#💡┃suggestions-web

1 messages · Page 3 of 1

vapid basin
#

I make tons of free stuff for people all the time and haven't monetized anything. I'm also one of the lead developers of Deforum. So, I don't mind being forced to give those songs away for free. It's just awkward when people are like "You should make an album with those!". It's hard to assure people that it's actually impossible.

frigid timber
vapid basin
#

Yeah, I'm a musician, but haven't done any music for quite a while. Suno.ai definitely makes it easy, and allows me to write lyrics for styles I wouldn't even normally mess with.

frigid timber
#

exactly! plus mixing up multiple styles and it somehow just works is the best

vapid basin
#

Suno is so far ahead of everybody as far as I'm concerned. Really great work, ya'll!

frigid timber
#

thought it was phone recorded initially with the song overlayed then those effects started happening lol

still forum
#

I've been making beats for almost 6 years. Would be great if I could upload my own work and have the AI write songs to my music. My biggest struggle in finding artist to collab with but if i could write my own stuff or have AI write over my beats i could have soooo much fun with this.

static mist
#

hii

bright mist
#

One suggestion for the website would be to be able to:

  1. Start by saying "I want to make a song"
  2. Within that song first find a seed / style that suits what you are looking for
  3. Then define some parts like an intro, verse, chorus, bridge, etc.
  4. Start generating those parts, with the ability to redo parts

Would love to have this instead of this current "chain" that we have right now, where you loose the seed/style if you have to start over.

crystal wedge
#

Hey Suno Team! I would love to talk to you guys as we just made one of our upcoming podcasts intro with this and would love to partner with you to promote you on the website and other things, can we chat in DMs or over a meeting?

gaunt creek
#

Search bar for library.

vagrant rock
#

Creating folders in the Library and an option to DELETE files. Without proper filemanagement this great tool is doomed

tidal wharf
tidal wharf
tidal wharf
#

Would be cool to have stable diffusion webui-style prompts with ((attention increase)), something AND something, and suchlike, work in style prompt

wintry forge
#

I would like to see an infill feature. Let's say that I get a generation and it's mostly good, but there's a section from 0:49 to 0:57 that has some sort of weird artifact to it. I'd like to have it regenerate just that section and keep the good parts.

wintry forge
#

Another thing I think would be cool for AI music generation in general would be optional hyper-granular controls on the output. I often have a specific idea for a song and would want it to sound exactly like what it sounds like in my head, so that I'm making my song, not just a song. This is more pronounced by the fact that the music I actually want to make is EDM, which is often instrumental, and right now all you can do is prompt "[Drop]", which is basically useless for what I'd want to do. Some ways that this could potentially be fulfilled with features would be things like

  • A way to input your own melodies and chord progression
  • The ability to iterate on specific parts of a track. Something like being able to ask it to change just the bass line or something.
  • Some sort of "img2img" equivalent. My dream workflow would be to sketch out an incredibly rough version of a song in an actual DAW, feed that into an AI, and tell it to make it not suck. Have it add missing fills and bass lines, do the mixing, and just generally make it go from a poorly executed idea to polished output. "Take this instrumental and this vocal track and mix it for me."

I'd imagine that a lot of these suggestions require groundbreaking new research to implement, so I'll be patient. But as a self-identified musical "ideas guy" who struggles with the more technical aspects of music production and is overwhelmed by how much it would take to overcome that, I feel like there's a lot of potential for tools that solve that problem for people like me.

charred breach
#

I want to create anisons (anime songs) in TV size (90-seconds long), But every time I try, it doesn't even get close to 90 s (The maximum time is 80 s each), but they barely make it to half (~45). Please, I hope you include prompts for song lengths and can last a maximum of 90 s.

Merry Christmas! 🎄🎁

tidal wharf
tidal wharf
#

suggest exports in wav. I understand that the quality isn't expected to be perfect, but I can imagine additionally decoding it into mp3 spoils it even further

raven comet
#

😆

tidal zinc
#

This AI is insane, I believe it can be even better if it gives us more export and editing options

bright mist
#

If the like/dislike buttons are for providing feedback to the team, then could we have a favorite button for us users instead?

tidal zinc
#

I signed up for a plan and honestly I didn't see any advantage other than having more credits

steep stirrup
bright mist
#

On the website, some kind of indicator / info message / red flag for things that will block song generation in the input fields. If that is possible.

wild ermine
#

#💬┃general-chat message:

One other possible piece of feedback: Include a link to the "raw logs" of the API calls to ChatGPT. If we are going to be paying and need to do a bunch of guessing, potentially burning through our GPU allocation, I think having the full logs is not that strong of an ask.

bright mist
#

Use indentation in the web-ui when making part-X of a song, so you can visually better keep track of which song is child of which parent.

left glacier
#

any tips on creating variation in clips? I'm working on an ambient project but I'm having trouble getting much variation most clips end up just becoming a humming noise or they just repeat

vast terrace
median olive
#

Suggest:
Can you add an option that allows the generation to continue but not from the end but from the beginning?

Because as the default algorithm is programmed with [Verse] and [Chorus] There are clips that have an incredible rhythm but without an proper Intro or breath kind of sound bad and unusable, to generate an Intro LATER, as the way the AI ​​currently works, results in SEVERAL lost credits.

Also:
The REMIX button, SHOULD give us a song with the same style/rhythm/composition, right?
kinda like it doesn't make sense how it works right now.

little cliff
#

hey @tiny bear @obtuse tree Plz take this in consideration, this is absolute easy solutions for this app :

1- Raise the song time for 2min(Max), 1 min(min)...double the chorus/verse if needed.
2-On "Song description mode" Make sure it has a short intro and an instrumental fade ending...not abrupt ends after final word.Or at leas give some seconds after final word and then fade out the sound.
3-Add a Reroll/chane lyrics or reroll/change instrumentals for a song created.Sometimes we got a killer instrumental and a shitty/bug/not good lyrics and vice versa...a reroll lyrics or Instrumental would help a lot.
4-Option of downloading the beat/instrumental track isolated.
5-option of female/male singer

brittle birch
#

Increase the character limits on the "Style of music." 2-3x the current should do it.

normal badge
#

Hello all! I have a suggestion to be able to remove clips you made in case of there are too many of them one by one or to be able to clear the history of what you did, also it can be helpful in case you experiment a lot with different generated things and you have so lots of them that you'd like to clear all them so it is easier to start new clips and find them easily

frosty topaz
#

They should add comments to the suno where you can comment on other peoples songs

sonic trout
#

Having PayPal as a payment method would be much appreciated

spiral crater
#

In the Suno UI, or Discord commands, it would be great to have a similar option to the "Continue" but to append content before and existing clip instead. For example, adding an intro to an existing clip of a verse. Additionally, it would be good to reuse clips like Chorus in other sections to build a consistent song and not having different chorus styles when generating a whole song from different chirps.

indigo narwhal
#

Add the ability to choose which key we want

tough hound
spiral crater
#

It would be awesome to retain a specific voice from old generated clips to generate new songs but always with same voice tone.

normal badge
leaden marlin
#

Hello, is there any Korean person around? I need help!

light skiff
#

Hey guys, i love your product. Any plan on releasing an api soon ?

leaden ether
#

Lovely product! Thank you. Please add the capability to edit the lyrics from a generated song. You can only continue but not modify the generated, and in my case I would like to keep the same voice/rythm/structure but try other words (custom mode)

balmy zealot
#

when Style: Random is used = define the style chosen for the .mp4 so it doesn't say RANDOM

tidal wharf
marble mountain
#

As I am new here, I gotta admit I am just blown away by Suno.
One thing I miss so far, is to make clips shorter before continuing them. Some times the generation is spot on, but at the end gets weird or ends in the middle of a sentence. Would love to make it a bit shorter instead and then continue generation from this clip.

tidal wharf
misty blade
#

I found that last days Chirp make full length clips even when it not needed, and the last sentence go on and on.. it harder to get good finish to song. Almost every generation is 1:20 minutes and every continue is 1:00.. I think it better how it was before.

marble mountain
#

I limit the lyrics each time, but it is hard to tell how far it will go. Some times it is too short, some times it is too long and adds a lot of senseless filler or lyric repeats in there

misty blade
marble mountain
#

@misty blade I learned about Suno 3 days ago and have a subscription since 2 ... so absolutely no clue on my side 😅

misty blade
#

I saw that when the generation is too short I got credits back. but sometimes (maybe because it in Hebrew) I get totally empty generation, somthing like 1 minute of silent. Can you give the credits back also in these cases?

marble mountain
#

No, that is absolutely the case. I did get refunds for too short clips too. I do not have issues with the actual length being too short. My issue is that it doesnt finish the lyrics. If I add less lyrics, it some times makes way too long clips with a lot of filler and repeats. Both could get solved by allowing me to shorten clips manually and then continue from there

misty blade
#

I agree it'll be great. I saw that sometimes [fade out] or [outro] can help you finish after the words

marble mountain
#

I actually do use those tricks already for the actual end. The issue is more when I want to piece longer lyrics from several clips

misty blade
#

still the best way is to take it to external editor to finish the project properly

marble mountain
#

I realized that I get constant copyright warnings when I ask to make a song with no lyrics or singer 😅
My workaround so far is to use Custom mode and write [Instrumental] as the lyrics... I think there are probably a lot of cool hacks there I dont know yet

misty blade
#

sometimes you prefer to duplicate the chorus in external editor

#

hm...

marble mountain
#

Yup, some stuff is better with external editors, but some stuff is better when Suno can continue itself. Both have pros and cons IMO
(Plus I am very rusty with audio editing, so that takes a LOT of effort on my side 😂 )

misty blade
#

maybe when you ask no lyrics you get something like real song. I wonder whether the fillers jibrish is taken from somewhere?

#

What you need to know about editing is how to cut right on the beat

#

so you can glue 2 clips together flawlessly

marble mountain
#

Suno obviously has to protect itself from plagiarism. So I guess no lyrics just triggers that and [instrumental] is probably on a whitelist somewhere.

misty blade
#

I prefer when suno make the chorus the same, because it sound more natural when it's not exact the same

marble mountain
misty blade
#

but sometimes it make every chorus different

marble mountain
#

Yeah, agree. Sometimes you want stuff the same and it gets very different. Lets see if they find a solution to it at some point. I have no good idea on my side how to do it even on UI/UX side

marble mountain
misty blade
#

I made some test, I copied wikipedia article into the lyrics and made a rap

marble mountain
#

Yeah... the copyright protection is not 100% yet. I tried some real lyrics too, just to see if it works... and it that case it did. Would never share something like this though

misty blade
#

ah I'm all the time put real lyrics.. its sort of "cover"

marble mountain
#

Yeah, true! Cover songs basically 😅

misty blade
#

it's very interesting to get new intepretation to existing song

marble mountain
#

I am not sure what this discord is like, but we might be spamming the suggestion channel a bit. Do you mind switching to #💬┃general-chat maybe?

misty blade
#

yep you right

#

but we also have here a lot of suggestions during the conversation

balmy zealot
#

When making a track public for the first time, perhaps allow a rename of the Title.

misty blade
#

I just asked ChatGPT to summarize what we suggested here:

  1. Adjustment to Clip Length: There's a need for better control over the length of the music clips generated by the app. Users are experiencing issues with clips being either too long with unnecessary filler or too short.

  2. Refund Credits for Empty Generations: There's a suggestion for the app to provide credit refunds not only for clips that are too short but also for those that are completely empty, particularly in cases where languages like Hebrew are used.

  3. Manual Shortening of Clips: Users expressed a desire for a feature that allows them to manually shorten clips and then continue from that point. This would give them more control over the content and duration of their music projects.

  4. Handling of Chorus Variations: Users prefer consistency in the chorus sections of songs. There's a suggestion for the app to ensure the chorus remains the same in each iteration, rather than generating a different chorus each time.

  5. Improvements in Lyrics Handling: Suggestions include better handling of lyric length to avoid overly long or short outputs and reducing unnecessary repetitions or filler content in the lyrics.

  6. Feature to Avoid Copyright Issues: Users discussed the need for a feature to circumvent constant copyright warnings when requesting songs without lyrics, suggesting a specific mode or tag for instrumental compositions.

stone solar
#

in regards to #6, I think having a "no vocals" option, and/or allowing the option to download separated vocal/instrument tracks would circumvent the copyright warnings and provide utility that has been requested

tidal wharf
#

copyright warnings not only come from the lyrics, sometimes it "thinks" (for whatever reason) that you are trying to replicate a specific artist, which is also prohibited

marble mountain
stone solar
#

yeah, but even if it isn't how the clips are made, doesn't mean that separated tracks aren't critically useful

marble mountain
tidal wharf
#

sometimes it's getting crazy, working fine and then all of a sudden blocking all my prompts although I didn't change anything

#

openai occasionally throwing a fit i guess

marble mountain
stone solar
#

yeah, and sometimes i'll just retry without changing the prompt and it works fine

#

if Suno operates in a manner where it can provide separated tracks without significantly rewriting the whole project, it would be optimal compared to AI separating the tracks after the fact

#

in regards to copyright warnings, sometimes I'll just insert a blank line between lyrics or something and it helps

marble mountain
stone solar
#

gotta keep in mind though, that there's a lot of media out there, and a copyright flag might be real in a language you don't speak

#

in regards to separate tracks, Suno also seems to develop the lyrics & instrumentals hand in hand, so there may be some issue keeping them separate if they're developed as a single output from the AI's perspective

#

its not like it just pumps out vocals and then says to itself "what music would go with this ?"

marble mountain
#

Yeah, exactly my expectation

#

About copyright issues: it shows quite some good thinking business wise that they dont want to risk it. It is overzealous some time, but better than getting into hot water at the start

stone solar
#

yeah, i find a lot of bad actors using press/clickbait on any new AI out there arguing "omg, look at what people are doing!?" so having anything remotely like cover songs could be difficult for Suno

feral aspen
#

If you get blocked for copyright even with purely original lyrics and you have a minute, add the prompt to https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/

Google Docs
Suno.ai - Song Moderation Mistakes

Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.

Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...

zinc willow
marble mountain
#

If it ends at the end of a verse of sometimes at the end of line, that works for me with a bit of trickery. If it ends mid sentence, I just couldnt make it connect and finish the verse

zinc willow
#

It be nice if it follows the song like Karaoke and then the Ai knows to continue

marble mountain
#

Yeah, it be nice. Mostly it works but in a few cases it causes issues. My main suggestion was to be able to shorten clips, which would also help if you like the start but just not happy with the rest

zinc willow
#

How do we had Instruments?

#

I got this song i want to add Australian and Chinese instruments

burnt yarrow
frozen osprey
#

Been loving the web app! I do second the volume control though, sometimes the preview player is real loud

ornate snow
#

Premium-access API to Chirp, please. Please

stable wolf
#

Could you put a button to select several tracks at the same time and download them?

#

☺️ please 🙏😄

stable wolf
trail hawk
#

It would be nice to if the BPM ofthe song is shown.

zinc willow
umbral coral
#

Let me distribute my song through Suno.

tidal wharf
delicate marsh
#

Suggestion: THe ability to download 'stems' of the instruments, or at least the vocals and instruments separately, for better mixing, track replacement in a DAW

tranquil cypress
#

I hope it can have the function to customize song details, keep the vocals or song melody, or keep both and just change the lyrics. also added the function of increasing or decreasing speed with the rhythm of the music

#

Many times I have a very good melody but it automatically creates terrible lyrics, or there is a voice I really like but the melody doesn't match, it makes me crazy.

clear flare
#

Please provide API interface.

winter oyster
#

Would the team at Suno please consider a Pay as you Go plan? Subscription plans are not ideal.

indigo kindle
#

Negative prompt text box would be useful and optional instrumental checkbox

dusty dove
#

I have some suggestions for improving the platform:

  1. Increase the time to at least 5 minutes so that we can better evaluate our compositions.
  2. Use of a "template" so that we can always create a composition based on the template that looks best.
  3. Preserve, for example, the voice and vocal style for future compositions.
  4. Continue the clip automatically when the time exceeds.

I think that at least this would help us a lot to think about the SUNO platform as our official composition platform.

marble mountain
marble mountain
# winter oyster Would the team at Suno please consider a Pay as you Go plan? Subscription plans ...

Once you have a subscription there is a pay as you go plan to get more credits, which is a lot more expensive per credit than the subscription. Hardware costs for AI are real and pretty huge from the numbers I read so far. Personally I just recommend to go for the monthly (careful, not yearly) plan and immediately cancel. Sure it is only one month, but at least works great with how my own curiosity usually works.

winter oyster
marble mountain
#

The cancellation is only active at the end of the period. So for a month you re a paid user then. With credits and commercial rights. It is a bit more expensive to pay monthly instead of yearly, but for me definitely worth it. Will probably extend next month again, for now i am happy I dont have any subscription to think about.

raw comet
#

i know this should come natrually as the technology evolves, but for paid plans i'm hoping we can eventually achieve a bitrate higher than 128 kbps

graceful dawn
#

Yo you guys gotta add an option to upload your own instrumental and then plug in ai or original lyrics. I would love to be able to use an ai background singer. Also the results on the web app are way better than through discord.

delicate cedar
#

It is still being worked on though and should be available in Discord Real Soon Now™️

sudden magnet
#

Hello! Really enjoying Suno! Is it possible to generate songs without any vocals? Only the background beats?

urban maple
#

Hello, it would be nice if there was a way to customize the music a bit, like choosing a male or female singer, or a duet, BPM, types of instruments we want, or changing the singing or instrument style midway through a song, the more customizations the better 🙂

willow bluff
sudden magnet
#

It would be great if we can still style it in the same way but just get the background without the vocals

oblique valley
willow bluff
oblique valley
charred breach
toxic haven
#

hello

#

Can you add the genre argentinean : RKT? is like reggeaton but difference

stone solar
#

if you haven't had luck with the direct prompt, you can try suggesting its influences and see if one comes close

#

like "Argentine minimalist Reggaeton with bassline"

toxic haven
stone solar
#

band names wont work

astral shard
#

sort liked clips by genre

robust jewel
#

Not sure if it would be possible with how things are currently set up, but it would be great to be able to extract and save a singer from a clip to use later. You sometimes end up with a great voice that would be perfect to use but not with that clip.

zinc gyro
#

I don't know if it's in Suno's future plans. But the possibility of synthesizing your own voice to use as a "singer" for the songs would be an interesting way to get away from the controversy over using the voices of well-known singers.

tidal wharf
ripe mango
#

A way to find Lyrics and a Full song filter

normal badge
#

Hi all! Here someone suggested to dislike song to get it removed from library but it doesn't work

safe sand
#

Not sure why I have to suggest this, but in the web version, can we get a DELETE button for all the clutter of songs we generated and don't want anymore? Or am I missing something?

zinc gyro
# tidal wharf >from the controversy is there?

I'm proposing a "plan B".
Certainly there are/will be legal controversies debated in the courts, brought by singers who have had their voices cloned. Madonna put a ban on the use of her image and voice in her will, even after her death.

woven merlin
#

I want to pay more. For a better sound quality, unprocessed song files.

frosty escarp
#

Ability to have more than just Intro, Verse, Chorus, Outro - but also include elements such as Bridge/Pre-Chorus. Also - to build everything in blocks (similar to Shortcuts in Apple OS’s)

split sage
#

hi - is it possible create song longer 1 munute with free plan?

delicate cedar
solid mesa
#

我有积分怎么无法使用

#

请问有人知道吗？

delicate cedar
solid mesa
#

我给他发了邮件，叫我耐心等待

rapid sierra
#

We need to be able to regulate the volume of the player.

golden hull
#

Is it possible to add a reusable SEED so that when you get something really good you can alter you lyric and get a similar style again, and again, and again. Like you can when you are on "continue from here"..

sage wedge
#

Better Sound Quality Please!

feral aspen
#

I might add bpm detection since that's also easy, if i get time

novel mirage
#

Allow us to choose the voice for the music

#

And maybe use ur own voice

urban maple
#

it would be cool if you could tell it to make a prequel to a song, like adding an intro to a song.

sage wedge
#

If the music cannot be downloaded in high quality in WAV format, it cannot be monetized, and therefore, it is not worth paying any kind of fee. This is something that should be improved.

visual rock
rare ice
#

devs must definately find a solution for the auto generated lyrics, i mean it looks like chatgpt's lyrics, but they are always the same, obsessed with neon lights, city lights, groove and funky beats. AlsoThere must be a prompting guide to enlight us how we can

#

navigate through the generation phase, otherwise, its just pure luck to find the good sound

#

the UI is also need an update imo, why there is no editing section where we can edit the song that we created?

feral aspen
rare ice
#

and 1 final suggestions, it would be awesome to be able to choose male or female vocals, i tried couple times but there is no consistency on that front

robust jewel
#

One option is to just use GPT yourself and prompt it in more detail about what the song's topic should be.

#

Honestly, though, it's easier to get varied lyrics out of GPT than it is to get anything other than AABB rhyming. Good luck getting something as complicated as ABAB, from my experience.

errant cargo
#

•bmp info and setting (typing the bpm in the prompt works 50% of the time)
•seed info and seed options (random seed, fixate seed)
•Quality Upscale
•Library with Folders or some type of tagging for clasify songs in the library (more than the like or not like tag)
•A little manual or prompt parameters handbook please

Pleaseplease, amazing service! 💯

last star
#

/How was this song made and is longer than 1 minute and 20 seconds?

#

/The songs I make cannot exceed 1 minute and 20 seconds, but how was this song on the trending songs on the suno website made that it is longer than 1 minute and 20 seconds?

distant moth
#

Will it be possible for us to train a specific genre on website? I want to generate phonk music but I think the Ai hasn;t been trained on that particular genre. (I am a subscriber monthly). Hope to hearing from you PS you're doing great service!

delicate cedar
#

I told you this when you DM'd me. Read the responses you get.

My apologies. I see by the time stamps that you posted this before you DM'd me. I hope the link I gave in the DM and above helps you make your songs longer.

upbeat sable
upbeat sable
delicate marsh
#
  1. Ability to specify BPM
  2. Ability to specify Key
  3. High Quality export as a single wav (48 Khz, 32 bit)
  4. High Quality stem export ie separate wavs for each instrument and vocals
  5. Upload an instrumental track and then have it add elements which are requested (eg: Vocals or guitar)
  6. Upload a vocal track and have it add instruments
tidal isle
#

Better payment plans with more song creations or only pay credits for songs that the user download as mp3.

rapid sierra
#

I think that Suno should update its copyright policy. I agree that free users should not use the generated songs commercially, but it should be clear, that Suno (or other Suno users) will never claim any rights to the users lyrics.
Some peoples are making their own lyrics instead of generating it, and thats is a hard work. Some peoples are using a copyright characters in their lyrics. Some other also puts their personal stuff in it. Some peoples are making covers of an other songs, even if Suno is trying to prevent it.
If Suno would be try to own it, it would lead to serious problems.

robust jewel
#

They made it very clear that you own your own lyrics

stone solar
#

thought of one of a silly little suggestion just now, "get rid of the birds". It would be nice if generated images for clips referenced previous prompts for inspiration to help generate images more relevant to the song as a whole and end the whole "bird image" for clips with only outros/instrumentals, etc...

weak elk
#

well the good thing about birds is that everybody's heard about the bird

delicate cedar
#

Of course! Bird is the word.

errant cargo
finite owl
#

Can't I make it longer than 1 minute and 20 seconds? The back end of my song gets cut off every time. Is the maximum length 1 minute and 20 seconds?

finite owl
#

agree

zinc gyro
# rapid sierra I think that Suno should update its copyright policy. I agree that free users sh...

I think time will correct some distortions. First is to define "authorship" in the lyrics or melody when something is created by AI. Being created by a free user is indifferent. In practice, this would mean exchanging the right of the author for the right of the "sponsor", which is a contradiction in terms. I find it contradictory, or at least strange, to contest authorship and determine that the platform should be cited when publishing anything externally.

hot flower
#

i need help , few days ago i made video about suno ai and someone texted me this..
I was very impressed with your Suno AI, through which I introduced a company in Vietnam - specializing in supporting young authors to compose music. And they asked me to connect to deploy a while lable Suno that can help the young musicians improve their composing abilities through the support of Suno AI.
I hope you can join me in deploying a customized version of Suno AI for my partners and if so, please give me the subscription fee you wanna qoute.
Thank you very much.

solid wagon
pure fog
#

Would be great to get an uncompressed output option , not just mp3 but wav etc

main kernel
#

The "500 songs" claim for Pro is not very realistic. As amazing as this technology is, I wouldn't exactly call a forty second to one minute snippet a song.

So, cut that "500 songs" in half, since you're not very likely to use both versions of the same snippet. Then the average song is 3:15 (195 secs) so it's going to take about four to five snippets to actually create a full-length song.

That 250 snippets now only really yields 50-60 songs.

Even if you're only happy with a quarter of those, it seems like a great value.

I just appreciate up-front transparency.

SUGGESTION - You should just say:

○ 250 Generations
○ 2 different Variations per Generation
○ 50 - 60 Songs *

  • Based on average song lengths and typical Generation length (results may vary)
tidal isle
#

This appears when I use my own written lyrics: "Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." Its a big nogo.

gray dirge
#

Hi, Any chance adding a pro feature where users can upload voice models (only the account can use) not public (like created from sovits, rvc, or voice banks), and being able to use it alongside song generation?

#

because sometimes you get a nice beat or generation but the voice/singer is not the one ☠️

Read, Some of the past suggestions above, hope some get added 🙏

quasi jay
#

One thing I want to see is the ability to create playlists where I can add generated tracks to which I can share

quasi jay
honest bridge
#

Minor suggestion: In the web UI > Create mode > the song list is force scrolled to the bottom of the list (newest songs) when new songs are generated. That scrolling behavior seems to assume that users always want to listen to their newest songs right away. Unlike images which only take a quick glance to measure up, generated songs require more time to take measure of and assess. Imo, a more user friendly way to go about this might be to instead provide some sort of notification bar similar to Discord where you can click on it and jump to the latest "unread" song, but otherwise leave the scrolling to the user.

ocean viper
#

Hey Guys,

  1. Would you mind adding possibility to generate one chirp (for 5 credits) instead of obligatory 2 songs (for 10 credits)?
  2. Maybe you could also add parameter to generate longer songs, billed 5 credits x minute? Allow people to input max 5x longer text for full song, then button generate (input) chirps in succession"?
    Both would be awesome!
small yarrow
#

Providing API service

brittle birch
#

A lot of times the start of a song, it clips, static or the volume is crazy high, clip and then goes to normal . This needs to be fixed.

zinc gyro
# quasi jay IMO the safest way is to write they lyrics yourself. I dont want to create every...

Most of the time I've done that. For Brazilian Portuguese, ChatGPT is terrible for lyric. It simply ignores rhyme structure. I've always had a knack for rhymes, so I prefer to refine or make them myself. I use GPT to develop ideas. It does that well. The chat rhymes are separate poor and obvious (at least to me). I've tested some of the available prompts and created a few, but I still think it falls short.

spiral crater
#

please add a functionality to have different projects or songs, so that we can group each short clip under same folder and not mix up when assembling a longer audio. Also add the option to delete or remove audios that are disliked, so that they don't remain in the library.

edgy bloom
#
  1. Ways to organize generations! Folders, deletions, renaming, and so on.
  2. Select in and outpoints for clips to continue from. Often weirdness happens at the beginning or end of a clip that you don't want, but the rest is awesome. If you could trim the clip before continuing that would help immensely.
  3. "Continue" from the start of a clip, backward. That is, allow for adding to the head of a clip, not just the tail.
  4. (This might be hard) Allow for more explicit, standardized commands that are used regularly, like [(Instrument) solo], [Instrumental], [Bridge], [Instrumental Intro], [Outro]. Also, for instruments, including male and female voices, choir, different kinds of percussion, etc.
  5. For sections like choruses that need to repeat, allow the ability to select specific timecode range in a clip and define that with a name that you can then include as [(section name)] in continuations. Bonus points for a weighting value that can control how much freedom the model has to be creative with the repetition (like CFG in Stable Diffusion). Also, a switch that indicates whether you want to repeat the instrumentation, the lyrics, or both.
  6. If it would be helpful for future model training, it would be good to be able to indicate why something was good or bad in a generation, and maybe a 0-10 scale rather than just thumbs up or down.
#
  1. Another possibly hard one, but some way to control the variation in the melody between lines. I find that many generations of verses use the same melody for all four lines of the verse, which gets old very quickly.
  2. Some kind of countdown clock to when your credits will refresh.
  3. Choice of number of rolls per button click. Credits are so valuable, I'd generally prefer to do one generation at a time to get a sense if my prompt is working.
  4. I've got lots of good ideas. Hire me to work on your team! This is a great product!
umbral depot
#

When I try to use Dutch lyrics, the voice is always from The Netherlands, would it be possible to add Flemish voices (from Belgium) as well? The difference is very obvious to someone living in either one of those two countries.

solar prism
#

We should have a channel for recipes, for different styles of music, combos that works well

#

Being able to create folder in the browser version

#

For better sorting etc

#

Teh ability to move songs that arent any good to trash bin that empty every "set days here"

rapid cipher
#

might be a good idea to say more than, "hey we will bill you monthly, but you can cancel at any time." How? There is literally no information assuring anyone that this can be done short of stop payments done from the bank.

bright lion
#

better audio quality than horrible 128kbps would be nice, mainly because at that bitrate there are extremely hard low- and high-frequency cutoffs and you REALLY CAN hear the difference between 128kbps and 320kbps with really high pitched stuff like hihats and cymbols as well as really low pitched stuff like drums and basslines

(oh, and i'll put another vote in for being able to delete, rename or organise items in our libraries, tho i can see that it is being worked on already)

frigid timber
rapid cipher
# frigid timber its on the stripe billing page right next to your plan information "Cancel Plan"

That's great, but it doesn't say anything about that when you commit to "monthly" other than what I said above. There are a lot of shady places on the interwebs and that statement doesn't really instill any confidence at all via a simple statement without some way to verify it before getting locked down into something you cant get out of short of a bank hassle and fraud complaints. Its a legitimate concern and one that may cost you revenue. I can guarantee you it did today. But by all means, don't change anything. It makes no difference to me either way. It was just a suggestion, in the suggestion channel.

robust jewel
#

They tell you that you can cancel it and then they give you a button to cancel it and you're somehow confused about if you can cancel it?

frigid timber
robust jewel
#

Were you expecting something more than a button to cancel when they told you it was possible to cancel?

rapid cipher
# frigid timber uhh ok, their terms of service says you may cancel at any time. Stripe handles t...

uhhhh.. before you commit, there is literally nothing else other than a little sentence saying that you can cancel at any time. this appears to be too complicated a suggestion, in the suggestion channel. ill just not worry with it at all and give zero dollars towards it. sarcasm on something that was meant with good intent is pretty childish and i am pretty astonished at the attitude with which it was received. thanks for the disregard toward it.

frigid timber
rapid cipher
#

Actually, I noticed that after my response. Who are you to even reply? Lol. Either way, your "help" was a disservice to whomever actually makes money here. Lol. Why would you even respond to a suggestion channel meant for people you are in no way qualified to answer for? 😂. Anyway, thanks for the laugh

frigid timber
rapid cipher
# frigid timber grow up dude

You've been on discord since 2020. You attempted, and poorly i might add, to answer a question in no way directed toward you and then tell me to grow up? Noted. Ill go do that now. just wow.

frigid timber
#

and to add as a paying member i know how the cancellation process works, so again was trying to help, clearly you enjoy being angry and stirring up problems.

rapid cipher
#

are you ok?

frigid timber
# rapid cipher are you ok?

im great, we can actually end this conversation. Hope they build a dedicated page for cancellations soon. Good luck

rapid cipher
#

i saw him typing.. had to block

slate geode
#

Can we increase the describe the music's character limit?

winter zephyr
#

Can we add a search function in the library?

marsh thistle
#

YES ADD A SEARCH FUNCTION i am wondering how trending even works at this point

raw bramble
#

I recently returned to Discord and learned about the availability of generations on the website. I'm intrigued to know if the website supports text-to-speech functionality. Furthermore, I'm curious if Suno allows users to utilize audio generated through text-to-speech for YouTube content creation. Specifically, I'm interested in creating comedy or entertainment videos where I can incorporate the generated audio, with the intention of referencing Suno in the video. This inquiry pertains to all plans offered by Suno

quasi jay
lost epoch
#

Add alternative payment methods like PayPal

tidal isle
#

Suno have no value, cause it dont accept all of my own written lyrics

#

Sometimes songs have words in more languages. So it needs a tag that tell suno that it shoud speak this word in other language. For example <the words in other language><spain>.

rare bay
#

I was running into some pretty significant bugs this morning (silence, shorter song lengths, etc.) and was commenting over in #🪲┃bugs-web , and the following came up: Musicians often start from, or get inspired by, a sound they are looking for. That might mean starting from a BPM, or more traditional notation verbiage (say, "alegretto"), or a solo instrumental by a specific instrument, or a vocal style (or even timbre). It would be great to dedicate more space to this in the description forms. Right now, the Create feature seems to lean pretty heavily on lyrical content as the motivation for creating a song. An option for downloading stems for individual tracks would be nice when a particularly good result emerges.

zinc gyro
#

Allow customization when putting together parts of the song produced in "Get Whole Sound". For example, you can exclude a bad verse without damaging the whole song.

tidal wharf
robust jewel
#

It probably also matters if the words being used have been... adopted... by English. There's probably a lot more Spanish that have been integrated than there is from Russian.

velvet sigil
#

I think it's absolutely ridiculous we only get 1:20! One of the most stupid ways to do it! The stitching thing is all broken, you can't edit the first part, second part just doesn't fit good. absolutely pathetic! Most songs are like 2 1/2-3 minutes and we should be able to create them seamlessly! It's the only thing so far I find horrible about this server. Why can't it just say charge an extra credit for every 30 seconds over the 1:20 or something?? It's very frustrating and for that reason alone I would love to find a different service to use! Also I hate it forces us to create 2 every time instead of giving us the option to only do one! like come on...

robust jewel
#

It's up to 1:20 from 0:40, with the change from V1 to V2. It's not limited to 1:20 because that's an ideal song length...

velvet sigil
#

yes, I'm saying it's stupid that it ONLY goes up to 1:20, I need longer and the stitching thing is a joke

grim belfry
#

I'm going to make a suggestion that I personally dislike, but here it is: make the clip time 60 seconds - because 60 second videos uploaded to youtube count as shorts, and get much better distribution 🙃

robust jewel
#

You could always get a first clip that you like the sound of and continue that with the thing you want to make a short. The continued clip is limited to 60 seconds, so if you don't choose to Get Whole Song, and just use that clip, it would work.

patent sonnet
#

Could you possibly make it so we can make "Playlists" on the web version?

Or folders. A place to sort your generated songs, just in case a few of our songs have similar themes or genres than others.

golden moat
#

so need vocal stems separated

marsh thistle
#

they need a way to split the lyrics and beat and be able to put lyrics onto a diffrent beat this process should cost 5 crdits!!!!!!! PLUS WHAT ARE WE GONNA DO WITH 5 CREDITS ANYWAY??!!?

tidal wharf
soft oasis
#

Can we have stereo audio?

green jacinth
#

y a t'il un moyen de traduire en francais tous ces messages

tidal wharf
golden moat
delicate cedar
#

And there are DAWs which feature the ability in their plugins.

robust jewel
neat gorge
#

Has anyone already suggested the ability to dictate song composition, along with the ability to control what instruments are being used, getting stems, maybe creating parts before a song, or in between another area. Could even go as far as creating a simple daw like interface as well over time.

#

This can go very far.

ornate wren
#

I am enjoying Chirp. I've noticed something while having it compose several songs: it seems that my preference for a song is often determined early on, either in the intro or at the beginning of the song.

Instead of having two options for an 80-second clip, I believe it would be more satisfying to have four options for a 30-second clip (with the same credit usage) or two options for a 30-second clip (using half the credit).

If I come across a clip that I like, I would prefer the option to 'Continue from This Clip', adding another 30 seconds to it. This approach, I believe, is a shortcut to creating a more satisfying song.

I read somewhere about a future plan to add a feature for restarting a song from the middle. Considering efficient use of AI resources, could you provide a way to specify shorter time spans in Chirp?

Thank you for the wonderful service.

tidal wharf
tidal wharf
# golden moat yeah, thats not great solution realy, not for pro use

The model doesn’t generate stems and having it generate those really means rewriting everything, and in a way that would require an enormous amount of pre-processing (eg you’ll have to get gigabytes of training data for stem-split songs). Not saying this is impossible but I doubt it’s anywhere near-future-possible. For now the model can only do the master track.

(Disclaimer: I’m not on the team and speaking just from my ML experience.)

frozen iron
#

In the library, instead of just a filter for "Liked" could we maybe get a pulldown that has "No filter, Liked, Disliked, Unrated" ... a lot of times I don't click like or dislike on a clip as I'm waiting to decide if something better comes along and a little more flexibility in the filter would be nice...

median coyote
soft oasis
delicate cedar
robust jewel
#

The quality should be better, in stereo, and the clips should be up to twice the length

soft oasis
soft oasis
bold wren
#

wish there was a channel here for song requests
that anyone willing could go and make some for people

bold wren
safe sand
#

Delete Track Button, Delete Track Button, Delete Track Button, That is all!

brittle birch
#

Separating the voice from the instruments

lost eagle
#

Feature request; upload your own "instrumentals" and get Suno AI sing/rap on it! 😊

(It could have potential for creating vocal, but I am actually thinking more for "toplining", ie. finding catchy melodies on the beat. Then writing and recording with a real vocalist later.
In that case it would be nice to give it the ability to just use jibberish words freely to have more freedom on the melodic rhythm.)

old nimbus
delicate cedar
ocean viper
# old nimbus HOW DID YOU DO IT WITH 10 MINUTES?

I think he didn’t, but it’s possible anyway. Yesterday I was making for 16 hours to spent all credits 😀 You can optimise work, just use multiple windows with several paid plans same time in each + Custom GPT to enhance/rewrite lyrics and advanced meta tags, to make 10 min time -> 1 min hyper quality music or 10 min time -> 5 min medium quality music.

delicate cedar
#

It absolutely is possible to do a ten minute song. You just need to be sure you consider your lyrics and your music to avoid too much degragation over the minutes. And being willing to try, try again when there are issues.

#

Note: The names were because I was trying to get a duet. It didn't quite work.

#

Oh and the line is actually hinky dinky but when I tried to use it it generated errors.

compact copper
#

can i generate in suno more length music like 2 at 5 min?

robust jewel
#

Have you read the thing literally right above your question?

#

As for how, on the website you can do it by clicking the button on the right side with three dots and selecting "Continue from this clip" and putting any lyrics/instructions for the next section of the song. Continue doing that until you have all the parts finished, then click the three dots and select "Get whole song." If you are doing it on Discord (I strongly suggest you use the website), you can choose Continue 1 or 2 to select which clip to continue from, then select the heart for 1 or 2 to get the whole thing.

delicate cedar
soft oasis
#

How about be able to make mono audio stereo? (like be able to add stereo to a V1 sample)

delicate cedar
#

Audacity can do it and is free.

soft oasis
#

Are you going to make Bark avaliable for other servers, similar to Chirp V1?

golden hull
#

Please add a “report” button for when the system refuse to generate but it’s obvious that it’s a wrong decision.

placid berry
#

Please add the ability in chirp to sync the execution of each word in the lyrics and chorus based upon tempo bpm and allow for the lyrics to follow an uploaded melody

golden hull
#

Record a snippet of my own voice - to be used in generations. That would be AWESOME!

urban ermine
#

Hi guys any plans on releasing a text to speech controlled generation model?

If you have cracked music you can surely crack highly realistic speech generation as well.

#

@sterile matrix

robust jewel
summer crag
idle ember
#

Hello again, Happy New Year. I hadn't used Suno in a couple of weeks after the big Christmas change. I see that there are changes in the management of the library, I like the change that indicates what part (fragment) of the song it is. I still think that the library would be managed better if the songs were nested, and above all, for your sake and ours, we need the DELETE option

#

And another thing I suggest for managing the songs in the library: The new songs should be saved on the new pages, not the other way around, let me explain, the first song should always be on the first page, that way it's much easier Browse and find songs. For example, if I have a song located on page 5, it should always be on this page regardless of new entries

#

If you think that visually it is better to see the new songs first, no problem, I think that is correct, but we would be on the last page, and page 1 is always the oldest

zinc gyro
#

Suno stands out for having opened up possibilities to allow frustrated poets (like me) to see their creations become songs. On the other hand, the rules of use could make Suno unviable when competitors (in this magical and surprising world) create similar open source versions. It's unreasonable for Suno to stipulate in its rules that everything created using free accounts is the property of Suno, so that it can do with it as it pleases. And the creator can't even use it commercially. This has nothing to do with making the project "sustainable". It is, in my opinion, a predatory rule. Humbly, on behalf of this fabulous project, I suggest that this rule be revised. Even if tomorrow (when Suno succeeds and brings the quality above 128k) I decide to go pro, I can't claim what I've created as free. That's unreasonable.

worldly valve
#

Can you please add a filter in library to show only “Full Song”? Thanks.

#

Also, it seems like the overall volume in the full song is a lot lower than individual clips, if there is a way that the full song has the same volume as the individual clips that would be great. Thanks.

worldly sphinx
#

is there (or is it planned/possible even) something analogue to img2img for sound? Like for example providing a music clip and adding vocals to it (like "inpainting" does for images)?

humble lynx
cyan sleet
#

i was wondering how to make 45 seconds of audio and which ends perfectly.

robust schooner
#

Would be great if you had some kind of non-subscription way to use this service. Like simply buying credits. I just cancelled my subscription, not because I don't want to use Suno, but because I simply don't need 2500 credits a month. I need 2500 credits that I can use when I need/want to

silk pebble
#

I would love for a possibility to create AI bands that make somewhat coherent music (at least the style) and vocals. For example I managed to get this awesome wide stereo voice in a chirp I'd love to reuse, making like a virtual band. In short "Design your AI Band"

silk pebble
# robust schooner Would be great if you had some kind of non-subscription way to use this service....

How many Credits would you be planning on using? I know it requires a paid plan to access, but 1000 credits run at $8, which is only $2 less for 2500 credits. Still, you should get to keep them at the end of the cycle. Also, I don't know if it matters but I think the terms say you can't make any revenue on content made from a Free plan. Which is like, if you had bought credits (while in Pro), then it rolls over to Free, and you use those stored credits to make a song, I assume you can't make any revenue from it (ads etc).

It's also something I had been thinking about earlier.

image.png
delicate cedar
silk pebble
delicate cedar
#

Sure, sure.

But I was thinking of something really far off in the future. Like a character creator in a game, that is crossed with AI in general to assist you in building an entire back story for a group, designing their look, designing their themes. It'd probably be some expensive $200 or $400 package but it would be a high end thing that would allow you to create the whole thing. Think like Gorillaz. It'd plug in your characters into an image generation tool, assist you with rendering their video performances...

Obviously something that wouldn't be available for a while. Maybe common place in a decade or more. But this is where it begins. With making music with AI...

robust schooner
oak echo
#

I have an exciting idea for a website and Discord integration called Suno Radio. It's a platform where people can come together and listen to endless music, with the ability to vote on songs and request custom songs for premium members.

Key Features:

  • Endless Music: Suno Radio generates music on the fly, creating a continuous and seamless listening experience. The music is generated while the song is playing, ensuring that it never ends.
  • Voting System: Users can vote on songs to influence the playlist. The more votes a song receives, the higher its chances of being played.
  • Voice Chat Integration: Suno Radio can be integrated with Discord voice chats, allowing users to listen to the music together in real-time. This creates a sense of community and shared musical enjoyment.
  • Custom Song Requests: Premium members can request custom songs to be generated by the platform. This allows them to enjoy their favorite songs or discover new music tailored to their preferences.

Suno Radio is a unique and innovative way for people to enjoy music together. It combines the social aspect of music listening with the convenience of endless, generated music. Whether you're a music lover looking for new tunes or just want to hang out with friends while listening to great music, Suno Radio has something for everyone

delicate cedar
#

It's going to take some work, but I think that's the eventual goal of it.

#

(Note: this is not inside knowledge, just speculation.)

zinc gyro
cinder blaze
main saddle
#

I'd like to redo a song with the same melody/music as before but with tweaked lyrics to better fit the melody it came up with. It takes alot of work to get lyrics to perfectly match a melody for a good song, this would allow iterations on the same melody to get a more fitting lyric.

zinc gyro
hybrid crane
cinder blaze
hybrid crane
#

Yes... I mean Suno 🙂

cinder blaze
#

in fact i basically only vc using a custom rvc model

#

ill have to mess with it this weekend and see where i get

#

ive never messed with bark (im actually not 100% sure what it is...) but since it runs local i imagine you can get higher quality from it? might be able to get the voice from there to put into the music

hybrid crane
#

I did.. that's why I said it... also tried it reversed.. take vocal track after stripping and try to add over a different voice.. it will have trouble with pronunciation.

cinder blaze
#

hm. with pronunciation sounds like it could be sample size. could probably make a song utilizing most phonemes a couple times to help that out a bit

hybrid crane
#

Some voices have too much effect on it, and are damaged with stripping, giving you realy weird results.. at least it did for me

high lark
#

I think it would be a good feature to be able to create more of a song before it starts, so like continuation but feeds into the start of the current song rather than continues after it

cold osprey
#

Hi sir can i create sinhala voice song , your website can create but discord its not working can you fix it?

torpid jettyBOT
frail widget
#

Hello everyone, tell me when I write [intro], and I want the loss to be 10-20-30-40 seconds? How do I do this? I write an [intro], and he makes music for 1 minute and 20 seconds, but I don't need that much, or if I write a [verse] right away, then he immediately starts performing it

autumn token
#

It seems that pansori, the traditional Korean music, does not work. Could you please add it? https://youtu.be/8Kt7YdXsWzg

YouTube
Great Big Story
Pansori: South Korea’s Authentic Musical Storytelling

They say that the finest work of pansori can make your heart stop beating. Pansori is a traditional South Korean form of storytelling, operatic in style, that relies on the authenticity, intensity and power of the singer’s voice to express stories of love, death, tradition, honor. A perfect performance does not strive for purity or perfection of...

▶ Play video
safe sand
#

Delete Track Button, Delete Track Button, Delete Track Button, That is all!

silk pebble
#

Not very intuitive but...

lost eagle
#

Feature request: continue on user uploaded content, would be absolutely fantastic!! 🤩

ionic wasp
#

320 Kbps mp3🙏

high chasm
#

I always dream of the ability to feed in a part of a real song and have the ai continue the song in its own way. That'd be so cool!

quick apex
#

Would be so nice to first generate the sound track (similar to AIVA), then be able to select a singer voice. Essentially more control over what gets generated, and especially be able to reliably repeat that same style

ionic wasp
#

It would be nice if audio can dowload flac format

echo flower
#

Someone said it above, but a filter for full songs only would be nice, and relatively simple compared to other requests.

soft oasis
steep holly
#

(not sure if possible but) would be cool if
Vocal and Instrumental were saved as separate tracks;
New tool to "sing over this instrumental" so we can change the voice params on a song, or arrange duets, backing vocals etc.
(Or even "play new instruments under this vocal" so we can switch that up)
Would help with edit and tweak, and also get around the missing vocals which sometimes occur (because we can just sing over a good instrumental result)

raw portal
raw portal
#

Wishes:

  1. Generation two suggested clips, often leave me a cool version I want to use, but then example the text pronounce wrong and I have to change some letters to get it right: I wish there was a option to "make same clip, with new text".

  2. The same as a bow, if the clip are , let's say 1:20, but the last 10 sec are "terrible", I wish there was a option to select a specific part, example: "continue from: 1:09" , or/and "skip 0:11-0:16" (some thing that makes it possible to clip long boring parts where the generated clip have a 1:20 long "self decided solo" and you want to cut it down because the rest of the clip are the cool part.
    Example are on this song I made: https://www.facebook.com/olejohanaanestad/videos/1310929436275376
    From 2:51 it start a "solo" that are "endless", and from 3:29 it start to sing something that not gives any meaning (sound like "eta gas, eta gas, eta gas...") - I would like to cut that part, but at same time the rest was perfect.
    (I know I could edit it in another program, but it would been nice if Suno had the option to "regenerate same "without a defined part")

  3. It happens often I would select a other combination of the #clip 1, #clip 2, #clip 3, etc....sometimes I even want to use clips of the other suggested versions to complete a song, so it would been nice if it was possible to select the clips to be combined and to decide the order myself.

  4. And of course, a option to choose male, female, duo, choir, etc...for the vocal - if there was a option to ad commands in the lyric field, it have been super cool.

Morten Aske, .....du ba om det :-D (De tar litt av i slutten, vet i...

Morten Aske, .....du ba om det :-D
(De tar litt av i slutten, vet ikkje helit ka det e det synge, gas, gas, gas...vet ikkje)

safe sand
# silk pebble Not very intuitive but...

Well it doesn't work now. I contacted these guys via email and they said they are making a delete button that should be done by now, and it still isn't. Extremely annoying to have pages and pages of song starts that weren't much good. It's a cluttered mess. You would think they would have had a delete button from the start.

stuck quiver
#

Suggestion/ After generating a song, there should be an option to separate each and every instrument, similar to Bandlab. I want to edit them in my DAW.

rich panther
#

I'm using Suno and I want do create a instrumental music with no voice, someone knows what prompt should I use to create this type of music?

delicate cedar
# rich panther I'm using Suno and I want do create a instrumental music with no voice, someone ...

You may find this instructive in how to get what you want: https://suno-ai-wiki.notion.site/How-do-I-make-an-Instrumental-352abcb0bf6a4f77b8fcf76a768f18af

Suno AI Wiki Notion on Notion
How do I make an Instrumental? | Built with Notion

Chirp can add an [Instrumental Break] between lyrics, or can create a whole song without words or singing. Depending on your genre, you may need to encourage Chirp to sing, or instruct Chirp not to sing. Here are some tips to help.

potent crane
#

This is a tip rather than suggestion - be wary of your lyrics being mis-pronounced, recently I had "live" (as in live a life) pronounced lyve (as in "I went to a live gig") and it mucked up the rhyme for the next line, also it pronounced "sympathise" wrongly as sympa-thees (not as in thighs), it might be useful to read through your words beforehand to try and spot any potential problems.

neon plume
delicate cedar
#

Question. Isn't that basically what the video download version does?

#

I mean it'd be nice if it did a thing where it highlighted the word being sung at the time it was being sung but effectively the words are there to follow.

hearty goblet
#

When will there be a better sort or filter option in my library?
i want to not see any songs I disliked... maybe filter for full songs, etc

potent crane
marsh thistle
#

Add and anime Style!!! it would be the best!!!!!! If somone was making an anime but didnt know how to make a full on intro this could make their dreams come true!!!

prime rock
#

I just got a five second continue which refunded my credits AND CUT OFF IN THE MIDDLE OF A WORD what sense does that make? The whole thing with ending in the middle of words costs credits and is exhausting to work with

prime rock
sullen garnet
#

Import to DAW

marsh thistle
#

it should cost 5 credits instead of 10 every song along with every version you end up not liking one anyway

keen solar
#

Improved audio quality /resolution with less artifacts

robust jewel
#

Yeah, I don't think you really need to suggest that one. Pretty sure they're going to already have been working on that, literally from the moment they started.

thick meadow
#

can somebody from SUNO start a room where ONLY full and finished songs can be posted please? the quality should be HD and mixed too. i.e. songs posted in there that are spotify ready

#

obviously that have been made on suno, like full 3 min songs.

#

then if there are any coders, who fancy a big one, please code a spotify for AI music, with a view to allowing conventional music on there in the future, whoever did this will make a lot of money and you will be bought out for 100x 12 months from now

thick meadow
ornate wren
#

I am generating 10 songs every day.
Each song has the same lyrics and style but slightly different titles,
distinguished only by the date included in them.

Two days ago (01/04), 8 'Good' ratings,
yesterday (01/05) only 1 'Good' and 9 'Bad',
and today 7 'Good' ratings.

I am curious to know if the date included in the song title significantly impacts the seed used for generation.
If possible, I would appreciate more detailed information on how the seed influences the generation process and what specific seeds are being used.

urban ermine
delicate cedar
thick meadow
#

Agree, made a couple of videos too, not great but still I bet theres creative people making superb videos & I wanna see them!

#

Please if there are any coders out there, think about the arduous task of coding a spotify for AI music, with a better payout model. You will make a tremedous amount of money, and you will sell for 100x in 12 months imo. Our coding team is busy, so we can't do it at the moment, we are knees deep in other projects, but mark my words.

#

If anybody does this, I will help with marketing, I can produce professional marketing content pefect for social media, with music of course ☺️ thanks to the new Frank Bruno, SUNO 🔥

delicate cedar
#

I honestly suspect the coding is the easy part. Hosting, storage, having contracts set up and bandwidth to go how it's needed, mobile apps to make it ubiquitous....oh and money to do payouts on. And partners for advertising. And a system for subscriptions. And advertising to get both AI music makers and to get word of the app existing out there...

thick meadow
#

Our coders had a look, according to them, who speak many languages, its a bit of a task... you're bang on though, about the rest I agree but can you imagine a new spotify? We can have a war with them, go toe to toe, what a terrific adrenaline rush this would be

#

we are reaching the stage where the music is competitive imo some beautiful sound coming out from people

delicate cedar
#

Useful to anyone seriously considering making a go at this: https://github.com/GorvGoyl/Clone-Wars

Just do a CTRL+F and search Spotify. No need to reinvent the wheel if the work is mostly done for you. Granted you will need to adjust the look and feel away from the original if you don't want to get in trouble over any look and feel patents. But it is not nearly as dificult as you might imagine to get started....

...says the non-programmer.

GitHub
GitHub - GorvGoyl/Clone-Wars: 100+ open-source clones of popular si...

100+ open-source clones of popular sites like Airbnb, Amazon, Instagram, Netflix, Tiktok, Spotify, Whatsapp, Youtube etc. See source code, demo links, tech stack, github stars. - GitHub - GorvGoyl/...

#

...errr obviously select one that has a compatible license with what you want to do and obey that license. You'll have to make sure you open source your changes if it's an open source license and all and so on.

dusk lintel
#

how i install teach me pls

hybrid crane
pulsar flame
#

When I download a clip or song from SUNO to my computer, it should embed/add the link to the properties section.

Maybe add link to comment section. So I can copy the link from there and paste in the address bar of my browser to go straight to the clip/song instead of going through my countless LIKED clips on SUNO website.

image.png
#

Categories for clips on the website as suggested by others would also be helpful.

Just like the Liked section/category add few more options or make the user able to create custom category.

soft oasis
#

Also make it so it makes the title of the file the actual song name instead of "5357415b-a202-4798-af1b-3bea6e8939bd.mp3" and the artist name could be "Suno" or "Suno AI". Also include the year maybe

fluid nova
#

How do I get the best results when using the Continue option?
What is more appropriate for me to write for Part 2?
İnfo : I have a subscription on website

cinder blaze
#

basically just the next 8 lines

ornate wren
#

I want an additional feature for the Custom Mode in Create that allows users to specify a seed value. I'm feeling incredibly frustrated because I made a typo in the lyrics input and now whenever I listen to the song that I very, very, very much liked, it bothers me a lot.
If it were possible to specify the same seed value along with the correct lyrics, I should be able to create the wonderful clip that I truly desire. However, with the current system, I am unable to do so.

solar prism
#

I really need a better way to sort the clips. We need a "Maybee" button as well as having the Yes and No

#

Would be great to be able to hide the refunded credits

foggy verge
#

Hello Happy new year to everyone. Is It possibile export midi file too?

solar prism
#

Being able to download a upscaled of the song image would be pretty cool.

#

Perhaps when you click full song, it upscales the image created for the song

#

A quick button so you dont have to click through alot of pages. Im currently on my page 54 😄

delicate cedar
#

If we're talking about customizable song images, the ability to select one of multiple images from the chain of parts would be nice. If I am successful at getting an outro it always goes for the generic chirp bird when the first or second image was more meaningful and nicer to look at.

solar prism
#

Aye, that'd be good

#

I saved them all from in a folder. Gonna try upscale them, though some images seem to be same

autumn pilot
#

Songs without name or lyrics should still download with an ID of some sort, before it did use to but then it stopped .

#

Now it just says mp3

drowsy hinge
#

So guys, when API ? 🙂 I just need the simple one shot custom mode

woven merlin
#

Ending a song is a too difficult task. It takes too many credits to find a decent ending. Some of the pieces have hallucinations, some repeating the same note again and again, and some just can't stop. In some cases, the pieces, which are not credited, are the best.

next jungle
#

When I "Continue from this track" to do Verse 2, it will not use the same melody. Any way to get it to?

snow coral
#

If you add a vocaloid to the style, sometimes you get a voice similar to Hatsune Miku. True, then it is difficult to control the genre; other words in the style are almost always ignored if there is a vocaloid. It would be cool to set the voice separately, and the genre of music separately. I am very attracted to the topic with Vocaloids and especially with the Russian language, it will be cool if this is improved.
I don’t use the service now, but just to support you, I earlier bought once a subscription for one month. You are cool.
I'm waiting for new updates.❤️

delicate cedar
mystic zodiac
#

New to Suno and it took some time to get the vocal i wanted, the feel, and the instruments and style. it would be powerful to be able to use that track as a reference for new songs that would use the same vocal, instruments, style, and feel. Either that or a drop box to select the vocal, instruments and such rather than just random.

covert oar
mystic zodiac
#

A great suggestion would also to be able to download each part as a stem for reworking in a studio like Cubase and such. Vocal, Bass and such.

echo flower
mystic zodiac
#

The model needs to pull certain instruments and a vocal from a database at random before processing them together

#

Then just a case of splitting out each part. VirtualDJ does that in real-time. not hard but would be handy here on download.

potent crane
faint peak
#

Is there any way to Subscrip on Suno "Pro Plan" for someone who is from germany?
I was thinking to use PayPal or are there any other way Suno gift card, amazon or anything else? is that okay if i ask here? - since like 99,9% (1:1000) of all people in EU doesnt use chose payment method at all (with the ecxeption of swiss).

thick meadow
thick meadow
mystic zodiac
solar prism
#

If you want to continue on a song and music is playing at the same time and decide to switch, you will end up with the wrong song

wary hazel
#

Pleas Help: hey, how to write that verse 1, 2, 3 should be sung by a man and the chorus by a woman? And there was a choir singing in the background?

dawn fjord
#

Need some kind of map to find things including a tree of connected parts. Getting hard to find stuff.

potent crane
#

When creating many parts or attempting the same part again you are faced with sometimes dozens of "songs" all with the same title and it just becomes impossible to remember where you are. The ability to rename parts on the fly would be a godsend, at least duplicate titles should default to a number (2), (3) and so on, to make it easier to identify.

cinder blaze
potent crane
#

No, I mean in the column where the parts are listed

wind crest
#

is there any way to erase and reroll just a part of a generation? i keep getting ones where the verse is terrible but the chorus is great or vice versa, or where in general it's great but there's a chunk where it gets stuck on a single note for multiple bars, or mangles pronouncing a word.

potent crane
wind crest
#

mhm i suppose that's where this tech will really excel at this point, as a springboard for a human to adjust and complete it

potent crane
soft oasis
#

Instrumentals in V2

golden crow
#

I have a question: How can I extend the duration of music creation or identify the appropriate subscription package for this requirement? Currently, with the subscription I have, I can only generate music with a maximum duration of one minute and 30 seconds. I would appreciate guidance on how to extend the music creation time. Can you assist me with that, please?

delicate cedar
sullen vale
#

Audio sample input

Like audio 2 audio

If text could be in the mix even better

Text + audio 2 audio

This is the best music generation tool We’ve seen to date.

Heavily testing has possible applications in CC.

safe sand
#

There is a few annoying "static clicks" on numerous tracks. Not all, but some and it sounds so unprofessional. Keep an ear out for them people!!!!

dense tinsel
#

How about a search feature? I want to explore the other prompts that the community tried, not just limited to the explore page.

potent crane
potent crane
molten egret
#

My top requested Features

#1 Duality (Able to produce two or more different voices for Duet Songs)

#2 Children's Voices (This could work for eerie songs like a FNAF Song or add extra layers of innocence)

#3 OST to vocals (Where you upload a Soundtrack and The AI figures out the best way to produce and Match vocals to that soundtrack, mostly thinking about Undertale and video game osts in general)

#4 (being trained and able to produce Disney like songs in terms of quality, especially Villain songs like Be Prepared, Ready As I'll Ever Be, Friends On The Other Side, etc)

rotund orbit
#

Hey guys. Okay, so I've been using suno, attempting to make some fairly serious art with it. And I've been having some really encouraging results doing it. Thing is, I'm finding that I'm spending a lot of time in the online version of the app. I've rendered thousands of individual tracks for hundreds of songs that I've been making with it.

This poses one big problem: scale.

The current web app is great for vibes or light usage, but when you've got a lot going on with it, it's really hard to stay organized.
I realize I'm probably an edge case, but I really like this app. I find incredibly satisfying psychologically and creatively, and I would like more options to organize my projects.

Just a thought.
Love you bunches.

delicate cedar
#

signed

rotund orbit
#

Another thought: Musical intros. The app changed a couple of weeks ago. And we don't get those 19 second musical intros anymore. That's positive, but now it's going straight into the song. Could we maybe get like a 3-6 second spot at the beginning of the track before the voice kicks in? Or, if it doesn't exist already, maybe a tag that adds one? 🙂

delicate cedar
rotund orbit
#

Also, kudos on fixing the moderation system.

violet citrus
#

Hey all, some suggestions from myself,

1- A drop down menu for genre, with each one having a 5/10 second snippet of an example of that genre, then also a section for other/custom so we can still type whatever genre we want.

2- A way to define a melody for each section. For example if I want a verse to have the same melody it would be nice to select that part of the track and say "match this melody"

3- A way to specify how long intros to songs should be, or even a way to add something to the start of a song. So if we get a killer song and want to add an intro we can specify whether to genere something for the start of clip or the end of the clip/

4- A way to edit the lyrics in a song we've generated while maintaining the melody would be great.

5- I know this feature is coming, but longer songs, and also clips that don't cut off your lyrics. A little warning box might be nice. Something like a red line through the lyric box that says "Anything after this point might get cut off"

6- A slider to specify the length of time for a clip from say "10 seconds" -> "2 minutes" -> "Decide for me"

Think that's it for now. Big shoutout to the Suno team. This AI is insane. Absolutely in love and can't wait to see what's coming up.

delicate cedar
thick meadow
thick meadow
woven merlin
#

The sound quality is exceptionally crap tonight. What happened? Also, hallucinations exist more than usual. He, guys, I'm not looking for the speed for song creation but better quality.

neat gorge
#

If we could control compositional elements like intro, verse, outro, breakdown, bridge, etc. That would be cool.

golden crow
# delicate cedar https://suno-ai.notion.site/b72601b96de44e5cacd2cd6baa985448?pvs=25#baf0eb82963f...

Hello, thank you for your response. I have an issue when starting a song, as the beginning sounds too abrupt. Is there a method or technique to make the song start more smoothly and end gradually? Are there any adjustments I can make to prevent the vocal performance from starting too quickly, opting for a softer beginning and ending? Currently, the song starts and ends abruptly, sometimes cutting off the complete reading of the song abruptly, affecting both the music and the voice. Can you suggest a solution for this issue?

delicate cedar
#

Which seems to help me but is in no way definitive or foolproof. At the end of the day an AI is going to do what it will do.

golden crow
ionic wasp
#

Would be cool feature if can cut off part off chirps. Sometimes clip sound cool but end of clip weird sound etc.

tidal wharf
neat gorge
#

Yeah but it doesn't always work

tidal wharf
#

That’s AI for you ¯_(ツ)_/¯

mystic zodiac
#

I'll add a request again. Not sure if these get read by anyone at Suno? The ability to add a reference song that has been created by Suno previously in your library in order to use the same vocal, instruments and feel for different songs to create multiple songs from the same band. Or, drop boxes to select the vocalist and instruments.

tidal wharf
mystic zodiac
#

The model needs to pull those aspects from a database and as a programmer myself, i imagine there would be code that pulls them at random to then go on to the creation algorithm.

#

Consistency has been and is an important factor in image and video ai creation and I think even more so here with music.

idle ember
#

Hello, I have a new suggestion: It would be very interesting if a part could be used as a seed for new songs.

mystic zodiac
#

Yes, that's a great idea as well as my suggestion above yours to be able to use a generated song as a reference.

wide belfry
mighty sun
#

I have a question if I have run out of credits for generating songs, do I have to buy more or will it be automatically unlocked?

delicate cedar
mystic zodiac
#

@delicate cedar Many ways it can be done and would only make Suno much more powerful IMO.

blissful current
# delicate cedar There is no public API access as of this time.

Sure, but... are there at least serious conversations about opening one up to the public at any point? Multiple people have requested it, and honestly it'd be so incredibly useful to me that I'd consider paying extra on top of my current subscription to just have API access...

delicate cedar
mystic zodiac
#

@delicate cedar Do you think there is a possibility of them adding more control like a seed or a previous rendering reference or similar in the future?

delicate cedar
main badger
#

chirp

torpid jettyBOT
austere quest
#

I suggest add a separate genre BRAZILIAN PHONK, because it cant generate it

safe sand
empty geyser
#

Agreement of transfer of exclusive rights of music by Suno AI to members (authors). I prefer that if Suno publishes a channel where the channel is dedicated for members to fill in the details and get a certificate of transfer of exclusive rights for commercialized songs or re-recording and re-composition for free, the birth of the song should be the basis for the agreement to support subscribers.

wide belfry
#

I'd appreciate the ability to create Acapella vocal tracks.

golden crow
wide belfry
#

Also, could we rename tracks after they're created. Sometimes, both tracks are worth pursuing but having the same name makes them difficult to keep track of.

delicate cedar
#

Quality of life suggestion: Add the ability to search your own lyrics in the webapp. I wanted to find some examples of things I tried previously that I liked and I have no idea how to do so. In the Discord version I could just search my DMs and find it.

topaz pike
#

Sound to music feature please. Kinda like image to image but audio instead. Basically you would import an audio file and it would generate a song from that audio. I’m probably not the first to ask about this but it would be cool

woven cradle
#

I would like a version of Remix where we could keep the current music and style but can just adjust the lyrics. From an implementation perspective it is a lot like Continue From This Clip but the remix becomes the new first clip.

lone roost
#

Would be amazing to accompany chirp model with an upscale model as in to improve quality for uncompressed product download. 👍

prisma comet
#

They could put a remix lyric option where you save the music original rhythm but get the option to modify the letter

zinc gyro
#

An interesting thing would be to respect the characteristics given in the style for voice and emotion. If I describe specific technical characteristics (timbre, mood, etc.) of a Frank Sinatra-type singer, it doesn't make sense to result in a Willie Nelson-style voice (and vice versa). These are singers with a "wildcard" voice that can fit into any style. One with a nasal, rebellious voice and the other with a sophisticated voice. They should work. A voice with similar characteristics.

Willie Nelson
country, Baritone, yodel, nasal, hoarse, soft, joy, love, naturalness, simplicity, nostalgia, rebellion, expressiveness

Frank Sinatra
Baritone, scat, soft calm and elegant timbre, charm, confidence, nostalgia, sophistication

#

Any suggestions for including a "vocal ornament" in the lyrics, such as "vibrato", "melisma", "yodel"?

elfin echo
#

[Build, Hook]

#

Build will give you that time before the text

#

Its better to use intro from personal experience, but it works maybe 30% of the time

#

like [Build, Intro]

#

The best thing for me so far is using it like this

#

[Intro: Casual talking]

Some text

[Refrain]

Some text

feral aspen
feral aspen
zinc gyro
wary hazel
feral aspen
hearty nest
#

I would love to see the ability to download separate stems for each song generated. For example, guitar, bass, keyboard/piano, drums, vocals, etc., might all be available as individual downloads at the time of generation. Are there any future plans for this?

mystic zodiac
#

I suggested the same for stems as well as the ability to use previous rendered songs as a reference to keep the same vocalist, instruments and feel for a new song. or a drop box to select vocalist, instruments and such.

hearty nest
#

Also great suggestions.

pure solstice
#

a form of "ai radio" where there is a playlist of public songs just playing nonstop

weak frost
elfin wedge
elfin wedge
#

Is Suno planing to add other payment methods? Im from EU where debit cards or paypal is usually preferred over credit cards.

mystic zodiac
#

i doubt the seed would be any good and just create the same one. we need to be able to use a previous one as a reference or have drop boxes for vocals and such

elfin wedge
mystic zodiac
#

ah. in AI a seed is what the model used to generate something

#

but yes. i and others keep suggesting that

#

it certainly would make suno much more powerful and separate them from the comp for sure.

elfin wedge
stray sparrow
#

Suggestion, the paid credits should be accumulative and permanent, if a user pays for a month and when the month ends the user still has a ton of credits left, leave them alone and don't steal the credits from the user account. Also if the user pays 2 months and when the first month ends the user still has credits left, just add the second month credits to the user account and don't simply reset the count. The current method you have is unfair and is like stealing the money of the users. I won't spend a dime until you change It.

mystic zodiac
#

i thought about that. so does that mean if you don't use all credits before rollover, you loose them?

stray sparrow
mystic zodiac
#

That's not good at all.

delicate cedar
#

Monthly credits only last month to month and are use it or lose it.

#

Extra credits which can be purchased separately and which are handled separately persist beyond the month to month.

stray sparrow
#

What I'm saying is Monthly credits are a scam, the expiring date is artificial, is an excuse to steal unused credits for users and make them pay another month.

#

I mean, I'm nobody to say how to run your business, but as user I think It's unfair.

olive loom
#

Hello there 🙂 I hope the devs look at this! I have some suggestions that would make Suno.ai so much better: #1. Add control over BPM #2. add control over the type of voice and pitch (Male, Female) #3. Add a way to select the length of generation #4. make the unused credits roll over to next month #5. Improve the use of the "[section]" commands for better structure of the generations

#

If you agree guys, please copy this so the devs can incorporate it into the service ❤️

delicate cedar
#

I have passed on your concerns to those who are in charge, and hopefully they can explain things a bit more. But as it is there are two tiers. One is the monthly allotment which is strictly use it or lose it. (Not an issue for me. 😅 I just about always run out by the end of the month on the Premiere plan.)

The other are extra credits which you can purchase with a paying account and these do not expire month to month and can be used only once your monthly allotment is used up. So they lay in store for when they're needed.

stray sparrow
#

I understand, but I think the system is not fair. I could agree with a system that lets you generate without limit's during a time, i.e. a month (I mean without a limits but following a fair use agreement to avoid overuse, misuse and reselling), or a system that lets you pay for credits that never expire. So If you use It a lot you can pay a mothly fee, and If you use It less you can purchase credits and spend them over a long time. But I think It's not fair to pay a monthly fee and have the disadvantages of the time constraints and the disadvantages of the credit system.

#

However, as I said, I'm only a user and I have no internal info so I don't know If the ends would meet If you follow my suggestion, I only can say how I feel with the current system.

potent crane
potent crane
mystic zodiac
#

@delicate cedar is there something in the web version that i can find which subscription i have? just checked account info and nothing there.

delicate cedar
stray sparrow
idle ember
balmy flicker
#

It would be nice to be able to put songs into folders for organization of each song and which to work on 😉

mystic zodiac
#

@idle ember A few as myself have suggested being able to use previously done song or part as a reference as you can with image creation or, to add drop boxes to select the vocal and instruments for consistency. I'm a programmer and not having seen their model but it makes sense that there is code to randomly select all of that from a database so make those a manual selection and the rest random. Would make Suno much more powerful.

robust jungle
#

I subscribed 2 weeks ago to premiere and I'm already almost out of coins (I have been for days actually), I'm just really having a lot of fun with it and find it useful for ideas when I'm composing) but buying new coins just doesn't seem worth it... Even if they last forever, the pricing for it is way too steep. I plan to integrate Suno into my workflow but this is legitimately holding me back. I wouldn't even mind paying double per month for double or 2.5x tokens beyond the premiere plan for example. Would it be possible for the prices of the credits you buy to be lowered or the credits you get to be increased? (See screenshot).

4000 tokens is less than half what I get a month, but it actually costs more... I really want to continue using Suno but I can't justify it to to buy them at this rate, I'd be out of them too quickly. Even if you double the credits you get per payment, that's still kind of excessive, I'd use them up, considering how many tracks I go to find ideas and gems that have decent quality audio for transcribing purposes. I digress, currently it just feels like poor value. I dunno.

image.png
#

Another suggestion I have is one for an API. It would be interesting to be able to develop a VST or something to fit my workflow more, or even a UI wrapper to drag songs directly into the DAW - automatically continue them etc. The more controls thing was already mentioned before I noticed but I'm going to +1 that

leaden marlin
#

Sometimes it just generates music instead of a song. I think there needs to be some sort of selection button that you can change to music or song.

woven cradle
gaunt prawn
#

i

#

I'd love to be able to take my own backing music drag into Suno and then experiment with it's vocal melody ideas over the top. You just tell it the key and the bpm. I reckon that would be a game changer.

prime rock
#

I often have a specific problem and I wonder if there’s a solution for it. When writing a song, if the chorus shifts tone very much from the first verse, it’s very difficult to get it to make a second verse, since it seems to only consider the last few seconds of the clip when continuing. If there was a change to the interface such that you could highlight a part of the song you wanted it to pay more attention to when continuing, that seems like it could help.

deep rain
# gaunt prawn I'd love to be able to take my own backing music drag into Suno and then experim...

That was also my thought. I write and play music, but don't sing. Suno can come up with solid melodies and vocals, but the backing music is not good audio quality. It's often difficult to hear clearly but a musician with professional tools and instruments can create good sounding tracks. If we could upload those tracks and allow Suno to create melodies, lyrics and vocals and I'd be very happy. Also, since the background tracks generated by Suno are very poor sound quality, but the ideas can be good, I'd like to be able to do the regular generations, download just the vocal track, and recreate the music tracks that were generated offline with better quality and insert the vocal tracks I downloaded.

orchid pond
#

can output like wav, i can contine edit in other music tool and have high quality ?

weak elk
#

no, only compressed

soft oasis
#

we want lossless

#

not a buncha lossy mp3s

weak elk
#

yes, we do. but there is none at this moment.

vital charm
tidal wharf
# prime rock I often have a specific problem and I wonder if there’s a solution for it. When ...

it definitely does not only consider the last few seconds, but a consistent verse 2 is a bugger indeed. One (non-ideal) solution I found is to make two separate continued generations, one continuing with verse 2 and the other one with the chorus, and mix them later in a DAW.

This assumes you only have verse 1 in the first generation, which is also hard to get. Hopefully soon the team will release "continue from anywhere," and it will become much easier and creatively empowering!

#

My suggestion to the team is to reconsider the approach to feeding lyrics to the model. Currently, when continuing, you're meant (I think) to input the lyrics for the new part only, leading to all kinds of artifacts, especially if the generation ended mid-lyrics.

Instead, I suggest that even when continuing, you're meant to enter the entire lyrics of the song, including the already-generated part. This will make it easier for the model to "tune" to the right moment (assuming it works similarly to how Jukebox did).

vital charm
# tidal wharf it definitely does not only consider the last few seconds, but a consistent vers...

"My suggestion to the team is to reconsider the approach to feeding lyrics to the model. Currently, when continuing, you're meant (I think) to input the lyrics for the new part only, leading to all kinds of artifacts, especially if the generation ended mid-lyrics." I have tried this. My problem is Suno.ai decided to make verse 1 again using a different tune. I think the only solution right now is to make all the verses in the same 1-minute 20-second song and then use an outside tool to remix everything together into a full song. But that takes away the functionality of Suno.AI with the understanding that it should be able to make a full song without any additional tools.

weak frost
vital charm
#

Well it sounds like they are working towards a solution which is a good thing.

#

🙂

tidal wharf
potent crane
#

It's really confusing that when you are in create mode the songs list from top to bottom, but when you switch to library, the songs are listed bottom to top.

#

It's annoying that you constantly get short clips, 10-20 seconds, yet you don't get credits for them, they are of little use and you should get credits for anything less that 20 seconds

potent crane
#

When creating multiple versions (I do 10 at a time most time) they title should default to title, title, etc instead of ten all with the same title, its frustrating trying to keep track of where you are with ten songs all with the same title

wary hazel
#

hey, I received a tip from you that if I want a woman and a man to sing in a song, I should do it separately, but how? this site is nice but lacks possibilities. When I type [woman] or something else, it's not a woman singing. How to make sure a woman sings?
How to make a song to be sung by a duet at a given moment?

delicate cedar
#

To answer your question, duets are something that is dificult to pull off right now. I wasted a bunch of credits last night trying to make one work myself.

The way I was told to do it is to do something like this:

[George] Some lyrics here
[Betty] Other lyrics here

#

I have sort of kind of managed it before but nowhere near consistely.

wary hazel
wary hazel
leaden marlin
#

Hi when is paypal ready for pay 🙂

knotty dust
# delicate cedar I have passed on your concerns to those who are in charge, and hopefully they ca...

I've been thinking on this for a while, and I'd like to suggest a potential compromise. 🤔

What if only accounts who allow their subscription to expire have their unused monthly creds burned, while users maintaining an active paid subscription could be exempt from the burn? So that as long as a user either has an active yearly subscription or continues to renew their monthly subscription, then their monthly credits get rolled over.

I think this could help better incentivize users to maintain their subscription, while also allowing Suno the benefits of burning some unused creds from users who cease their spending.

Personally, I'd be interested in purchasing a year of Premier soon, as long as the monthly creds rollover within that year.

Otherwise I'm reluctant to commit to a full year; since in order to avoid burnout I'll mostly likely end up wanting to take a break for at least a month, or maybe even 2 over the course of a full year. So would be nice to not have to worry about deadlines every month.

knotty dust
#

Also while I'm not familiar with the costs of running Suno, speaking strictly from a customer's perspective, if the above compromise were implemented then IMO the current monthly and yearly pricing for Premier seems fair.

Ever since the free accounts were switched to 50 creds per day, the Pro plan seems less appealing though.

IMO even with little or no change to the current price, Pro could still work out ok as a cheaper way to test out paid Suno, and as a way to drive people to the better value Premier.

However, would also be nice if it were reworded to be more clear upfront how upgrading means losing access to the free daily creds, meaning Pro only offers 1k more creds than the free plan.

Also, it seems problematic how after upgrading to Pro, spending the 2500 creds, then upgrading from Pro to Premier mid billing period; it only added +7,500 creds to my account instead of the full +10k.

I also noticed my billing date wasn't automatically extended from the upgrade. Fortunately I was pretty early into the month so didn't bother requesting a manual extension, but considering the credit burns this could be an issue if someone were to upgrade near the end of their billing period.

Thankfully, once contacted the support team did well in adding the missing 2500 creds without much delay, and was nice how they they threw in a few hundred surprise bonus creds. 🫶

gaunt prawn
cinder blaze
#

a thought for the continue clip feature. trim excess silence off of the clip thats being continued from. that should prevent generating long periods of silence, or the errored couple second clips

tidal wharf
jade bramble
#

im making lyrics for a song i need help what should i add to this
my head is filled with parisites
they tell me i am always right
they tell me to never go down without a fight
cause they control my life
yea my head is filled with parisites ya lucky im type to comply not the one to lie
sometimes they tell me im above the other guy but all they do is lie lie lie lie lie

robust jungle
meager snow
#

what should i type in music type for getting joyful party rhyme for kids

stray sparrow
cerulean flower
#

Can you put an intro in front of the song that is being generated?

hybrid crane
#

I paid for Midjourney for a whole year almost 350 dollar and ended up only using it a few months..

shell quest
#

Hi, I'm downloading as video and I notice the equalizer is gone, I would like it back. Also I think the title would be more useful than the style at the top. Finally the lyrics are written too small, they could easily be 1/3 bigger.

rotund orbit
#

How about a search function for the library? It's tough to slog through tens of thousands of takes

delicate cedar
#

Yes! Something that you can search metatags with or lyrics with!

rotund orbit
#

exactly!

#

Or even if all we got was song titles

#

that would work too

#

My library is like 600 pages long right now

light pawn
#

Amidst the tapestry of life, she emerged as a radiant bloom, navigating the seasons of growth with grace. From the tender buds of innocence to the full bloom of wisdom, her journey unfolds like petals unfurling in the gentle breeze of experience. Through challenges and triumphs, she painted her canvas with resilience, courage, and the vibrant hues of self-discovery. Each chapter tells a story of metamorphosis, and as the sun sets on one stage, it rises on another, casting a glow on the incredible tapestry of her evolving essence, blackground, 8K --niji 5 --style expressive

delicate cedar
weak elk
#

It's Sunojourney

zinc gyro
delicate cedar
#

[To the melody of Some Kind of Wonderful] 🎵 It’s some kind of journey 🎵

potent crane
#

I really don't see the point of the "video" - what purpose does it serve? I already have any lyrics printed out or on my computer, if for any reason I use the suno created lyrics then I would have to type them out to use in my studio, and it's not really a video is it? It's just a still image with words which could be offered as a jpeg rather than an mp4. A waste of bandwitch in my opinion, unless someone can convince of their usefulness.

mystic zodiac
#

Not sure but I would imagine the use might be that when shared here, people can see the structure that was used to further learn how to use it to achieve the results others are looking for? Just a guess.

fossil sail
#

hello I'm interested in making music. Suno AI processes the voice of artificial intelligence on the beat very well and adds beautiful functions to the sound. What if I trained my own voice, wrote lyrics, and made it sing with my artificial intelligence voice?

potent crane
fossil sail
potent crane
fossil sail
south geode
#

this place will steal your credits for bad outputs! I gotten output so bad that they came out in some alien language not from earth and they still take the credits that's the worst of it! you think with 2500 credits your be able to produce a few songs like more than 20 or so but your wrong as you have to keep generating song clips to get it right lie so many times! and then some songs don't even complete your shorts lyrics! and still they take your credits! what a nightmare this is! if i hit the thumbs down button it only removes the song clip but won't credit you back for being bad so your stuck with nothing at the end of it! you won't get your money worth! and that's the worst of this! they need to fix this mess!

robust jewel
#

It would be nice to be able to pin something to the Create tab so it doesn't get bumped off when you're trying to generate the next clip.

tidal wharf
# south geode this place will steal your credits for bad outputs! I gotten output so bad that ...

you were never supposed to, or told that you would, get your money back for bad generations. Just like any AI, this one is “as is” and “use at your own risk.” However low the percentage of good generations is (which, in my experienced it isn’t), you expected to factor that in in your expectations of how pricey using the app will be.

My experience is: I wrote and published four 30-minutes albums for twice the monthly Premiere price, i.e. one minute of final song “cost” me $0.5. That’s more than a generous offer if you ask me.

robust jewel
#

For those who intend to pay, but who are worried about wasting cash while still learning how to prompt, it's not a bad idea to use the credits you get on a free account to practice with for free.

leaden marlin
#

Please make some folders in the library per project.
At the moment I help me with suffix in the titles

E.g. to manage multiple versions and parts I managed it that way
Title - 1.V1 means Version 1. Verse 1
Title - 1.C1 means Version 1. Chorus1
Title - 1.V2 means Version 1. Verse 2

And so on

#

Perhaps some subfolder per version of project would help a lot. Have meantime 3 pages in library

coral harness
#

Suggestion: Add a "Add an Intro" to created song (not continuation). Often it starts a bit chopped off.

leaden marlin
#

perhaps you can do that by adding no text?

coral harness
#

Yeah, tried that already with an empty [Intro] but it is still happening.

tame ruin
leaden marlin
tame ruin
#

i dont know and i wont look 🙂

leaden marlin
#

Other suggestion: some intellisens in the text If you type [ than all possible commands appears

raven python
#

hey all, when Api access?

leaden marlin
#

But he add some vocals to it (makeing no sense)

tame ruin
tired yarrow
#

Suggestion: Would be amazing if we were able to make a FULL song in exchange for more points than usual

#

Maybe by selecting the duration as well

delicate cedar
#

You can make a full song. You just need to continue from part to part until the song is complete.

hardy arrow
#

Would appreciate a little longer length in order to do one verse and one chorus per song I have to make them pretty short. Also aj idea I had if you get a theme you really like it would he cool to produce new songs with the same seed.

I had sone luck picking an exact year and style don't know if it understands exact years but it hit the nail on the head

#

For instance this is 2007 rap hip hop and its right on point

#

Captures that Era perfectly.

faint ore
#

how to make a full song?

delicate cedar
hardy arrow
# faint ore how to make a full song?

choose continue from this clip and when u like a clip you like keep doing that then click get whole song (btw I use the website I feel it makes better music than the discord)

#

is it possible with the paid subscription to get unlimited relaxed generation? its something midjourney offers with their subs as well for art wand it would be nice since the free program gives you essentially only 1000 credits less just spread out ove the month

#

or sell the software that you can run localy on our own computers with a subscription or one time fee

#

and provide consistent updates to that

#

a lot fo people like running it on their own machines not relying on the cloud

mystic zodiac
#

@fossil sail you can use something like Bandlab to extract the vocal track from your suno song then use something like Replay RVC to change the vocal to your AI model, export and piece it back to the music in something like FL Studio.

left glacier
#

is there a way to control the bpm? song key?

cyan jetty
#

how to edit a song with specific seed

left glacier
#

any tip on how to get consistent variation?

zinc gyro
# cyan jetty how to edit a song with specific seed

That's not possible (yet). The closest thing is to use the "Continue" option using a new letter and keeping (or trying to keep) the "seed" of the voice, the style... and then edit by deleting the previous part.

zinc gyro
#

Or another more appropriate term

granite silo
#

is it possible to make a track without lyrics? just instrumentals, assuming there is a metatag for this

feral aspen
delicate cedar
# granite silo is it possible to make a track without lyrics? just instrumentals, assuming ther...

The unofficial community wiki also offers some advice on this. You can find that here:
https://suno-ai-wiki.notion.site/How-do-I-make-an-Instrumental-352abcb0bf6a4f77b8fcf76a768f18af

As always though experimentation is the key. Try new things, do stuff no one else is doing in order to create your own unique style.

Suno AI Wiki Notion on Notion
How do I make an Instrumental? | Notion

Chirp can add an [Instrumental Break] between lyrics, or can create a whole song without words or singing. Depending on your genre, you may need to encourage Chirp to sing, or instruct Chirp not to sing. Here are some tips to help.

granite lynx
#

Suggestion: being able to import a sample and let the AI work around it.
Suggestion 2: when a song is created, being able to download a .zip file with all the extracted individual samples (drums, melody, acapella, etc...) so that you could remove the items in the song that you didn't like and rebuild it using the first suggestion. I understand this might not even be possible with the AI model, but i'm curious about your thoughts.

wide belfry
pseudo surge
#

Can we get a volume control dial on the website? Always scaring myself when I go to play my generations and forget my computer volume is too loud

wide belfry
supple lake
#

Is it just me or did the quality sort of get boosted lately?

feral aspen
#

as far I know quality is 100% unchanged. Maybe you just get better at prompting...

potent crane
blissful current
cinder blaze
#

except [cowbell solo]

blissful current
cinder blaze
mystic zodiac
#

That's talent lol

cinder blaze
#

so much talent in fact that now south park exists

coral harness
mystic zodiac
#

The biggest suggestion more than consistency would be to fix the quality degradation from "Continue" and in general. Sadly Suno is unusable for me. I'll end my subscription and keep checking back when I can to see if it's as good as I think it can be. Good luck.

#

@cinder blaze My sense of humor 🙂

wet berry
#

#💡┃suggestions-web: it’s more like a complaint. Why do I get only instrumental music when I clearly ask for vocals??

remote cliff
#

Hello, I previously created a song using the Suna AI, and I was very pleased with the melody and vocals that resulted. Now, I want to create a new song using the same melody and vocals but with different lyrics. However, when I try to add new lyrics to this existing song by selecting 'continue from this clip', the system interprets the new lyrics as a continuation of the previous song and attempts to create a longer song. It perceives the new song lyrics as a continuation of the old song. My goal is to create an independent song with completely new lyrics, using the same music and vocals. How can I achieve this?

cyan umbra
#

I don't know if this is at all feasible, but it'd be really amazing if there was a "Continue reverse". I've gotten what I thought was a perfect first segment several times, but then had buyer's remorse on a word or a line as the song progresses or some audio distortion I didn't catch at first. If I could use segment #2 to regenerate segment #1 with minor alterations, it would be an amazing addition.

tidal wharf
remote cliff
potent crane
mystic zodiac
#

@remote cliff I've been asking for this. Either to use a previously rendered song as a reference or, have drop boxes to select the vocal, instruments and such but not sure if they actually look at this? At the moment I think the quality of a full song degrading after each "continue" is the most important feature.

cinder blaze
#

I think it'd be a good idea to provide the 50 free daily credits to paid subscribers who have ran out. is it because I'm greedy? yes. but I also think it'd prevent drops of usage, and people unsubscribing from the service due to being unable to use it through those empty periods (short of paying more than the subscription fee).

keeping up that consistent usage of 5 generations a day, would likely increase extra credits purchases as well since songs typically take more than that to get done, and when people get in a flow they're more likely to make a purchase. so it'd be a good idea on both a marketing and retainment standpoint.

worthy bloom
strange plume
#

I am sure I am not the first one, but whats about getting Stems instead of a mied song? And setting BPM by parameter?

novel pulsar
#

Hi guys, I just used up all my credits for pro plan, really suggesting to have at least some extra free credits distribute per day as it doesn't feel quite fair that free basic plans have 50 credits per day that makes up 1500 credits a month, pro plan is only providing the free credits at once with 1000 more credits... I am very interested to continue using Suno but with limited budget, therefore would love to voice out so that we can all enjoy using suno in daily basis as a solid paid supporter 🙂 This AI means a lot to me as it helps me to navigate & survive through difficult times.

tidal ridge
#

Hello, I would like to know if Suno has any partnership programs with websites. I would like to establish a partnership for mutual benefit, promoting Suno.ai in exchange for a pro account. Who could I talk to about this?

weak elk
#

dude, just pay for it.

potent crane
upbeat sable
#

we need better vocal quality

#

or offer stems so that again. we can get better vocal quality

wooden lodge
#

Would be cool to have ai upscaling with the service, i dunno if this is in the works

full lotus
#

We need to have a control over number of generations, may be from 1 to 5.

restive plume
#

It seems that the more tomes you "Continue from this clip" the worse the vocal quality gets. I'm sure this has been brought up, but it'd be even more amazing if we could get a fox for that.

full lotus
brittle birch
#

Split all the individual instruments as well as the vocals. This way, it keeps everything clean. Also, the sound quality needs improvement, especially the vocals. The voices are almost always distorted.

torpid lintel
#

Great work, developers! This is the best AI Audio program I have come across. My suggestion would be: To allow higher quality results with the use of more credits. I would prefer to use all of my credits for the day on one song where your system spends more time working on the song, resulting in a higher quality result. It could be accessed using a --q parameter like was used in MidJourney. Thank you for considering it!

drifting mica
final mural
#

Hi! You need to make the STUDIO VERSION urgent! Where you can have more control and edit each part of the song generated.

hardy arrow
karmic yacht
#

what if between verses and chorus and bridges and verses and basically all of that what if there's like a smoother transition because I've noticed that when it gets to a verse to chorus it kind of speeds up when you're doing rap and I feel like it would make it feel more realistic if it kind of Blended in a little bit more
😄

blissful current
# torpid lintel Great work, developers! This is the best AI Audio program I have come across. My...

I'm not an expert in Suno's models, but I don't think that's how it works. From my understanding, the --q parameter in Midjourney controls the number of iterations of diffusion; but Chirp isn't a diffusion model. It generates one token at a time and doesn't refine existing tokens the way diffusion models refine their images. So the only way to increase quality with Chirp's model is to either make it bigger and/or train it more.

Again, I'm not an expert, but that's my understanding at least.

grave oasis
#

How does the Free Credits work? It should be explained in the How-to-Chirp if Credits refresh every week or something.

tiny vortex
#

Can there be a top liked suno create song? I can see there is a liked-clips-feed, however i would like something like the trending list on the website. It would be awesome, if we also have submit your favourite ai song, and everyone on discord can vote, as an event.

sick wolf
#

There is any way to continu a music who didn't come from Suno ?

delicate cedar
restive plume
young mirage
#

Please touch up accessibility a bit. Live Region for when songs finish loading, headings or landmark for each generation, a way to quickly find newly generated songs.

river falcon
#

Can we add a pop section to the genres?

median coyote
# robust jungle I subscribed 2 weeks ago to premiere and I'm already almost out of coins (I have...

Yes, this has been discussed at length. In my opinion, it's totally backward right now, and this pricing model encourages users to simply create multiple accounts to get the best pricing, rather than rewarding existing users and encouraging them to purchase more credits and build their accounts.

Suno has acknowledged that this is a known issue they will be addressing, but I assume it's not a priority since most people don't seem to buy additional credits.

urban sleet
#

Can you please disable that awful feature by default? It's cool that we cannturn it off, but I can't imagine people prefer this to be turned on, and It's just driving me crazy all the time...

SmartSelect_20240120_024035_Samsung_Internet.jpg
zenith gorge
#

It is off by default?

torpid lintel
#

I have used the Continue function below one of my generations. It asks for Genre and Lyrics. When I proceed, I get something that is generated from scratch. I want to see if, when I use Continue, the program continues using the same song instead. Alternately, if you could offer an option where a user can generate a song that lasts longer and uses more credits to do so, that would work too. Thank you for considering this!

wide belfry
#

A tag to stop Suno from creating gibberish lyrics or reusing lyrics from same song would be amazing!

shrewd widget
#

A feature that analyzes a pre-existing song a user inputs and gives a style prompt that allows you to replicate the general feeling but not copy it specifically.

rapid sonnet
#

i need my Delete button i be spamming 💀

#

Improve the [Breakdown] [Bass Drop] [instrumental] Callout they rarely work for me in the Metal Genere.

stoic stone
#

Do you have an affiliate program?

woven merlin
#

What's wrong with the vocal sound? Most of the time, it's not clear; over-compressed definitely, but something else, too. a kind of crackle in vocal sound.

cloud wadi
#

Maybe a way to choose what instruments are use at certain points

onyx plaza
#

My personal suggestions are:

  1. to get a credit refunded for any clip we rate with the thumb down (a clip that, of course, you can’t use anymore);
  2. the “Rename” feature, to avoid to get confused when you create a lot of clips.
delicate cedar
cloud wadi
delicate cedar
#

(Assuming you are using custom mode)

cloud wadi
#

What's custom mode?

#

I use the discord version

delicate cedar
#

Oh!

#

Did you know there was a webapp version?

cloud wadi
#

No xD

delicate cedar
#

It's higher audio quality, has a longer time (first generation is up to 80 seconds, subsequent ones are as much as 60 seconds, and audio quality is higher)

#

Oh I see that I mentioned the higher audio quality. Did I mention the higher audio quality? 😅

cloud wadi
#

I'm subbed so will it connect?

delicate cedar
#

It should.

#

Just log into the same type that you subscribed with.

cloud wadi
#

How do I use a meta tag

delicate cedar
#

Here's a list: https://suno-ai.notion.site/Using-Metatags-34944efe09ec41d693e314f13af44695

(Make sure you're on the ones for Chirp and not the ones for Bark!)

Then put them in the lyrics section in brackets. EX.

Screenshot_at_2024-01-20_13-20-26.png
Suno on Notion
Using Metatags | Notion

Metatags guide prompts but aren’t verbalized themselves. Like stage directions or parentheticals in screenplays, you can use metatags to describe a specific part of your prompt (e.g. [laughter] and piano interlude) or to influence your entire prompt by describing speaker characteristics like [whispering] or music genres like rock.

cloud wadi
#

Oooh. Those things. I've just been doing [verse] or [End] And stuff

delicate cedar
#

Yeah. There's more that can be done in that manner.

#

The coolest thing?

No one really knows how far they go! So the only way to get stuff is to try it. And try it more than once if it doesn't work the first time. Because it might sing a metatag one time and call the music instrument or something else the next time. We just don't know. So experiment!

molten torrent
burnt igloo
#

There is a paid plan that will generate me a song over 1:20

#

?

mossy badge
#

melody only not make a song, what do I do ?

potent crane
steady harbor
#

I deposited 30 dollars... and it says that I have no credits on my account

#

This is a constant glitch, I’ve been using your service for a month now

craggy patrol
#

I would trade 1000 points for higher audio fidelity and NOT getting cut off at unpredictable times. Just sayin'. Suno is already amazing so we're just impatient. Stay the course.

sage wedge
#

When Sonos improved the quality previously, were songs created before the change able to be downloaded later with the improved quality?

limpid pollen
#

please add paypal and or other payment methods

wide belfry
#

I'd pay more for tracks that are 3-4 minutes long.

#

And that would accept more than just a verse or 2 at a time.

potent crane
rapid sonnet
delicate cedar
#

Lyrics: (c) Joeseph McCarthy 1919, Public Domain

mystic zodiac
#

@delicate cedar Depends on how many donuts are on hand lol

robust jewel
weak elk
#

indeed. I remember that old Mac parody where he says "It's not so much operating a computer as it is sort of tricking it - fooling it - into doing what it is you really want it to do. You kind of have to sneak up on a Mac. I don't feel I'm operating the Mac so much as I'm just there sharing the Mac experience, and if I can do something useful following the Mac's willing, so much the better."

idle ember
#

Hello everyone, I haven't written in a while. I make a summary for the devs of my suggestions:

  1. DELETE option
  2. Nested structure in the library
  3. The last song should be saved on the last page of the library, and the first song on the first page
  4. Any part of a song can be a seed for new songs, maintaining voice, style, rhythm, etc... (How many times do you like a part but we have to change the lyrics and start the song again?)
  5. If the seed is implemented, you can choose what to keep, style, rhythm, voice (keeping only the voice you can make an album of the same singer 🙂 )
shell yoke
#

Would love to be able to Create Playlists, Search for Keywords in Title and/or Genre and maybe an option to Check out profiles to find other Songs the same author has published (maybe a follow option for this as well to find back to all people whose music you liked more easily?)

deep ocean
#

My one and only suggestion (which is obvious and likely stated many times by now, and might even be impossible because I know nothing about the tech behind this service) is to make a way to export stems. Or at the very least just the vocals. I love the vocals. But I would like to re-record the music myself and use the vocals. But the vocals simply do not separate cleanly using any vocal remover/stem creator software/service. So I would love just the clean vocals from the tune. Hey one can wish. Suno is great and has been a real blessing to have. Keep up the great work!

tidal wharf
dapper jolt
#

Make Suno Free

misty flare
#

What do you do with the created content? Do you monetize it? If yes, how? I am curious, please tag me in the response

balmy zealot
#

Please ripoff the way that site Newgrounds’ Portal works for Suno Explore.
Why?
Theres a rate system and weekly rewards with tiers. But i like best that it is not curated or sorted. A true explore! Maybe only allow us to sort by language. What do you think of that for a starting point¿

wind crest
#

the ability to lock a specific album art to a multi-part generation
(eg i really like the image that generated with part 3, lock that in so that is the art used for the completed song)

twin cypress
#

I have a suggestion. Give me a seed option so I can work with a song with consistancy, let me be able to tell it in a prompt specifically what strings, chords and progression of notes I can use so I can continue. Currently I'm using a lame backwards way to jailbreak so I can do those things and while I get what I want more often than not, I'd rather be able to do so far more efficient. Example :

twin cypress
#

Also an option for HZ, like so I can tune things or at least guide it in a direction of tuning.

#

I hope if this idea catches you use the old french standard (435 Hz = A4)

sharp kernel
#

Hi team I would like to change my subscription form the yearly to a month charge I have emailed the team for a refund but have not heard back

#

Lots of bugs to be sorted out and not worth it for me at the moment

woven merlin
#

One thing I really like to have is ability to create more songs in a same style you already have a fancy song. So, a new text and a new composition, but the same vocalist, same band and same style.

median coyote
twin cypress
potent crane
wide belfry
idle ember
#

New suggestion

  1. Allow editing the lyrics of finished songs, not to change the lyrics in the song, but the saved lyrics attached to the song, let me explain:

Preliminary lyrics of my song:

Phrase 1
Phrase 2
Phrase 3
Phrase 4
Phrase 5

When creating the song, it only goes up to Phrase 3, I continue the song where it left off, adding again Phrase 4 and Phrase 5 (which have not been included in the first part of the song), and I also add the Phrase 6

Finally the lyrics attached to the song are:

Phrase 1
Phrase 2
Phrase 3
Phrase 4
Phrase 5
Phrase 4
Phrase 5
Phrase 6

Phrases 4 and 5 are repeated, although the song reflects the 6 correct phrases.

I hope I have explained myself well, thank you devs.

potent crane
wide belfry
robust lark
#

Seems others have the same or similar suggestion but ill say it again,

please add ability to alter lyrics but keep the same "seed" (same vocalist & beat) , this feature is pretty much necessary, would be 10x better outcome with this ability, its super frustrating without it.

ocean viper
neat gorge
#

Stems Plz

wind crest
rotund orbit
#

Okay, here's what I want. It's a quality of life upgrade. When you click the remix button, I want a radio that allows me to include the verse it's extended from if a verse exists. Like, it would only come up on an extended clip, like part 2 or part 3.

copper temple
#

Is there a way to Upload own Beats ? Would be insane to see an AI sing on ur own created stuff

split remnant
#

make a button where u can request that the song ends

#

a button to request female or male voice

feral aspen
onyx plaza
#

Another suggestion: they need to find a way to deal with "uneven" left credits.
Due to the "Credit refunded" mechanism, I only have 5 credits left: I can't use them (because you need at least 10 of them to be allowed to create a song) and yet they will expire at the end of my monthly subscription.
Ok, it's only 5 credits...but I've already had a lot of useless clips (totally silent, crazy sounds...) and that's unfair 🙁

stray sparrow
copper temple
#

We need continuation on the website too please

wide belfry
delicate cedar
urban sleet
#

Please make [Intro] work...
seriously, it works like 5% of the time, and this is the most basic feature and I'm sweating with anger most of the times when a good song starts out of nowhere and there's nothig to do to fix that.

idle ember
cinder blaze
delicate cedar
tidal wharf
#

@delicate cedar hey, is there any place to see the release notes for the latest update?

delicate cedar
tidal wharf
#

I see. Did they roll out “continue from anywhere”?

formal bear
#

unless if you want to volunteer to test...=p

cinder blaze
#

I do poi_hisalute

tidal wharf
storm cargo
#

I have a few suggestions for the website; they might have already been mentioned, but I'll highlight them again:

Include a 'Delete' button.
Implement sorting for full songs and group parts with the full song.
Add a search option for easier navigation.
Enhance the form to follow a standard song pattern like [intro][verse][chorus][outro] to facilitate the application in picking up required fields.
Consider adding a dropdown for selecting the style of the song."

storm cargo
#

I see now we have a move to Trash and Style feature.. Thank you guys.. Suno rocks

copper temple
copper temple
shell dune
formal bear
formal bear
potent crane
#

There should be a dialogue box included in each entry for notes - to be faced with a page of entries all with the same title makes it very difficult to locate the one you want without having to play through them all.

near solar
#

/template load name_or_id: gRdhrapAe3TV

edgy wren
#

would love to be able to input my existing audio; and have Suno add vocal melodies overtop!!

idle ember
#

I'm going to saturate the trash! Finally that option! THX devs!

idle ember
twin cypress
#

I like to actively song-write. It means I use a lot of credits. But I would like to take a song I'm working on and have a function to 'redo' it with the lyrics as I modify them without having a totally new generation and variance.

wide belfry
# copper temple It still got a max length then ig

1:00
+
1:10
+
0:50

3:00

What "Continue This Clip" does is put the next piece in line. I've had 8 parts in a single song.
Then once you have a good string of parts is "Get whole song" and Suno splices it together to become a 3 minute song.

supple lake
#

Is there a search that I am not finding or a workaround?

storm cargo
feral aspen
delicate cedar
swift cloud
#

Is it impossible to download lossless sound quality in WAV?

left glacier
#

does anyone have a solution for the tiny clipping sound that happens when you create a full song. it seems that where each clip is spliced together I can always hear a clipping sound which is ruining my songs. is any one else experiencing this?

terse thistle
# neat gorge # Stems Plz

This would be a huge step forward- in the meantime, in our lil apartment's experimentation, when we produced something we were really into we had good luck using a separate AI splitting tool (first one we found worked well https://vocalremover.org/splitter-ai) to get workable stems and import into a DAW!

terse thistle
# split remnant a button to request female or male voice

Through experimentation, I've found that adding "female" or "male" to the beginning of the genre accomplishes this pretty effectively, probably 9 times out of ten it properly generates a voice of that gender (e.g. "female trap" or "male country")

knotty condor
#

Please give us a way to either rename songs after they're generated, or to add a note to them after they're generated on the Suno website. Thanks!

feral aspen
agile badger
#

I understand your need. Here's a brief summary in English of what you're looking for:

I am seeking an AI solution that can analyze the audio of a story narration and automatically create corresponding background music. This AI should be able to detect the varying emotions and themes within the narration and compose a fitting musical score or sound design that enhances the storytelling experience.

#

AI make background music for elevenlabs voice narrations...

#

"Does anyone know of something close to this?"

#

"I have been trying to use the tags from suno.ai to control the voice and background, but this still does not work. Suno.ai is indeed intended for music, not for storytelling."

Please note that this is a translation of your request about using Suno.ai for storytelling purposes.

#

"What I have seen are tools that can generate tracks and sounds based on descriptions, but there is none capable of importing an audio, or even creating a harmonious background track along with the narration of the text."

atomic pond
#

Change - Error Message

"Couldn't generate that.
Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names."

Change - INCLUDE WIKI LINK - Because the message was irrelevant. It was overly explicit was the reason, after testing.

harsh musk
#

Also Is there a way to bulk donwload the songs ? For exemple all liked song?

#

Also add the possibility to add folders in library

vagrant rock
#

I would suggest to be able to get ownership (Commercial terms) for songs created in free mode retroactively by paying for the time the user has been active in free mode. For example when a user created an account on 14 november 2023 and decides on 26 january 2024 to want to get the ownership he should pay for that period (PRO/PREMIUM?)

raw cave
#

How about a toggle switch for either Start Song (A.I. prepped for gradual start or intro instead of jumping into the song), Continue Song (jump right in), or End Song (A.I. prepped for fade outs and ends). Also, a toggle for creations that are either Long (120 sec.) or Short (less than 30 sec.) for quicker starts or ends. Hopefully, the Short ones would only be 2.5 credits.

While I'm at it, a Seed slider (from 0 to 4) where a 0 would copy the seed from the previous clip and a 4 would adjust the seed for more variations. And a slider (from 1 to 7) for Instrumental-only (1) to Voice-only (7)...might help prevent blank lyrics from getting cut instead of continuing the song and/or it might nudge the A.I. when it's refusing to sing.

storm cargo
dense depot
#

Can suno do intelligent substitution of lyrics on the existing fixed songs?

deep imp
#

Can we have handpan instrument voice in Suno?

potent crane
delicate cedar
deep imp
round nebula
#

How about disabling censorship of lyrics? I might actually pay if that was possible. Also, the interface should have the option to choose different eras and genres of music within those eras so that you get a sense of the style that the song will be created in as opposed to using a text prompt that might not create the song in the style you desire, which is a waste of time/processing. Fix those two issues and I would pay for suno ai

potent crane
# round nebula How about disabling censorship of lyrics? I might actually pay if that was possi...

I don't the ability to swear in a song is a priority for the app tbh, I would rather see the option to rename songs instead of having 30+ parts all with the same name, I would like to see a dialogue box for notes next to each song creation, I would like to see lossless wav files for premium users, I would like the ability to bulk delete files, that's four things off the top of my head without any thinking, I could find a dozen more before having the ability to say sh*t in a lyric

woven pilot
#

can the ai give the BPM of the created songs ?

mystic zodiac
#

@potent crane Hopefully the ability to use a previous render as a reference or at least drop boxes to select the vocalist and instruments might be on your list? 🙂

kind ocean
#

I need one suggestion, If I go through "Pro Plan" then Can I use my generated song on YouTube for my music channel without any copyright and reused content.

woven merlin
#

I would like to have a possibility to tell Suno I'd like to have, for example, a 10-second intro [guitar riff, drum fill]. Now intros are mostly far too long, or maybe 2 notes.

delicate cedar
potent crane
potent crane
potent crane
rotund orbit
#

Hey guys. As of now, I've spent around $140 on suno this month. Can I please get a bigger subscription plan?

glass cliff
#

@delicate cedar (or whomever is taking suggestions): Please consider revising the price structure for purchasing additional credits. The people who have already subscribed to your service are existing customers, and should be incentivized to purchase more credits. As it is structured today, it makes FAR more sense for me to create multiple accounts, rather than to pay literally DOUBLE the monthly subscription costs for credits that will expire in a shorter amount of time, since credits don't roll over month-to-month. The last thing you want is for paying subscribers to be rationing out their credits rather than re-upping at a reasonable prices.

Also, please consider granting paid subscribers the same free 50 credits/day that non-paying users are granted. As it stands today, paid subscribers are actually LOSING 1,500 credits per month (50 free credits/day * 30 days), At this time while we're working out the correct language to produce the results we're happy with, those 50 free credits/day would be EXTREMELY useful, especially while working out new ideas to help get pointed in the right direction. SO many credits are burned off with unexpected generations, and both these suggestions would go a long way toward encouraging more usage.

Thank you for your consideration.

image.png
severe wyvern
glass cliff
potent crane
glass cliff
potent crane
#

No, I am a humble user like anyone else (pro plan)

low slate
#

I have been using suno since early beta. The BEST song i created was the very first songs I created with the prompt "an ethereal postpunk song about urban romance with poetic lyrics in the style of joy division." The song sounds more like the Psychedelic Furs than Joy Division but the bot obviously understands artist names and I was amazed at how EASY it was to get EXCELLENT results. I was in awe! Now it seems they created this amazing tool just to tie its hands behind its back and put a blindfold over its eyes. Now it mostly generates generic crap because using a genre and "vibe" is not nearly enough information to get usable results. I wasted all my credits yesterday trying to make "1980s miami bass freestyle electro hip-hop in a freaky ghetto style." and it had no idea what i was asking for. none. So If we can't use artist names to point the bot in the right general direction, then we need MUCH longer prompts to describe every aspect of the style we want. To recreate that first song I made so easily on my first try I would need to describe each instrument as well as how it is played and the effects processing applied to it, the qualities of the singers voice and where he is from, the time period, the style of poetry etc. etc. And even then I might have to use 1000 credits to get something close to what the bot could easily make on the first try if we could give it an example (like an image reference in midjourney).

glass cliff
low slate
#

exactly. i wanted 2 live crew

#

i can't think of a better way to prompt that style. I could maybe add 808 kicks, 16th note closed hi-hat pattern, call and response vocal chants, etc

rotund orbit
low slate
#

but that is my point. if i could just say "in the style of 2 live crew" we wouldn't need to spend all day playing word games because i am sure the bot could get it right on the first try.

rotund orbit
#

I would like to see it have more training on obscure musical topics like bollywood or arabian techno.

#

I mean, who hasn't wanted to write a song about food in hindi, right?

low slate
#

again... this is the problem! I know the bot can create something close to what i want but if we can't give it a frame of reference with an "in the style of" then it simply has too many possible ways to interpret the prompt.

#

I made some "middle eastern electro dance punk with arabic melodies" recently most came out pretty good!

rotund orbit
rotund orbit
low slate
#

Another of my first prompts was " a song about outer space in the style of daft punk." suno made a killer track... catchy as hell... funky. can't do that now...

#

I feel like they have crippled the tool on purpose. thats all.

glass cliff
#

I like the tool, its one of my favorite AI tools available right now. My biggest issue is the 2 credits issue I mentioned above, and that at this time, we're essentially Beta testing their model to figure out HOW to use it b/c its by and large undocumented.

plush oyster
#

got a suggestion.. allow a longer cap limit to properlly set up the prompt for both the style of the track and the lyrics. i feel its handcuffed with not being allowed to express completely the instructions, so the transformer can correctly execute and hit the nail on what we are trying to create.
also if we wish to continue generating on one track it should be able too , and give us access to the .wav file if we are paying for comercial use.. with a mp3 file there no sense of being the creator of it and we music developers clearly understand that we need the wav file NOT the Mp3 file. thank you

umbral barn
# low slate I feel like they have crippled the tool on purpose. thats all.

They do this shit ON PURPOSE.

NOW you get it.
They take away everything that makes joy for us.

The older you get, the more you'll see this.

That's why I no longer make a big issue or comment when I like something. Not even in the grocery store.

((They)) WILL take it away if you like something. 💯

delicate cedar
#

Do I dare ask who they are?

glass cliff
#

You can blame the people coming after AI technology for that. Some are willfully ignorant, some refuse to acknowledge what it can do, and a growing population is having an existential crisis about it. The ones on this side of things are pioneers, hobbled by the fear and ignorance of people not ready to embrace the inevitable.

glass cliff
delicate cedar
# glass cliff You can blame the people coming after AI technology for that. Some are willfull...

That and the way music is licensed is messed up.

There's a performance copyright which includes any recordings and a paper copyright that covers the songsheets and lyrics. So even though the songs of Bach, Beethoven. Brahms, etc may be public domain, any modern recordings likely will not be unless specifically and correctly released to the public domain.

NOTE: I am not a lawyer and my understanding may be incorrect or misinformed. This is just how I understand it.

As for the Edison recordings, even those need to be checked carefully as there are some which are now public domain and some recordings from those time periods which have been legislatively prohibited from becoming public domain.

Copyright law is a joke and the more that the rentseekers attempt to lock things down and create perpetual rent for themselves the worst things get.

#

You could say I have strong feelings on this matter...

potent crane
#

I have had a response from Suno regarding my suggestions, out of the ten I proposed, 2 are apparently in the pipeline for implementation, namely "bulk delete" and also the creation of folders in the library to assist organization of files.

vagrant rock
#

Please add Paypal as another payment option

ember stone
#

For the web interface, I would very much like to have a timeline layout when continuing songs. At the moment with the playlist layout, if you have to go back a few steps to regen or hear a reference to a previous section, it can be very difficult. There are a lot of gens in the way. This is especially problematic if you do any other songs in-between. Having a timeline to isolate the linked parts of the same song would be great.

cinder blaze
twilit crescent
feral aspen
cinder blaze
#

that works too. either way kinda reaches the point:

  1. better incentive to purchase subscriptions
  2. no lul in usage from paid accounts
  3. more incentive to purchase extra credits
delicate cedar
glass cliff
glass cliff
cinder blaze
# 
| Tier    | Time   | Credits/m | Cost  | Fee  |
| ------- | ------ | --------- | ----- | ---- |
| Premier | Annual | 10000     | $0.24 | $288 |
| Premier | Month  | 10000     | $0.30 | $30  |
| Pro     | Annual | 2500      | $0.32 | $96  |
| Pro     | Month  | 2500      | $0.40 | $10  |
| Topup   | Month  | 4000      | $0.75 | $30  |
| Topup   | Month  | 500       | $0.80 | $4   |
| Topup   | Month  | 1000      | $0.80 | $8   |
| Topup   | Month  | 2000      | $0.80 | $16  |

it lays out like a standard tier model. maybe turning the topups into a better tiered pricing structure would be more enticing for them, but other than that, it seems pretty straightforward and appropriate. biggest issue seems to be in the free credit distribution. its not quite right to compare it to '50 credits a day equals 1500 a month', because, it doesn't. i imagine a very small number of the userbase used 50 credits a day, every single day, without fail. but it would be appropriate to distribute those same 50 to people regardless of if they're paying or not, as paying for the service does remove that benefit (which is weird).

#

dang I almost made that table fit right on mobile lol

native yoke
#

hi, can let us choose male or female vocal? or one song mix with [male] [female]?

feral aspen
queen glade
#

hello is it possible to continue old songs we've made with chirp? I want to continue an old one (like 2 months ago I would say). It doesn't work on the website. Thanks in advance

feral aspen
delicate cedar
#

In 2022, under a new law called the Music Modernization Act, decades of sound recordings made from the advent of recording technology through the end of 1922 went into the public domain. In 2023 there was a pause, with no sound recordings entering the public domain. Now, in 2024, recordings from 1923 are open for legal reuse. You can download, remix, or use them in a soundtrack. Yes, these recordings are a century old, but better late than never! As you look through the list above, note that only the 1923 recordings made by these artists are entering the public domain, not their later recordings.

https://web.law.duke.edu/cspd/publicdomainday/2024/

If those Edison recordings are made after 1923 they are still under copyright. Congress has no problems retroactively yanking items from the Public Domain back under copyright again. They've done it before.

Public Domain Day 2024 | Duke University School of Law

Tweet January 1, 2024 is Public Domain Day: Works from 1928 are open to all, as are sound recordings from 1923! By Jennifer Jenkins Director, Duke Center for the Study of the Public Domain CC BY 4.0 Please note that this site is only about US law; the copyright terms in other countries may be different.[1] On January 1, 2024, thousands of copyri...

#

That being said if we really must continue this discussion, how about doing so in the #🗣️┃chill-chat channel so we're not cluttering this one, which is supposed to be for suggestions.

queen glade
feral aspen
delicate cedar
delicate cedar
craggy patrol
#

How about a toggle for Instrumental Only. You can use [Instrumental] but it's hit and miss.

rose seal
hybrid crane
#

I know a lot of people asked for stems.. I want to add to that request that is only makes sense if they are pre-processed, not post-processed. The latter we can already do ourselves, and it always has "bleed" from instruments. If it would be done before creation then I assume in theory there is no bleed.

rose seal
molten egret
#

What is a stem?

ember quarry
#

How about a separating the music and vocal tracks to download for each clip (song)? That way in post, we can have more control over each track independly.

craggy patrol
#

It's ideal when you want to lipsync a 3D character to the lyrics and not have the music mess up the mouth movements.

robust jewel
charred epoch
#

I tried entering a lot of things in the description, but it doesn't work to prevent the lyrics from appearing. In custom mode, if you delete the lyrics, a song without lyrics is created, but only pigeons appear and the quality of the songs is very poor.

versed sandal
#

I'd like the ability to crop a clip. Similar to how, after continuing, you can choose Whole Song to get a new clip that's all the component clips combined, this would create a new clip that is just the first (x) seconds of an existing clip, based on user input. Sometimes it does something really good, then keeps going and ruins the clip.

#

I'd also like to be able to take a clip that isn't part of the current generation and add it to the end of a different clip. Possibly with an option to have the AI try to smooth the transitionm

potent crane
potent crane
feral aspen
#

Cropping of a sort is coming, arguable a more powerful use case, using custom crops of the previous clip for a custom continue. So you can choose the spot where something realy good is happening, and before it keeps going and gets bad, to continue and add onto the song.

#

I don't know about more convenience style cropping, but technically it wouldn't be difficult at least, so maybe!

weak elk
#

that's great, so if a song suddenly starts fading out, you can grab it before that happens

feral aspen
# weak elk that's great, so if a song suddenly starts fading out, you can grab it before th...

Yeah, that's the idea, I think the possibilities it will enable will blow people's minds. Though it's important to remember it's still going to be a lot of RNG. I am certain that some clips will fade out even when you clip it well before the actual fade out. In Bark this drives me crazy, it's like there's some magic token in the generated audio that says "5 seconds in the future, change the speaker voice" sometimes, so odd.

weak elk
#

This morning I had several 40-60 second starts that just ended early, I wish I could force them to continue

feral aspen
#

I am quite curious if it also has the semantic properties of Bark, which are a bit like sheet music sometimes. So if you remove one tiny segment in the past, it can change the background sound of the whole clip.

#

It's like that one segment has some command that says, "Start playing a continuous sound here, until I say otherwise"

weak elk
#

I sometimes continue on 4 second continues just to see what happens. This morning a track just kicked into punk rock gear after lots of hallucinated speech

feral aspen
#

If you force it to continue, as in literally force the AI to model to not stop and keep generating, it may just hallucinate, or basically start a brand new song. But if you carefully choose a spot to pick from, I htikn you will be able to rescue some of those clips and have them keep going, by clipping off the right part of the end.

weak elk
#

yeah that would be awesome

lapis plank
#

Hey, I'd like to suggest to be able to enter as an input a .srt file in order to respect the timestamp, it would be very helpful

ember quarry
opal pulsar
#

Suno UX improvment proposal :

When you hav multiple project it's a bit complicated. We need something like a project manager to store and sort generated audio.
It will be also usefull to be able to rename an audio part
Another improvment will be the abilities to generate a whole song from selected different part.

All this improvment will save us lot of time and make your product.

Feel free to contact me for more information UX / UI proposal etc (it's my work 🙂 )

feral aspen
opal pulsar
#

nice is there a beta program or something like that

#

because it will save lot of my time and make my work better. The lack of those fx slow down our work a lot

feral aspen
#

not sure, I only know from reading what the devs post about upcoming feature in this Discord, we're all eagerly waiting together to manage our growing libraries

delicate cedar
#

Hmmm...

#

This is not a hint or anything but I only tend to use that one when I know something good is coming and I'm waiting impatiently for it just like everyone else.

...but not a hint. 😏

opal pulsar
#

^^

#

in the roadmap : "A button for quentin when he click on it someone run to give him his cofee"

delicate cedar
atomic bloom
#

Not sure if this has been suggested before. But I suggest somehow add a negative prompt? For example: if someone doesn't want vibrato, you can put that in the neg prompt in an attempt to at least not add too much vibrato in the singers voice?

opal pulsar
#

not sure if the AI model they use can do that

atomic bloom
#

This is why we should be able to continue from a part or feel from prev gens. as part whatever it may be at that point. part 2 3 4 5 6 7 etc.

#

Maybe I am one of those dodo's that don't conform to mainstream and will not make mainstream garbage...?

atomic bloom
#

of course, i may be trying to push Suno too far lolol.

feral aspen
atomic bloom
#

LOL

#

then heres one more suggestion. code a way to create an ai asdr while generating a track. For examples all the clicks and pops have to go hahahahahaha

potent crane
warm cargo
#

I have used up something like 5,000 credits to get a finished 4 minute song and I have done this for 6 full length songs so far. These are a lot of wasted credits in not being able to prompt things like specific time signature changes and I have to reroll over and over until I get that one that sounds right... only to find it ends in a way that continuing it creates some ridiculous transition. I have a few things I'd love to see to improve upon this amazing tool. 1- Grab your mental visualizer for this one or I'll draw a picture if needed: Node Based Editing. sometimes I'll spin off around part 3 and end up multiple variations of the second half of a song. Sometimes part 5 of a branch would better fit part 5 of a different branch, so I'll download the individual part 5 that I like and bring it to DAW only to find that there's no formidable way to graft the branch onto the trunk and roots of part 4 of a different branch from my initial part 3 spin off. The ability to mix and match branches and twigs off the same root would be huge in saving a ton of time and manual modulation. 2- As the continuances go, there's obvious generation loss and more hisses, static, loss of quality over all, starting with the accompaniment if it's an acoustic instrument. Basses will go moog. voices devolve into white noise on essing... perhaps having a continuance start its generation with consideration to the original seed? Or even better, if Part 1 is the hook/chorus - give the option to reinsert Part 1 at any point between continuances to keep a clear hook at top quality, from which the next continuance will have the original gen as its sample source. 3- I love the hidden treasures in using strange metatags, like obscure music theory terms and getting results to match. I know nothing about coding.

dawn fjord
#

Bulk download would be good. Grouped by song and parts (folders) would be better, obvs.

short cave
#

more credits 😦

zinc gyro
#

[Fade IN] and [Fade Out] are commands that rarely, very rarely work on tags. It's more a matter of luck. If possible, add the option for ALL songs to start and end with FADE. It sounds obvious, but having to edit externally just because of the FADE is nonsense.

surreal chasm
#

Duration limit:5 or 7 minutes for free users

edgy wren
#

Would love the ability to add/update specific parts to completed songs. i.e. adding an electric violin solo in the middle of a song.

brittle birch
#

Needs a unique branch identifier. Currently if multiple branches are generated it's nearly impossible to tell which is which without going back and listening to each one. For example: Continuing from clip one title X. branch can be X_1, X_2, X_3, X_4, etc. Continue from part 2, X_1, X_2,X_3, X4 etc... something like that... This way it's easier to track on a list and then go back and eliminate the non favorable ones. 🙂 Also needs a tree, so you can go back and trace from start to end.

hard quiver
#

API plz 👍 🙌

glass cliff
glass cliff
glass cliff
brittle birch
zinc gyro
lofty yarrow
#

this could be a goldmine. feature up v3. GUI interface, add in eeffects, basic playlist layout like in a daw, or can do by {} in prompt, go as far as buiilding like modules in Native Instruments, this thing could be HUGE all i know if your melody and precussive algorhithms on point, could interpret Fruityloops riff generator and FPC to use percussion and synth presets? HELLLLL dude big money time export as a VST module you will be minted cashmoney over night

potent crane
zinc gyro
sick wolf
#

A possibility to upload your own sound, or cuted generated (by Suno) sound in Suno App would be interesting

granite kindle
#

Suno AI music site improvement ideas:

  • a volume slider (having to go into Windows Volume Mixer and lower the firefox volume to make suno songs sound quieter, is no good.)
  • when you continue a song, let it carry over the song title and style of music. (right now it keeps the song clip but makes everything else blank, so there's a lot of copy pasting just to actually continue a song.)
  • a way to mass download audio files, with them being named correctly (I'd pay 10 bucks to do this.) Midjourney lets you tick boxes by date and mass download. good feature.
  • a way to search and display only the family tree of a single generated song.
  • a way to properly rename songs.
  • if songs in the library were listed in reverse order (pages count up as new songs are added, instead of shifting the entire library to keep new ones on page one), people could remember the exact page a song is on, and it would not change, making it easier to find the songs.
  • ability to trim a previous generation and continue from that point. (for example I generated an amazing start to a song, but it "ends" during the song, so I can't continue it.)
    Suno is amazing! I used Suno to make this music (westingtyler, "Midnight Overture" on yt) and see the amazing potential. these features would make it even more amazing.
marble turtle
#

Not sure if this has been suggestedf before: Could you explore adding a feature focused on generating songs with persistent vocals. This would enable users to create tracks with vocals that maintain a consistent tone, and style, similar to a specific artist's signature sound.

raw cave
#

CREATION LENGTH
• Long – up to 120 sec. Any creation less than 30 sec. automatically gets refunded.
• Short – less than 30 sec. for short intros, ends and instrumentals breaks.

CREATION TYPE
• Complete Song – A.I. prepped to begin and end a song within 120 second limit
• Start Song – A.I. prepped for intro or fade in. Vocals held back from the beginning.
• Continue Song – Repeat pattern from previous. Vocals can jump right in. Could also be used as a song start if you want the vocals to jump in at the beginning.
• End Song – End or fade. Vocals could automatically refrain out.

TEMPO CONTROL
• Self-explanatory

VARIATION CONTROL
• Controls the A.I. seed randomization. Continue Song selection might defualt this setting to low whereas, Start Song or End Song settings might default this setting higher.

VOICE CONTROL
• Controls instruments vs. vocals. Higher settings could possibly force the A.l. to sing when it's refusing to do so. Lower settings could possibly reduce the potential for vocal hallucinations.

image.png
vagrant rock
#

And please get rid of all the censorship for not being able to use lyrics...

#

"Couldn't Generate That", a minute ago you could!!!

vagrant rock
#

With Remix use the continue from from the original please

tacit mist
edgy oasis
#

What an amazing app. Thank you all for creating it. Here are a few things I'd like to see: 1.) The ability to save the voices generated as an artist profile. 2.) The ability to describe atmospherics in a song (crickets chirping, small creek water flowing) 3.) Should be #1 but 16bit/44.1khz for pro accounts. I am taking this ai very seriously, as it has the potential to be stellar. 4.) A GUI similar to art editors in allowing modification of different sound elements and vocal inflections on certain lyrics.

old quiver
#

one thing that I'd like to do is be able to use tags to overlay one set of vocals or a specific sound against another

real root
#

If you lot need someone to train or submit some voicemodels for higher quality output (not sure if it even words that way) I have plenty that I made personally at 1k epochs of myself and my singer friends. I can also train new models. We’re a multiracial polyglotic group from New York with a wide vocal range.

merry hatch
#

i think it would be great to be able to generate a song and be able to regenerate the lyric with in that song with out have to get a new track, to help correct typos or just refine the flow of the lyrics

glass ivy
#
  • free credits plz
desert condor
#

Add more pronunciations for unused words like totally

bold cedar
#

the ability to generate intros and outros for already tuned songs, please

hybrid crane
bold cedar
#

but now it's really worth paying

hybrid crane
bold cedar
hybrid crane
#

And if it is not worth paying for, I don't understand you want it for free... lol
Clearly you love it, but want it to be free... Yeah sorry, but that is not how the worlds work 😉

bold cedar
bold cedar
#

🤫