simply code flac to wav ffmpeg on your backend?

[meager igloo]

Can you add an option to download subtitles? Manually configuring subtitles is very troublesome for me now

[indigo ginkgo]

Could we get a specific model to end songs... something trained to end a song.

[teal shore]

A donation system wouldn't suffice, where we can donate 50 points (per day/month) to each other. Use it to boost inspiration of users by giving each other credits to improve our sounds, help each other out, for collaborations when one person has credits and the other doesn't instead. Could work 👍

[sly delta]

Use [end] or [fade out] or [finale] in the lyric section with an extend.

[sly delta]

I don't know of a way to share credit for a track bounced around. That's really more of a distributor thing.

[indigo ginkgo]

It only works in v3 and that degrades the quality of the entire song. (and even there not always). I think having a model trained at ending songs would help a 1000.

[sly delta]

Works in v3.5

I don't know about v4.0 - I do mu creating in v4.0 and then I Remaster. But I never Generate or Extend with v4.0 to avoid the Shimmer and also because v4.0 is more clicheic for pop styles.

[vague dawn]

Suggestion : langage filter in searching list menu , if I look for UK drill I don't want to hear something russian, German, Japanese ,etc...)

[arctic python]

YES! Also when doing this just to add the outro that Suno had cut off and using specific prompts it will still often just start adding other parts/verses/chorus from the original song when you only need the last two lines and fade out grrr I have wasted hundreds of credits just trying to finish the song as I wrote it.
Because you can't crop an extend before adding it, and you might need to crop the front part out first.

[graceful mica]

Hi, I just wanted to ask if there is a way to mass download all creations? I have quite a lot and I would like to organize them.

[sly delta]

No that would load their server. You need to develop a workflow where downloading and cataloging is part of the process.

[deep monolith]

I noticed Suno now has a new sweet interface for replacing sections and the waveform displayed looks really gorgeous. I was wondering if you tech guys could put the waveform displayed thinner but as it looks at the bottom of my songs that I play on a regular instead of the plain simple white line that we now have. It would help a lot in giving us feedback on information on where we are going with a song. It would look super gorgeous even if it was just an option setting; the plain white line or the waveform.

[deep monolith]

I am a super producer, I made close to 2000 or more songs using about 7 different DAWs over the years. I have been producing for over 25 years. Now thanks to Suno I can produce even faster and more accurately. I know a lot about interfaces and workflow. I think a waveform display would be really nice.

[distant sigil]

Will there be any features that the basic users can enjoy? It’s not fun being left out if everyone else is having fun using the v4 features except the extend feature

[sterile stream]

Dreaming of the day when we'll have a collab tool!

[median wave]

good idea. I usually indicate [end] in the lyrics, but it doesn't always work out - what I have to do is Extend the song, then crop it at a silence. Speaking of that...there should be a configurable silence at the beginning and end of the audio. One common need to end songs is to [fade out] but I don't think Suno knows how to do that...but it's sorely needed (otherwise I have to post-process the audio to perform the fade and add silences.

[deep monolith]

I agree. The ending of a song is very important. I spend a lot of time fixing outros.

[lament gale]

is it still not possible to filter the library by workspace? or filter out songs in workspaces?

[runic pier]

THANK YOU FOR FADE OUT!! we asked and we received! thank you! (now just sort out other types of ending and we're really gonna be in business!)

[median wave]

been playing with the Persona feature because I realized that listeners look for music by genre and then by artist. That's what a persona is. People don't want to hear a collection of random and different sounds, they become familiar with a particular sound (voice and genre) and begin to follow that persona. Consequently, I re-rendered a set of my songs into one persona so I have the beginnings of an "album". That's the next abstraction Suno has to make "genre/persona/album".

[median wave]

how do you do fade out? I wanted it but don't see how to do it.

[deep monolith]

Click on the Edit button and it’s at the top to the left beside Crop

[runic pier]

only just been added as an option in browser. probably in the "crop" section (or new equivalent since it has been all put together in to one interface by the sounds of it)

[deep monolith]

I noticed when I was using the crop feature that it was only designed to keep a portion of a song. We will need to cut out and dispose of sections in the middle and not just the left or right also and have Suno flawlessly merge or splice the remaining portions.

[median wave]

right. deleting a mid section that you don't want isn't possible, leaving you with the task of reverting to a prior generation that didn't have the "erroneous" piece.

[deep monolith]

Did anyone else notice there is an advanced editor bottom at the top left corner in the bottom of the drop down menu, I can’t wait to see what creative feature Suno unleashes.

I am working on this little song titled “Turn It Around “ about turning negative events into positive ones and I find myself taking the song into the edit mode just to visually see the waveform. It’s much more visually enjoyable and informative. I will be using this new feature constantly. Whether I am editing or not seeing the waveform is very comforting and fun.

[unique surge]

I'm dumb and didn't realize that the crop tool automatically saves the segment. It faded to black and then came back with a zoomed in waveform, so I thought it wanted me to do more, but nope. It already saved it in my library 😂

maybe have it say "Cropped clip saved!" after that fade-in

[marsh lichen]

New editing interface looks nice. A couple of suggestions:

1. Please add the possibility to make several crops at a time, in different sections of a song, and I think it would be better if the Crop feature would remove the highlighted section, not keep it. Or add the keep/remove switch.
2. Add Fade In in addition to Fade Out. It would be only natural.

[slow walrus]

I wonder whats inside the coming soon advanced mode

Sounds promising

[median wave]

Just tried the new editor...fade out - perfect! changed settings to "keep duration" so I don't get 2 minute replacements of 30 seconds. OK. Also, Is it cleaning up sections you don't confirm? not sure. If so, that's great.

[deep monolith]

Mastering hopefully. Mastering would be super cool.

[unique surge]

ability to expand the Style prompt box like the regular prompt box

[median wave]

Agree, but I think the waveform would be more informative if it plotted the vocal and other instruments with distinct colors. I know most waveform displays are aggregated, but I don't see a reason to keep doing that.

[marsh lichen]

Please add the option to create only one track - at least for replacements, extensions, and remasters. They are quite often literally two copies of the same.

[slow walrus]

I want edit functions

copy, paste, cut

[runic pier]

OK, I just tried the Fade feature. You did the minimum with it. Fine, it's still good news but..
it's only a volume slider, so why don't we have the ability to affect the end point of the fade?

Here's how you change it:

1. Red side of selection indicating the loudest (or least affected) part of fade,
2. Blue side of selection indicating most affected. you can move both siders to silence any point in the song you want.
3. You can glide them over each other to invert the effect and Increase volume as well. (perhaps by a set amount over the length of the selection, perhaps by a custom amount)
4. You can add more than one selection on to the playback bar at one time, allowing for silent parts within a song.

(obviously going too loud would cause clipping issues but that's at the user's discretion how they apply that)

[indigo ginkgo]

lol, quickest fade out option dropped, haha. Still would love an ending model, haha

[slow walrus]

What I want is:

• copy, paste, cut (ideally from one song to the next possible as well), and AI just involved in the transitions - not the singing like replace does ..
• Possibility to select more than 1 persona OR possibility to change persona on an Extend - because currently I have to decide between my male or female singer and one of them is then a random new voice. I'd like to keep them both
• gender tags in lyrics should be commands, not suggestions put on a dice. its the most frustrating one and burns tons of credits if it constantly does the wrong gender in verse sometimes
• I want to be able to pick any extend and be able to append it on any song on any timestamp. This would help soooo much with suno's remembering of the song (max 2minutes back) and with the quality. e.g. if you have a 5minute song and extend from timestamp 2:30 you will get the quality from 2:30. If I now could apply that extend at the end of the song that would be AMAZING

[deep monolith]

I just tried the fade feature, it’s simple and awesome. Reminds me of the one in Cubase or FL. Studio.

[deep monolith]

I put my gender tags in the style box not the lyrics box. (Male vocal with backing vocals) most times.

[static coral]

It would be amazing to have the ability to direct the melody trough the tags. Like you writing [Pa, pa, tatadatadada!] and next riff trying to reproduce that sound. If such thing would be possible, it would give us much greater control of our songs, allowing them to sound just like we want them to.

Also having the ability to use more than one persona in song would be cool.

[deep monolith]

Yes it’s possible

[nova dawn]

My ui don't have any new features yet.

[slow walrus]

that works when you use only 1 singer the whole song

But I need my female singer at specific verses

[runic pier]

with voice tags, put them with <these brackets>

[sly delta]

I'd love to have an internal crop tool to fix these problems

[slow walrus]

Maybe its possible in the advanced mode. Would be nice

[sly delta]

If Suno would let me reupload the song I'd fix it in the DAW but uploading 5mins present "legal issues"

[slow walrus]

I mean
[female voice]
works, but as always with tags: Only if Suno likes to fill your wish

[runic pier]

{<female singer, powerful voice>
blah blah blah,
this is my song bit.}

that might work.

[deep monolith]

Yes I use one voice for the whole song. If you tag it in the lyrics box you’ll get to have it switch between two voices for example a singer and a rapper. You can even specify the Rap verse or chorus.

[runic pier]

don't use square brackets, and definitely don't use square brackets if there is text on the line above, that makes them less reliable.

[slow walrus]

hm

[deep monolith]

Get Audacity it’s a free Microsoft Editor. You’ll love it.

[slow walrus]

I do edit all the time external, but people would like to be able to have some songs already in good state on suno 😄

[indigo ginkgo]

If you do a crop, you'll see them...

[rocky falcon]

Hi, I restart web and late restart pc and a message continue saying "A system error ocurrred" anyone know what happend?

wait, wrong chat sorry.

[deep monolith]

Yes it’s there hit the “Edit “ button to see it…

[sand granite]

Ok lots of people lots of questions. I signed up...I'm burning through coins... my songs aren't completing, they are ending early.. and I don't understand why... or I make a short version to test... and then it doubles it up...

[deep monolith]

Keep an eye on the amount of text letters you use in the lyrics box at the bottom it say 0/ 3000 to guide you. Limit yourself and Suno will do the rest.

[sand granite]

way under 3k. It counts me between 1500 and 2k

So I can do a song.. full song and it never completes. I do a piece of the song, and it doubles up that piece,.... starts over, very strange. annoying obviously

[twin quail]

Would be definitely helpful or maybe possibility uploading it to external storage like Google Drive or Dropbox or FTP

[sly delta]

Extend is part of the workflow depending on genre some tend to end early. Tou should only expect about 5-30 seconds of good audio each extension. When you extend you need to tell it where you are in the lyrics.

[deep monolith]

Break down the structure of your song into blocks. From Intro to Outro.

[sly delta]

Sonauto can reference external audio links via their API. But they are the only one and they only gen 1:32 songs.

[sly delta]

Really important to use structure tags: [intro][verse][chorus] Also use vertical space and punctuation. There's more on this in Suno's official guide to prompting.

[sand granite]

what does that mean.. pretend I went to public school

verticle punctuation

wait a guide?

I've been guessing

[sinful merlin]

I haven't checked into the new feature yet and I hope I'm putting this in the right spot....but I had this idea fur Suno a while that'll be even more helpful when editing songs.

Where we can add different prompts like bubbles that have separate lyrics where can select if it's a intro, verse (as many and as long as needed), chorus (any many needed and maybe enough fur 1,000-1,500 letters/characters), bridge's if wanting, one just fur talking if we want it ta just talk in those part as many as we need, and the Outro scene.

Also have where we can add multiple different persona's to each one where needed. If not the above part at least adding more personas to the music would make it feel as if you are doing a collaboration wit another artist. Weither its on the same edit or when you extend the song.

[sly delta]

How you write lyrics changes what you get from the model greatly.

I spend 10 minutes sometimes between generations changing punctuation and spacing on my lyric before pushing the button again. You can put (insert words) in parenthesis. You can break up a long line to change delivery by adding line breaks. You can change the emotional delivery of a line with punctuation like comma, colon, question, and ellipsis.

For best results use 6-8 blocks of prompts excluding [break][interlude][drop] consider [intro][outro][solo] to count in those sections.

[sly delta]

I've been begging for the multiple persona thing. But really we need to split the persona into rhythm, vocals and instrumentation. Persona as it is is not granular enough.

[sinful merlin]

True. And they don't really quite sound the same as if you used it in different songs.

[deep monolith]

Wow guys look at what just came up. In Edits…. The whole songs waveform is broken down in the basic colored structures from Intro to Outro. It looks beautiful. It wasn’t there before….

[sly delta]

I might be one of the people responsible for personas. I sent support an email in late July explaining how I was getting consistent personas without their feature. That still works... Sometimes better than personas. (especially personas this week...)

If you're interested in how to do personas without personas DM me.

[sly delta]

This means that we can finally have some visibility on which lyric section tags really work.

[median wave]

yes, this is very important for a persona, that it does not change its vocal character. Otherwise the listener get's confused "oh who's singing now? some new person(s)?" Instruments, melodies, rhythms change, but not the voice.

[median wave]

maybe we're all guessing? Suno should get a technical writer/ presenter to explain how you "influence" the sound, because you can't command it, like traditional software...in fact, if you try the exact input many times, you don't get identical output.

[sly delta]

I'm working on it. There are only a couple of us that use Suno enough to know. I'm not sure the devs have found the tricks of us super users.

[arctic python]

I tried the "Exclude" feature due to wasting a vast amount of credits getting a male singer despite clear prompts asking for female only and all the way through
And rather than "Exclude" it "included"male singers 😭
Also getting a metal growl male vocal for some words/lines despite the entire prompt being in an electronica type of theme with a female vocal, apparently using words like "epic" and so on in the prompt will do this? Sometimes I want a hardcore gothic metal female song and it automatically give male vocals. this is annoying to the point of madness at times.

[sinful merlin]

Sure that sounds interesting

[sinful merlin]

Yea i know it's a working progress I still like it. It's not 100% there yet but the more I use it more I'm learning myself. I just found out how ta do different voices in same song but then it starts blending back ta same voice.

Far as persona goes kinda the same but it's still interesting to keep a female and male voice of different pixels in sound. I just can't wait til they update it where we can allow our own voice and save it as a persona. I wouldn't use it ta upload like that but ta know how ta sing better wot real voice.

[slow walrus]

the exclude box is an include box

what you type there is what you will get 😂

[floral pewter]

Cool! I hope that will be able to figure out, still super new at all of this especially, editing! 🤪

[distant sigil]

(1) An indicator of some kind that lets me know who’s unfollowed me? Why bother following me if people are just gonna unfollow later. (2) I guess a block feature that still lets people listen to the songs on the page (3) credit donate feature or a subscription donate to other creators thank you for reading dev team

[ornate osprey]

For the new editor: -1) I prefer old editor as I could exactly mark where to stop in the player. It's not easy to do in this new editor. Gives us the option to use the old editor (or add that old functionality, maybe you don't know which one...). -2) Make easier to found "same duration", in a tutorial or something -3) Make less costly the replace option, 2 credits. Make free cost for the fade option, adding an option to add or not a maximum of 10 seconds. -4) We can't "confirm section" as easy when leaving the editor on the workspace.

[slow walrus]

I can't select "same duration" like earlier

you can

there is an options ⚙️

[frigid timber]

This new editor looks amazing, I am hoping that this is a baseline for a potential web based DAW in the future, I can see alot of resemblances. being able to slice, cut and add in other generations seems like a way to go with this, multiple track lines probably further down the road, for keeping vocals and instrumentals separated. Very nice though! Kudos team!

[median wave]

I find it sometimes ignores that and creates a section of 2:04 minutes (that I can't use).

[leaden marlin]

save draft function for songs you don't finish all at once.

[hollow arch]

I suggest adding the option to move sections and a cutting option instead of replacing. All of these should be updated according to the sections of the placed lyrics. It would also be useful to have a copy section, as the choruses are often not identical.

Example:
Verse1, Bridge, Verse2, Hook, Verse3.
→ Verse1, Hook, Verse2, Bridge, Hook, Verse3.

I really like the new editor, but I’d like it to be more accessible.
I would suggest adding an "Edit" option instead of "Create" – or making it so that when you enter the creation process, you are actually in edit mode. This way, you can create your song directly in edit mode.

Also, there should be a separate section for audio, so that the vocal track is independent of the instrumental. This would make modifications much easier, especially if the instrumental is perfect but the vocal has artifacts or defects. Additionally, every time a replacement is made, the vibe and rhythm change automatically.

Thank you for your hard work!

[lethal pivot]

I believe that you can also write [Chorus 2], to get a slightly difference Chorus..

[worldly sphinx]

Not a suggestion, just a big thank you for the new editor on the browser, I like almost everything about it and finally we have a "fade out" option 🙂

[honest bridge]

Suggestion: The mobile web experience is pretty frustrating, primarily due to accidentally clicking on things because it’s too condensed. I constantly click the buttons beneath the track progress bar when trying to click on a later part of the song (which I do for every song initially, to get a feel for it). It’s just a constant headache, and it makes the software not fun to use. I have lots of larger devices I can use, but most of the time I prefer to use the one that fits in my hand, and the web experience on that one is quite bad.

[worldly sphinx]

this might be a transient problem but shoudn't the "extend" option in the create sub-menu open the new editor interface for extending?

[humble summit]

Looks like a plugin for a DAW

[stone steppe]

oh crap

geez wood get over yourself

ill take that. thanks bro

[humble summit]

But I do hate my round buttons in the browser xD @stone steppe

[stone steppe]

hahahaha

I thought you might!

yeah I hate too

i made it a bit ago. dont know what I was thinking. i was too much into the analog design

[humble summit]

You'd be better off to have a drop down I feel titled "actions" or something

You can stuff it with more stuff later and have a title+icon

[stone steppe]

thats a really good idea. I had thought of something similar with a drop down menu and sorta copying the style they're doing with the edit UI. Its pretty smooth but i feel like the timestamp scrubbing is a bit annoying

i might make a v2 of this

[humble summit]

Well I see why timestamp scrubbing is a thing, you can do it pretty much in any DAW. You could also select any time stamp region and have a menu pop up on the selected regions and indicate there what you wanna do, if you want to move the edit menu over there. This way you can also do multiple edits via making multiple timeline selections (with a button "new slection" or something). This will save credits and give you most edits for your buck

[stone steppe]

oh no totally! 100%. I would be so mad if there wasnt timestamp scrubbing. I kind of meant the way they have the functionality right now, it doesn't focus on the time marker when you zoom in with your WASD keys. It also removes the selection when you accidently click somewhere else. I feel like the responsiveness could be just a little bit more user friendly but thats it

[humble summit]

For that you probably would want a desktop app

[stone steppe]

Suno App

true

[humble summit]

Although browsers are good enough

[stone steppe]

when

[humble summit]

Suno DAW when

Or alternatively

If Suno can have the ability to separate vocals and instruments

You can just import that into a DAW

And do whatever you want there

Stems is not really enough for that (even if they were working lol)

[sly delta]

@humble summit much of what you are asking for is provided outside Suno. I frankly don't trust them with a good DAW workflow. I'm much more in favor of them opening an API and letting someone else figure that out.

[humble summit]

Well as you know it’s not available and against ToS. I am aware that it exists in standard DAWS. I’m just saying

[sly delta]

Those that want a pro workflow and use pro workflows want to integrate Suno into our workflow not replace our workflow with Suno.

[humble summit]

And who is “we” in your context?

You mean musicians who make music themselves I’m assuming

[sly delta]

Pro audio uses of SUno

Those who consider SUno a professional audio tool

Sonauto has a real API now. If their next model is as good as 3.5 I think they could become the backbone of AI music production if Suno does not change their stance on API.

[humble summit]

I think you’re going off topic here as this is a suggestions channel for browser workflow it’s pretty simple really what we were discussing here lol

[sly delta]

hehe yup

Bringing it back. Browser is NOT a DAW and doesn't do a good job a working like one.

[humble summit]

What I was saying is modern browser has enough capabilities to act like one. In fact there are browser based DAWs already

[sly delta]

Yup and they all suck!!!

[humble summit]

Do you not know GarageBand? It can be ran in the browser. Lots of albums were released with it

[sly delta]

I'm a 25yr veteran web programmer and music producer for the last 30ish years.

You can't do what is common in real music production 50ish plugins on 30ish tracks. The browser memory model has problems with that. Even the web audio stuff is only designed for about 10 processors max.

[humble summit]

Yes you can, it’s called cloud

Just like NVIDIA has their cloud for gaming

You can do very similar thing with this

Presuming company has the hardware

Or access to it

It would be an interesting approach to have Suno be basically a streamed DAW

But this far in production I’m not sure how viable such a big architectural change is

[sly delta]

So now you want a dedicated cloud CPU/GPU for each user on the edge? Audio production is highly interactive. You're constantly changing things. The only way it works is local. If you're pushing commands for control ribbons to render in the cloud your'e introducing latency in the workflow...

While WASM Web Audio is getting better as I said the memory model is NOT ideal for the amount of processing you want to do. And if you're doing this locally on device anyway - why not have a local interface that's tied to the 100k plugins already existing?

Adding an API allows that to happen. Better model for Suno. Less effort. More profit. Less chance of user backlash if they change the UI... hehe

You could do GarageBand in the browser but not Ableton, ProTools, Audition or Nuendo... Tht's why SSL and Harrison are building off the open source Ardour for their stuff.

[humble summit]

Well if you really want API access I don’t know what’s the point of this discussion as the answer is pretty simple “NO”

[sly delta]

My point is that a Suno DAW would make the entire things less stable because of the amount of front end design resources necessary to make that work cross browser. And we already have horrid frontend problems right now in ALL UIs

So I'm arguing that although I'm in a DAW every day I think it's a BAD idea to have Suno build a DAW

[humble summit]

Other thing that in THEORY could work is a desktop app connected to Suno that is capable to take Suno output and properly segment it

[sly delta]

That is what I'm advocating - but it's not a browser suggestion

[humble summit]

If made by Suno they would keep their API private

[sly delta]

I'm also suggesting that Suno let others do that rather than their current NIH syndrome.

[humble summit]

Well send them an email xD

[sly delta]

The current lawsuits kind of tie their hands on the API

[sly delta]

I've been dialoguing with them about this for the past 5 months in email.

[humble summit]

And I’m assuming there isn’t a hint of agreement

[sly delta]

Your assumption is incorrect

we basically need the lawsuits to blow over so they can stop being afraid of the big back record labels then we'll get a lot of this.

[humble summit]

We'll see, I am a backend dev so it would be interesting to see what happens

[sly delta]

I'm DMing you privately to continue the convo. There's things I don't want to share publicly...

[gaunt wave]

Image you can use plugins directly on Suno 🤔

[austere crystal]

We should be able to rearrange songs in our playlists! I used to be able to do it in a janky way, but now it seems like you can’t anymore.

[gaunt tinsel]

I think Persona should be able to be set always.
I cannot set Persona on Extend....

[dire iris]

I wish teh Style Of Music box was bigger, I swear its even smaller now lol

[amber violet]

Generate an improved 'instrumental' from a song and create a fancier instrumental version with more distinct layers (drums, bass, guitar, piano, synth, etc) -- and we can generate those parts and reassemble and mix on our own.

[lament mantle]

a SHUFFLE button would be pretty cool, on all platforms for playing of playlists. ... the workspaces thing is cool, howabout SUB-SPACE under a Workspace. so each and multiple songs can be created under One Overall Workspace as the CoreSpace, like windows would stack folders in a briefcase.?
but first, A SHUFFLE BUTTON ..!

[haughty dome]

Make it possible to set up FADE and REPLACE before committing the song to be remade.

[hybrid hatch]

+1

[vocal wyvern]

I would love a Mass Download function for songs you select

MP3 and WAV would be amazing in 1 click

[haughty dome]

When Exploring for styles, and you click "Create With V4" button, the Create window opens. I'm surprised that it doesn't already carry the style over to the Create window.

[wise blade]

Here is a suggestion I have. Get rid of the extend to all feature in favor of collaborative workspaces. Here is how it would work. Nobody can freely extend tracks throughout Suno UNLESS they are added to the collaborative workspace. Once they are added to the collaborative workspace that a particular user sets up they can: extend, replace, remaster and fade. Both users can work much easier that way and it isn't as violating as your track being out there to be passed around and extended as anyone pleases. You as the user would have more control over editing rights being shared or not and who they are shared with. Also all collaborators on a particular track are listed on it just like Newgrounds does with its games and movies. 🙂

[wise blade]

Here is how the Newgrounds page looks.

So on Suno you have the Creator of the track then you have collaborators. Both would be listed right on the page so people could easily click links to collaborators Suno pages.

[opaque gull]

Hi, if a singer is jumping in at the beginning without a musical intro or the song ends abruptly, I suggest two functions (make intro and make outro, with entry for seconds).

[wise blade]

I agree many features are controlled by meta tags that don't always work. The same goes for pitch changing and tempo changing. None of these things can be done with the click of a button. It's throwing things like [no vocal intro] your song lyrics [song instrumental] and finally [instrumental outro] into the lyric section before generating though it is great for outros that they have a fade out now.

I mean think of HTML / PHP. Sure someone can put in <b>word</b> but they also have buttons that will do that for you automatically. So when it comes to SMTL (Suno Meta Tag Language) there should be a button you push that can give you a particular result.

[humble grail]

📌 Pin Comment Feature

[lament mantle]

really though... a shuffle button wouldbe great.

[muted anvil]

Fix the replace section..last time its great blend into instrument but new one replace all then lyrics completely out of hand lyrics..please fix it

[supple current]

if DRM/audio sigs are self-contained within the track, maybe create a thick-client app for windows/mac users to download their entire repertoire with IDs/structure (say in json) and allow offline drag-and-drop re-organising of music into albums/workspaces and/or cover art/lyrics and then sync with cloud/service. a) could be useful for mobile/low-bandwidth users/rural/countries like Australia and its big fat modem link and b) [more interesting imho] to collect some user analytics (such as what DAW, how often used, other valuable info for investors)

[honest bridge]

Just wanted to express my gratitude for having access to Suno, since that didn't really come through in my last suggestion about the browser experience on mobile.

Now that I got that out of the way 😅 , a little more hopefully-useful feedback:

Description: Workspace filtering request
Expected behavior: I expected to be able to filter by Workspace in the Library view (or something like a dropdown selector)
Actual behavior: I'm only seeing a Workspace dropdown on the Create screen (which is great to have in that place)
Additional comments: I think it's good that songs default to the main workspace if one isn't explicitly selected. The thing is, when you want to listen to those songs (only listen, not create) the selector is only available on the create screen - which isn't always the intent. Thanks again for the awesome product.

[supple current]

community competition or dev hackathon on genius ways to overcome the only valid argument i see: how to combat ai slop, both others and self-created. i like gamification, so i liked the 'influencer' suno cred idea based on remixes. perhaps a slop-o-meter would be good idea. it would self-score easily as a start and provoke further engagement to reduce the sloppyness of a track via edits or regens or whatever, effectively turning the tide on signal-to-noise ratio, quality over quantity cough 60,000 followers, etc

[honest bridge]

Description: Intimidating "Trash can" icon in the upper right of the screen
Expected behavior: Heh, well... I expected one of two things: either take me to the current trash folder, or delete the current song (I made sure I was on a song I wasn't in love with just in case while testing).
Actual behavior: It took me the to Trash folder, phew!
Additional comments / context: I think I understand the reasoning, or at least what mine would be: "Do we really want a "Trash" menu in our app?"

The button is just a little scary, for a few reasons:

• The button being round is different from the other rectangular navigation menus, such as Songs, Playlists, History, Personas.
• It's an image only icon instead of words, which contrasts with the other navigation menus.
• It's separated from the other navigation menus, which gives the impression that it isn't one.

Suggestion: Perhaps instead the concept of "Trash", or "throwing a song in the trash" (which implies that you don't like it, as opposed to just not having a need for it), instead use the concept of an "Archive" and the icon can be like a box with a down arrow over it (or whatever the standard is for that).

So maybe present the not-scary "Archive" option at the song level, and then present the concept of the "Trash" once you're within the archive. Then you could tuck the Trash navigation menu within that Archive section.

[severe glen]

SUGGESTION: One-Click Empty Trash Can....

Also, chord detection for video exports to overlay chords for either guitar or piano.. for creating video guides with chord progression overlays..

[wise blade]

One that lets you delete tracks whether they are attached to personas or not. Oh and they should allow you to delete personas you don't want.

[hollow arch]

Nice point and if you write simple Chorus should be the same.

[modern summit]

if we dislike a clip then it directly goes in the trash

[stone steppe]

thanks for basically adding this Suno!

for those who don't know, precision timing for extends now exists within the new editor.

HUGE

[tacit epoch]

Hello. new here. not sure if it's been asked already. The ability to extend a song with a specific persona is not available.
so if i want let's say a long instrumental track and then to extend it with lyrics i cannot choose persona

[tacit epoch]

it really is cumbersome trying to organize workspaces

[tawny relic]

I don't think Suno uses [...] VST
Even if their models directly generate pcm streams, they could still implement that into a neat little VSTI that would let us generate directly into a project inside a DAW

[sly delta]

Open click move to trask. But then two click empty trash like the OS. You don't want to inadvertently delete something you can't get back without a lot of help. Honestly I wouldn't use this now that workspaces are available. I often go back to trash rerolls and pull out interesting sounds

[fair nimbus]

Please remove the lyric moderation when it comes to explicit and sexual content, and instead allow us to add an explicit badge to our song titles if we wish. A “explicit song” toggle on our end that adds a 🅴 badge. The real music world has content of this nature and it is celebrated; don’t limit the creativity of your users. These topics add human texture to a song.

[severe glen]

SUGGESTION: Ever seen Fadr AI? It has MIDI extractor.... we need that here....

being able to export time aligned midi chords, bass and treble

even melody from voice

[marsh lichen]

I also asked for it. The lyric moderation in the music generator is a thing no one needs or wants.

[twilit hedge]

Much like in public image generator... NSFW is prohibited. We can only be creative with that if we run the model locally... I suspect it's the same deal with lyrics generator.s.. they want to be on the safe side. With that said.. some moderation in lyric in Suno sometimes is a bit ridiculous such as common phrases...etc...

[sly delta]

You can get pretty graphic with styles I know. You can also get around lyric moderation by using weird spellings so the whole think is pretty much a joke and annoyance at this point.

However I think they are trying to avoid more law suites right now and that's what all this moderation is really about.

[twilit hedge]

Yea, there's some way around it for sure.. ie.. spelling it phonetically for some words

[sly delta]

You have to do that half the time anyway just to get the thing to pronounce words correctly

[formal bobcat]

Is it possible to change a song made already, using Cover, to just change the style of music, or just change the voice of the singer, without changing the actual tune or words?

[marsh lichen]

Please add the possibility to turn on Persona in Covers.

[stable spire]

Please add the ability to delete multiple songs at once.

[twilit hedge]

I think so .. that's what Cover main function.. cover the same song in different styles.. but like anything on Suno... it's sometimes doing what it wants.. just keep at it!

[turbid drum]

yea you can and its not that hard tbh

[marsh lichen]

Can you please return the v3 model to the Edit menu? It was the best when you wanted to end a song without it continuing for 2 minutes. And please make the Edit model more visible - now it only shows in the Edit menu and is not seen when you make changes in a waveform.

[pearl trellis]

I came up with a feature request last night. A karaoke mix! Get a gen that you really like, hit the karaoke button, it gens again with the same song note for note but minus the lyrics. As a plus, the new song could possibly add the lyrics in text with l the bouncing ball or highlighting. As a plus plus, the user could (tech dependent) listen to the song with lyrics on ear buds while video / audio recording him/herself singing while watching the bouncing ball and the app combines the person singing along with the audio track with no dialogues. As a plus plus plus there could be fun ai backgrounds.

[sly delta]

Three dots -> Create Get Stems. Wait for the generation. Remaster instrumental track.

[pearl trellis]

Not bad 🙂

[sly delta]

The stems separation takes a bite out of the instrumental so that's the reason for the remaster

It won't be exact but its usually usable

[pearl trellis]

🙂

[formal bobcat]

Yes, I have tried, but it never gives a satisfactory result

[slow walrus]

that would be amazing

absolutely amazing

because I have 2 singers and its annoying that persona saves just one, so this would be a good solution

And:

Playback Queue:
Let us open the current playback queue list, enqueue songs there, etc just like on any streaming service

Playlist:

• Search function in Playlist
• random playback

[ornate osprey]

try Replace, it is a similar voice... but yeah, Extend doesn't work. Try to make less than 2k character songs if you use Persona. It would be good if we had Persona working on Extend function. It's the functionality I want the most.

[slow walrus]

I think he means that we can change to an other persona midsong

That would be really cool, because then I could start with my male singer persona and use later my female persona singer as well, otherwise she would have a new voice

[ornate osprey]

maybe that doesn't work good because Persona keeps the style, everything of the song and probably changing would end up with weird results. But keeping the same Persona on Extend should be easier and it may work better.

[slow walrus]

hm I have no issues with keeping the same persona on an extend though

well both persona's are "deutsch rock", so that should fit good together in my case
But maybe it would be not the worst idea to be able to select if you just need the voice, or the style as well

[glass nacelle]

Please add a song interrogater. A way to feed a song clip into the AI and get prompting descriptions. This would be an awesome feature; particularly for those like me that are not musicians. I would absolutely spend credits on this add on.

[slow walrus]

https://chatgpt.com/g/g-y2J4bAO8A-ai-music-prompt-generator

that one understands band names

and will give you prompt for suno

of course if you ask for Rammstein, you wont get rammstein from suno, but you will get music style fitting to it

[fair turret]

BULK REFUND TOOL- Please add tool to give bulk refunds on multiselected clips with the following options for reason: GIBBERISH, SUNG WRONG,TAGS SUNG, NUMBERS SUNG, SUNG OTHER THAN INSTRUCTED

not right we ahve to keep paying for mistakes of the Ai

[glass nacelle]

Yeah, I already do this. But I need an interrogater. For example, I hear this cool music clip. I feed that clip into the interrogater, and the AI spits out a description ie, syncopated bass with rolling hi hats at a 1/64 , etc

[slow walrus]

I got a song with a sinus wave instead of music the whole song through.
I didnt get credits for this back either 😄

[reef basalt]

I think it'd be nice if Suno could auto generate a .json file for subtitles. This would make making YouTube lyric videos for your music a whole lot more streamlined instead of having to set up the lyrics yourself. Vizzy.io actually has a lyrics section that takes .json files and automatically adds them in and sets the lyrics wherever you want them on screen but if you don't already have that file type for the lyrics you have to add them manually which can take some time if you have long songs.

[viscid locust]

Would be awesome if you could feature a persona in your songs.
For this example the persona is John.
[Verse 1] - [Chorus] - [Verse 2 John] - [Chorus duo] - [Bridge John] - [Chorus duo]

How to start a duo?
You'd go to the persona page where a new button called 'Feature" or "Duo" would be, you click that button and it drags you to the create page.
Or it would drag you to the library where you pick a song you have created, you edit the song and on the edit part you highlight the part you want the persona to do instead.

[sage gorge]

Not sure why there isn't an option to purchase a license for a single song made as a free plan user.

Don't use this platform often enough for a paid plan but there are a few outliers that would be cool to have rights to.

[sweet dragon]

is there a place i can easily find a list of all the commands/instructions you can use with brackets while making songs? ex.) using [chorus] to denote the chorus, while below it using [verse 1] to signify the start of the first verse

[reef basalt]

This is what I've used in the past. Unfortunately, suno wiki has died so can't rely on it anymore. But generally you can find some helpful stuff here though as people update it in the comments.

https://www.reddit.com/r/SunoAI/comments/1c6qevk/megathread_suno_tips_tricks/?rdt=60897

[balmy zealot]

..
Downloading: making the "Where from:" data the actual link to the song on the website would be great. right now it's "blob:https://suno.com/####-####-####-####-########"
I tried removing the blob: and those links are 404 error
I also tried looking at the url structure on your website and saw /song/ missing. That doesn't work either because the ####-####-####-####-######## doesn't match anything on the site

[mighty tartan]

11

[balmy zealot]

Users can see if a song was "Extended From" with a link to the song.
Users can NOT see an "Extended TO"

This is making it challenging to update and organize old creations into the new feature of Workspace; particularly ones that aren't complete with a "Get Full Song" tag because users can find the end of songs but cannot begin from the first generation and work their way forwards to the end of where they left off.

[terse rapids]

My suggestion for the "View Profile" UI is to add another tab where all songs are listet having a comment/feedback from other users. As it would be for the artist easier, finding those and reply to the song feedback/comment.

[formal bobcat]

What's a sinus wave?

[formal bobcat]

I have not found one yet

[slow walrus]

https://www.youtube.com/watch?v=PyD9cMarVJk

Sine wave, seems like I did translate wrong,sry^^

[formal bobcat]

LOL, just kidding with you

When I create a song, and it's all good, and then I want to Get Stems, i.e. make an instrumental version without the vocals, I can still hear quiet echoes of the vocals in the music. How do I get rid of that? Surely when creating the original song, Suno should create the music and vocals separately so that they can be then provided as separate streams, as well as the complete song with music?

[slow walrus]

Stems uses a center channel extractor from the final mix, just like you would do in an audio editor.
I've heard currently the AI doesnt output these things seperate so that stems cant access it earlier

[gentle galleon]

FYI. Shimmer is still a problem. Just Incase you were trying to ignore it. It’s a real problem.

[peak knoll]

Can the AI sing the lyrics in written order when we extend the songs? I typed in 2 stanzas and the AI sings the bottom first then the top. Sometimes it sings the bottom stanza 3 times and completely ignores the first.

[peak knoll]

The edit mode on the website for section replacement needs work. I have 4 lines and I want to change only 1 line. It ignores the voice/style of the other 3 lines and composes a completely different voice/style for the line that needs replacing. Spent over an hour trying to change only 1 word in the entire line but it won't match up with the other 3 lines.

[pulsar estuary]

new edit mode changes the style alot, not sure what that's about

[proven dagger]

hi team can we add option to automaticly add tune to social media,or youtube, or how we can dowland more then one sone at once , it would make life much easier

can we add promt to change instrument form playing sone to any other?

[pearl condor]

Bring back remaster feature for free users or what I mean is at least make it a limit to 5 remasters of songs per month

[teal schooner]

Notification Options

Please can we have the option to select which notifications we would like to receive?
I would like to know if someone has followed me, liked a song, commented or replied to a comment.
But since I've followed quite a lot of people, I don't want all the notifications about them having dropped a song.
Therefore it would be useful to pick the type of notifications we receive. 🙂
Thank you

[eager igloo]

Is there any way to download a SRT file for the songs so that can create our own videos?

[worn moth]

please when creating next part(extend) of a song make the limit for more than 2 minutes. Thanks!

[upbeat rock]

Extend should have a duration window. So when I'm looking for just a 15-30 seconds outro, I can specify 30 secs length for extension

[upbeat rock]

when remastering a clip extension, it no longer allows you to "get whole song" - would be nice to remaster a clip and then flatten it to full track when it's ready

[keen marsh]

i dont like the new extend function 😦 it doesnt work rly well...

[proven dagger]

i can build a code for a ai agent to atomically uplode to YouTube or other social media

[twilit hedge]

I had better results when Remaster the instrumental track... it'd remove most (if not all) the vocal bleed in it.

[twilit hedge]

I think it'd be best if there's an option to choose how long the extension would be.. a range from min to max.. sometimes I just need to extend 10s or 20s ... etc...

[glossy spear]

version 3.0 produced better songs. Why is it that now when I want to enter lyrics and generate a song I get an error when I select the 3.0 option?

[tepid trench]

v3 generates shorter songs and the lyric length is lower (I think it's half what v3.5 and v4 allow?) So try removing some lyrics. This is the most common reason I'm aware of for this problem.

[glossy spear]

Ok thank you, and why does version 3.5 perform worse than 3.0. The songs generated are stylistically inferior and the vocals sound very artificial. The best songs came out for me in version 3.0. I'm thinking of buying pro, but for now I'm discouraged.

[tepid trench]

I don't know the reason for that. Probably just trained differently as there were certain things about v3 that quite a few people didn't like (vocal layering happening a lot for example.) Most of my V3 songs degraded in quality so I never went back to it after v3.5 and now v4 is here I almost exclusively use that.

[glossy spear]

Ok thank you for the information, and is there any way to make the suno listen to what it is commanded to do? Because I often see that it generates songs that don't stick to the layout?

[fair turret]

EXTEND POINT when doing extends etc pleasea allow us to choose the same granualr point in a clip that you can in replace rather tahn truncating the number

DRAG AND DROP SECTION; Please allow us to select a section of a song and drag and drop or choose an insertion point by beat elsewhere in song.

VISUAL BEAT INDICATOR . Please give an onscreen indicator of drumbeat

SELECTION LENGTH - PLease allow us to control the length of an added section of song especially at the end

[tepid trench]

No way to force it to do something. Closest way is to just repeat the same prompt more than once and hope it's "seen". #📚┃suno-school may be a better place to get general advice on how to do things (this channel is for suggesting features)

[fair turret]

Add more of the same Metags order matters

[glossy spear]

Okey thanks

[fair turret]

best you can do

[sharp meadow]

ooooo I thought of something, make it so we can possibly create transitions between song to song. So if we want to make our own little suno album, we can make our songs transition into each other on it

[nimble shore]

DELETE SECTION: How about in addition to REPLACE have the DELETE SECTION option? Most of my replaces happen when Suno starts hallucinating and either goes really off or start spinning the record at the same tune.

[wheat dragon]

Is there anyway to expand the (styles) section on the creator on browser? It's like ever since the new update it's smaller and it doesn't expand like it used to so you can see all of your styles and genres

[hybrid tulip]

There are two things I things that would really be cool to have is:

1. Word replace: I know that I’ll have a nice version but suno will go off the rails on one word and ruining the song. The current replace usually complete rewrites the section completely.
2. A swap feature where if you have two versions of a song both with really cool parts you can move one part to the other song and have Suno match the section up to the song

[crude river]

Should we also Select parts for vocal melodies?

[marsh lichen]

I didn't manage to replace one word until I turned off the manual replacement toggle.

[radiant shell]

I've been wanting those two things as well.

[hybrid tulip]

This is a good point. Maybe there could be a selection for music, vocals or both

[broken canyon]

Could we have multiple persona's per track please 😄

I would love to have a persona breakdown so we could say
Persona 1 : Option = vocal / instrumental / none
Persona 2 : Option = vocal / instrumental / none
Persona 3 : option = Vocal / instrumental / none.

Then within the track be able to influence the vocals by a prompt similar tp [vocal 1] or [persona 1]

It would allow for a lot more recycling of bad gens rather than binning them and would also allow for the featuring of persona's on other tracks

[bitter vine]

hello

could anyone help me?

[scarlet cairn]

I would love for when we generate the instrumental stem from a song, we have a video karaoke of the song to go along with it!