#💡┃suggestions-web

1 messages · Page 2 of 1

fringe knoll
#

It's easy to agree with a strawman argument.
You were baited into supporting a falsehood. When you verified it, you discovered it wasn't exactly as told.

This is a good thing to learn. It's why I didn;'t put anyone in time out.

(and remember who was deceiving you. They will do it again.)

thick iris
#

...why are people fighting here?

fringe knoll
#

they were being censored.

molten orchid
thick iris
#

Ah. Well, its much harder to be nice, so I appreciate anyone trying to get that energy going.

On another note, is there some sort of cheat sheet or prompt building for Chirp? This technology is amazing....

fringe knoll
thick iris
#

!!!!

#

AWESOME! Thank you 🙂

molten orchid
#

Also I don't believe he had any ill intentions, it happens sometimes that something doesn't work and we assume we know why. Just like the web app says: "use another prompt and try again" might lead to the assumption that your prompt is wrong/censored, while it's not the case and you just need to try again with the SAME prompt later.

fringe knoll
#

this is true. I had the same prompt censored and uncensored, like 5min apart...

molten orchid
#

So I think it's a good suggestion to edit that error message to avoid confusion

leaden marlin
#

Hi, i made a great song snippet on the main Suno page, and want to extend it, but it seems that i'm unable to... Any ideas? Thanks in advance

rose seal
leaden marlin
#

Awesome! Thank you! 🙂 (y)

gentle wasp
#

It’s seriously like 90% male voices even when I specify “female singer”. I’m just throwing credits out the window is all.

delicate cedar
#

If you don't play you can't win.

feral aspen
serene shadow
#

I appreciate its multistage process and Quant on diffusion might be hard. but are there any plans to put in place a "tiny" model that would be practical on the web?
I need about a 2x speedup to get realtime and all batching and faffing in the world aint gonna get me there, AFAIK

rose seal
#

is this re bark?

autumn pilot
#

PLEASE, make the lyrics avoid having mention of the genre or instruments, it's so annoying to have the lyrics mention "In this nostalgic lofi beat ..." , "with the catchy guitars ringing in my ears", " In this lofi jazz, my mind takes flight" , "In this indie pop symphony, we'll never be the same" .

If there's anything that I dislike about the service is those cheesy lyrics :(((

feral aspen
granite cape
#

Regarding the web app, please have the song download open up in a new tab rather than replacing the existing one. It's very annoying to have to reload the page every time by going back in order to continue making songs.

rapid bison
#

Do we gonna have an api anytime soon?

molten orchid
#

Btw, is there a way to cut the song at specific time to then continue from there? If not, then I leave it here as a suggestion in case no one mentioned it

waxen sand
#

A way to buy rights (same rights as in paid version) for a clip / song you made with the free version (If full song, pay pr. chirp). Idea of how to balance to avoid free generation abuse:

1: Requires active subscription

2: User has to pay a higher price (either with money or using many credits as payment) for the clip to get the rights

3 (Would be nice to have even if feature I talked about above wouldnt be added): Possibly some icon to display if the clip is generated with or without subcription (if paid for like explained above, it will change status as made with subscription) so its easier to know which you have the rights to or not

delicate cedar
formal bear
#

given the list of things we have to prioritize...

#

we will do it at some point

delicate cedar
#

Awesome! Thanks.

shell quest
#

I would appreciate the chirp generation error messages being clear. Like the fact that artist names are now rejected. Same thing for lyrics censorship, I'd like to know which key phrase the model (GPT ?) refuses to process. In fact I would want an uncensored lyrics generator, it's simply unacceptable to have artistic limitations in genre such as rap where social criticism is fundamental. I tried other models but they either also censor or are not creative enough. I know that there is one earlier model from openai that is trully uncensored.

fringe knoll
#

but agree, some better feedback as to WHY it was blocked (or just glitched?) would be appreciated... so it feels less like gaslighting.

#

(sometimes just re-arranging the words will be accepted... 🤷‍♀️ )

shell quest
#

Sometimes yes, sometimes not. Sometimes the lyrics will be generated but the speech will be blank on one of the 2 versions. Seems like a second censorship.

fringe knoll
#

if the lyrics are there but Chirp didn't sing, that's not censorship, that's just Chirp being random...

#

(not that THAT explanation gets the song you wanted)

#

several AI all have to cooperate at the same time...
🤖 🎲 🤖 🎲 🤖 🎲 🤖 🎲

shell quest
#

Other suggestion : I think we need a new interface, it's just not handy to scroll for miles, then switch pages and scroll more just to find an older song.The generated image helps a lot to identify a song, but we need a text based search. I'm thinking of a sort of catalog which lists all prompts (or genre + first line of lyrics), the duration and some buttons to download, continue generation etc. Clicking on the text would display the full lyrics and large image in a separate area of the browser.

leaden marlin
#

the ability to edit already made iterations would be nice for those times the lyrics jus need a tweak but the general idea is near perfect.

#

not to mention to reuse the same rhythm for the sake of consistency

rose seal
leaden marlin
orchid lava
#

Just a small suggestion, I'm sure I've seen it before, a delete option for chirps.

feral aspen
#

Suggestion: Navigate to a clip based on filename. Ideally as a link you can save for handy reference like app.suno.ai/me?chirp=913c8e28-0c62-4d13-8cf8-d859ba42a29b but text input would be fine too

feral aspen
#

Feature Suggestion, more of a future thing I'd guess. Sketch Mode - generate a larger number of very short clips at a cheaper cost, meant for exploring and testing prompts instead of creating a final product. Testing tags or styles I often really only need just 10 to 20 seconds of audio, and it's fine if that audio just cuts out and stops abruptly - it's just to test prompting before doing the real gens later. So basically you could spend 10 credits to get 4 - 8 very short clips instead, whatever makes GPU sense on the backend.

waxen sand
rose seal
orchid lava
#

A useful idea would be if you could generate a continuation from a previous clip, but add it to the end of a later clip, so let's say a clip has a part where it goes silent for a few seconds, can sound great in the right context, but everytime I try and generate from that, it literally just creates a few seconds of nothing. Whereas if I went back to a previous iteration and got a generation from that but spliced it together on the latest one, that would be fantastic, I know I can do that in Adobe Audition or similar, but it breaks the flow a little bit.

granite cape
#

Feature Suggestions:

  • Providing Chirp is a modified Bark (I'm unsure of the specifics). It'd be really cool and useful to be able to save/extract the generated speaker for a generated song so the voice may be used within Bark for other genres, songs and perhaps even just speaking. This could also allow users to experiment with outputs using Bark and then use Chirp for full attempts.
  • Relating to the previous, but rather than Chirp -> Bark, being able to save a speaker for use with other Chirp generations (Chirp -> Chirp) without continuing a song would be very useful as well. If they could be saved (preferably by name/seed) and imported, this also allows the community to share their favorite speakers with others to use with their own creations.
rose seal
oak echo
#

what is radio mode?

#

make the bark tts longer

spice tartan
#

I wish combining multiple songs would be done using AI, so the transition is smooth

#

And or sometimes lyrics end too quickly, it should try to continue them until the end then continue with the new lyrics

autumn pilot
#

Damn, props on the new UI ! Tons of new features, I'll be busy the next few days ! 😄

#

btw I love we don't need to add lyrics and it works phenomenally ❤️

bold cedar
#

Possibility to generate longer songs, please, even if it would be a paid option

autumn pilot
#

There's the new "Continue from this clip" which allows you to generate infinitely long songs I believe.

bold cedar
#

Yes, but it's not coherent

fringe knoll
azure sandal
#

Hello everyone, first and foremost, congratulations on the update! It's fantastic. I'm not sure if this feature has been mentioned before, but I believe it would be highly beneficial to implement a "seed" function. This would allow us to continue generating content within a particular genre that we've discovered through specific prompts, as currently, using the same prompt always yields different outputs.

bold cedar
fringe knoll
bold cedar
autumn pilot
#

It's hard to piece out what continuation belongs to what clip, and since I get 2 clips per continuation it gets exponentially harder with each continuation. Also I think there's a bug with get the whole song because it's giving me parts I hadn't generated

autumn pilot
#

Do you guys think a continue from this clip feature could ever be applied to non-chirp samples? Like using my own music to generate longer sections , that would be really cool !

supple lake
#

I really am loving everything in the update. The only thing I really would like at this point is some sort of seed or way to have a similar sound again and also a way to have more of a sense of how to prompt for specific genres

#

But thank you thank you for this. This is honestly a dream.

frank sequoia
#

I had an idea. What if it was possible to, instead of describing the style of the song with words, you could link a video or upload an mp4/mp3 so it can have a style similar to that?

trim siren
#

Integrate this as a plugin into some music production software. Allow use of any audio as a source, arbitrary lengths of generated audio etc.

#

The web interface is relatively easy but the audio generator tech is so absurdly good that it's handicapping it. You could be selling this to music production studios.

#

Like, I can imagine selecting the wave form of a drum loop and prompting the generator to create a drum solo from it.

#

Ideally down the track it'd be nice to have everything isolated into separate stems somehow, though I imagine this would be hard.

spice tartan
balmy zealot
#

It’d be a courteous gesture to allow a refund or maybe split the credits used when Chirp glitches with only a 7 second clip haha

amber kernel
#

Hello everyone ! An API release for the generation of musics would be amazing. I'd be your first client ! 😊

zenith gorge
#

Discord - Make suggestions and Bugs forum channels so people dont clog up the channel/ask or report the same thing repeatedly and or keep these and have ones that are asked repeatedly have a post in zed forum channels?

final sphinx
#

I'm discovering your website and it's fantastic. I would like to make a suggestion: could you add an option to download the voice and track of the generated song? It would be great to be able to re-edit or master it 😍 That would be awesome guys. I hope you keep it in mind, I'm really enjoying the experience. I also hope that complete songs can be made or at least that the AI can give a better ending to the songs

wraith lava
#

Use terms about commercial license should be simple & clear to understand

delicate cedar
#

The problem as I understand it is that the music industry has its own special legal language that is different from how licesing generally works in other types of situations. And AI is such a disruptor on so many levels in an industry that is rapidly changing that it's hard to make anything as clear as anyone would like it because so much is in flux right now.

Until one of us makes a hit that completely dominates the market and makes a boatload of money, this situation is likely to continue to be in flux. Once that hit happens though? Expect lawsuits from every direction as everyone tries to see if they can get an interpretation of the law that benefits them established as precedence.

Reminder: I am not a lawyer, I do not play one on TV, this is not legal advice. If you take actions based on this, your mileage may vary, objects in mirror are closer than they appear, do not taunt happy fun ball, and do not pet the sweaty stuff. My horoscope said "you will meet beautiful stranger" but so far no luck. Sigh.

zenith gorge
#

Web - If the clip is continued from something show what its continued from

image.png
mossy nebula
#

Loving the new website! I have a suggestion, it would be great to have the stems so I can create idea from it.

fair loom
#

Main Suggestin: 1- Add Stems Export , - Acapella Export, so users can have all the instruments used in the A.I. Song - (For premium users OFC)

frank sequoia
sly furnace
#

Suggestion: Please let us delete chirps in the web-app. It's not possible to work with chirps if there are hundreds of them

molten orchid
copper plank
fringe knoll
dusk venture
#

It would be nice to have some kind of chirp merge tool, where we can fine tune that spot where the chirps merge together while creating the whole song. The upload on the left has that empty 3 second space around the 56 second mark. I did a crude Vimeo edit for the second upload. Some kind of mini tool that lets us just control the merge point would be helpful.

autumn pilot
#

just happened again... So annoying.

autumn pilot
#

It keeps consistently happening......

fringe knoll
# autumn pilot just happened again... So annoying.

is it possible the chirp you are continuing from, has a stelth ending, that it has ended the song a second or 2 before the chirp ends...

then the Continue is trying to continue dead air, creates a few seconds and quits because it thinks it is done?

fair loom
#

Important suggestion For the future , let' the user choose vocalist voices , male , female, tempo , key of the track , and also have a consistent main idea to continue the song? 🙂

#

important to : Choose the style between parts , if is rapping or singing..

autumn pilot
blissful current
# rose seal This will take some R&D but is definitely on our list!

OMG, being able to save and reuse voices is on the list? I can't wait! 😁 It'll make my Synthia Nova project so much more consistent... Though we still need an actual API (or at least an API way of requesting/refreshing auth tokens; I'm willing to manually code the rest!) to fully automate this stuff.

ivory yacht
#

Train it on more specific music styles. For example, DPRK type music doesn't work. Also many others (Even more normal ones) unfortunately don't work from my experience (Sometimes makes long clip with nothing but weird noises and sometimes random music that almost sounds randomly generated)

neat berry
#

First, this is an amazing product!
I want to suggest in the style, to be able to write "in the style of band X".
That would capture a lot more information than a few prompt words, and technically should be possible by fine tune the NN model to their songs.

slim aspen
#

Echoing the request for Stems. If you had that available I would subscribe to a top tier in a heartbeat.

fringe mountain
#

May have been suggested again, but you need to be able to edit the lyrics on a generated song cause if you like the result, you might want to change some things after the generation

autumn pilot
#

Are lyrics censored ? :/

#

Wanted to make a meme song but can't because of the topic lol

cursive stirrup
#

does suno has API to use?

feral aspen
agile forum
#
  • having a way to generate a song without the form (just writing inline, example: /chirp lyrics:"my awesome lyrics", genre:"rock"
  • being able to download the music separate from the vocals
supple lake
#

I am LOVING the new site. The one thing though is I can get confused as to which clip I'm continuing since both clips generate with the same picture and lyrics, unless there is something I am missing.

#

And I've accidentally hit the wrong one before

#

Being able to piece the whole song together though is amazing.

#

Another thing that could be cool is a way to prompt an introduction. So to do a lead into the start.

minor wedge
orchid lava
#

What would be amazing, is to upload a song you've made yourself and have a singer put your lyrics to your song.

#

So let's say I've just made a track on a daw, to upload that track and have a singer, sing onto your track. Imagine...

worn horizon
#

I made unofficial ChatGPT bot for Suno AI https://chat.openai.com/g/g-HYh6nK5UK-suno-ai-music-creator
Allows to search internet for news or memes, convert them to music one click compy prompts for Suno and provides link to app.suno.ai
Here it suggested prompt about Black Friday that resulted in this song 😄 https://youtu.be/Z-wonI-uj68

But could make sense to make official one + action for others to use. Could allow people to find creative ways to use Suno with ChatGPT

ChatGPT
ChatGPT - Suno.ai Music Creator

Unofficial Suno.ai Bot - Transforms conversation themes into musical genres and titles to turn in to music

YouTube
Eduards Ruzga
Song about Black Friday made with Suno.ai and ChatGPT

Made with https://app.suno.ai using https://chat.openai.com/g/g-HYh6nK5UK-suno-ai-music-creator

Allows to turn news, memes or anything else in to song ideas for suno.ai

Have fun

▶ Play video
shell quest
#

Would be cool to base the image generation on both the genre and the lyrics. Right now I have a dozen of undiferenciated birds because I'm doing instrumentals.

fringe knoll
quiet turtle
#

Is it possible to influence the song length?

glacial solar
#

Is there a guide on the all ya need to know like current styles and prompt examples ?

fringe knoll
#

OR:
cut the chirp shorter by supplying it with fewer lyrics.

earnest storm
#

A way to make the AI remember past sections so the chorus can repeat and stuff

fringe knoll
fringe knoll
earnest storm
fringe knoll
earnest storm
#

Okay

#

Thanks

gray hornet
#

Hello: Is it possible in app.suno to create a prompt so that the chorus is sung by a woman and the lyrics is rapped by a man?

fringe knoll
gray hornet
fringe knoll
gray hornet
fringe knoll
#

Mmmm always a roll of 🎲
I sometimes add a human 'warm up' sound " ahhh AAHH ahh" or "MMM-mmm-mmmm"
before [Verse] it seems to trick Chirp into singing right away when music starts...

listen to the begining here:

molten orchid
#

Will we ever be able to separate instrumentals from vocals in the future? If so, I'm curious if it would be applied to existing tracks made

zenith gorge
restive quarry
#

hey guys can i fine tune suno for other languages that are not currently listed?

true bridge
#

Hi,

I wanted to share a success story of how I was able to use Suno.ai to create six different genres full songs in three different languages (English, Hindi and Filipino) within just 24 hours.

Here are the songs produced:

https://soundcloud.com/drshell-1

I also have some product improvement suggestions:

  1. Ability to choose language, voice (Male/ Female), Scale and Tempo
  2. Ability to restart song composition from one of the clips produced with in the process
  3. Ability to join separate clips into a song in a separate composer window
  4. Ability to export audio in other high quality audio formats and not just mp3
  5. Separate Suno.ai music store where creators can post their music for people to buy
SoundCloud
Shell
Shell

Listen to Shell | SoundCloud is an audio platform that lets you listen to what you love and share the sounds you create.

▶ Play video
quasi moss
#

if you add ukrainian landuage its will be amazing

sharp perch
#

Will it BE possible to generate Longer Songs in Future?

true bridge
#

I already created it @sharp perch See the post of souncloud profile. All full songs.

#

You will have to create song part by part and join it.

sharp perch
true bridge
#

Listen Electric symphony of love @sharp perch

sharp perch
true bridge
#

Once you like a clip, choose continue from clip in the options

#

Create song part by part and join them together as you go

sharp perch
#

Thanks for help

true bridge
#

My pleasure

sharp perch
true bridge
#

Thats right prompting. Try different prompts till you get the voice you like. I am not sure why my model always produce great proffessional voices.

#

This is the unexplainable part of AI

sharp perch
#

Ok

minor wedge
true bridge
fringe knoll
true bridge
#

@fringe knoll Imagine six full length songs in six different genres in three languages within 24 hours. This will revolutionize digital music production.

#

I could not sleep whole night hahaha

#

@fringe knoll try out the song Electric Symphony Of Love, its crazy

minor wedge
true bridge
#

I am half dead as I could not sleep whole night hahaha

true bridge
young basin
#

Suggestion:
let COPY STYLE from any result(!) :

++ add "GRAIN" (!)

For example: I liked THIS result (No. 2)
#🐣┃chirp-beta-3 message

I want to generate a NEW song (different words), but IN THIS STYLE (!)
I would like in my new request == Grain from the previous result I liked, and get == a new song in the same style

Thank you 🙂
YOU DID VERY COOL :)))))

=========================

Also add:

++ emphasis on the desired syllable - in the form of a “Capital letter”, for example:
"mAke up", or "make Up";

++ duration of sound: in the form of “a”, “aa”, “aaa”, for example:
"mаааke up", or "make uup";

++ add “backing vocals” to the background - in the form of light reverberation, so that together with the main vocals we hear parallel backing vocals with light reverberation (!)
for this we need the "=" symbol at the beginning of the line:
= This line will have NO backing vocals, but will sound with a slight reverberation" ;
== this line will be clean, without reverberation, but along with it, in the background, backing vocals with slight reverberation will be duplicated;
=== combination of the two previous options (!)

++ similarly with ECHO:
the symbol '*' will mean "ECHO" (!)
application options - as in the case of reverberation (!)

++ enter CONTROL to completely end the line at the end of the composition (!)
(because it has been noticed many times that the last word in the composition = INTERRUPTED, and this is offensive!) 😢

++ prohibit long intro (!)
maximum 1-2 bars (!)
(because it generates an intro of 4 bars, and it is 43 seconds long... there is no time left for the song (!) 😰

limpid nymph
#

FUNCTIONS WE NEED
-ability to add our beat
-ability to select a certain voice
-ability to automatically generate new songs that are complete with intro and outro
-ability to add a song to use as "inspiration"

#

-ability to generate separated audio files for vocals

young basin
severe basalt
#

-Tone control.
-More flexible instrument control.
-Beat rate control.
-Sound effects control.
-Intertwined vocals.
-Voice effects control (whisper, muffled, etc.).

delicate cedar
#

Separate channel exclusively for music videos of the not text and generated by the Suno Chirp process variety.

#

I think as things continue to go along we're going to see a lot more of them, it makes sense to me to start a channel for them apart from the rest.

#

This could also be a benefit down the line for some kind of 'ChirpTV' MTV style channel on various platforms, such as Twitch or YouTube that would expand awareness of what can be done.

heavy mantle
supple lake
#

Is there a possibility to create a song database?

#

Or am I missing it. Like where "Turkey Beat" is, can I make my own?

rose seal
limpid nymph
#

IMPORTANT

#

we need a library of beat intros to select from in order to be able to select the genre in an easier and more accurate way

#

I have been trying for hours to make an eurobeat song with no success

rose seal
supple lake
limpid nymph
#

beat intro selection proof of concept

#

i can provide different intros if needed

supple lake
#

Awesome

limpid nymph
#

Idea: when hitting "generate full song" a segment of 2 seconds of silence is added at the end of the song. At this point the AI has to create a bridge and make an outro

#

Not sure if it would work

idle ember
#

Hello, first of all I would like to congratulate the team that created SUNO. I have a suggestion, there should be an option to edit lyrics before creating the song along with the current option to create lyrics and song in one step. I mean the lyrics of the song automatically created in Create, not in Create (Custom)

feral aspen
idle ember
#

That is, the lyrics of the automatically created song serve as inspiration, and I think it should be possible to edit

limpid nymph
#

ability to regenerate lyrics for a great beat would be a game changer

buoyant marlin
#

Devs, u r awesome! Thank u so much for this project. It is amazing, mind blowing and unfuckingbelievble.🚀🚀🚀

supple lake
limpid nymph
quick edge
#

I am not proficient in English, so I am sending this message through a translator. Please excuse any awkwardness in the language.

Since the latest update, the app now generates audio clips up to 1 minute and 20 seconds in length. However, this is causing me a problem.

In the previous version, I used to generate multiple audio clips of around 20 seconds each, select the best ones, and then combine them to create a complete song. With the current duration of 1 minute and 20 seconds for each clip, this method of creation is no longer feasible. As a result, I am unable to produce the music I desire, leading to a decrease in quality.

I would appreciate it if you could address this issue.

#

Please stop generating longer audio clips. I would like to return to the method where I can choose and combine shorter clips.

I am likely one of the most frequent users of this app. I create and post at least one song every day. You can see my work here: https://www.youtube.com/@kats130/shorts

My request is based on my extensive experience with the app.

As of today, only clips of 1 minute and 20 seconds are being generated, which is causing me significant trouble. This makes the app unusable for my purposes.

YouTube
k m
supple lake
#

Why can't you just edit them?

feral aspen
#

I understand what they mean, it's trickier to have fine-grained control when all generations tend towards 80 or 60 seconds. So if, for example, you want to change the style of music precisely from 20 seconds to 40 seconds - that is harder to do. All things that should improve in time.

rose seal
frank sequoia
#

A lot of the time, Chirp just cuts the song in the middle of saying a word, so it makes it really hard to have a smooth continue. I'm trying to make it continue but it just refuses to speak English and just starts speaking something that might sound like a language if you don't pay attention to it

#

So it would be nice if you could make it so it tries to end it at sentence ends instead if there's not enough time

sage mirage
#

I understand not wanting to be able to let people actually input audio itself and use that for generation, BUT having the ability for this model to look at audio and derive information from it ala how the /describe command works in MidJourney would be decently handy

#

Mostly because the audio we get out of Suno isn't described in a way that we can easily re-summon it, you know? There is data we don't see as users

#

The exact same prompt will kick out two wildly different outputs, to try to capture the one you like again is impossible. There are other ways to mitigate that, seeds would be one.
Being able to continue from a clip but then snip the original out would work too

#

Generate a sick jam, continue from it, change the direction of the lyrics, snip the originating clip from the chain

sick zenith
#

I suggest adding it for a fee or not to train the neural network for the user’s voice and he can choose to have the track generated according to his voice. there will be a choice and he will cast his vote.

#

make it for the mobile version of the site so that when you use custom text and style settings when you tap generate, it doesn’t immediately go down to the new generated tracks. In order to generate a new track, I need to move up again. or make it generate immediately as a batch, for example there will be a window where the number of tracks is. it is more comfortable

obsidian canyon
#

you should have a channel for showcasing projects using Suno, where they have been taken out of context - i have been using LALAL.AI to isole the vocals generated, and combining them in a DAW, with a cohesive track

balmy zealot
#

What’s frustrating is there’s too much overlap as a paid subscriber. A subscription can use all credits and then we miss out on the 5 free daily! 😭now we must wait and wait and wait and wait and wait and wait and wait and wait and wait and wait and wait and wait and

#

if we unsubscribe we get the 5 free daily again ?

#

Or we have to wait the 30 days before they’re activated, I wonder.

sick zenith
#

make it possible to change the track you like by promt. that is, you change the text in it or remain acapella but change the music or vice versa. set the bpm of the track. edit a track by text so that you can select a part of the text and choose which voice and style to perform. It’s up to the user to decide what time the track is like in mubert. also give a choice whether you need a track or a beat. can make a gallery of user tracks on the site and also do the same as in image generation applications, the theme of the day only for generating tracks. whoever wins + the number of energy to generate tracks. It’s better that the video with the track is not the text, but like on YouTube in the clips where he reads the text and you can see the words he reads.

delicate cedar
atomic bloom
#

i may be propmting wrong. But i am using the website version. most outputs come out distorted in one way or another. like the volume is peaking above 0db and or over compressed? not sure how many are experiencing this? Maybe have the output at a -3 or so db to prevent distortion artifacts? then the user can normalize themself? thought i would mention it.

feral aspen
#

I don't think it's a 'bug' in the sense that it's a simple software fix. Just something the model does and I'm sure the devs will hope to improve over time. (They already have, if you compare to earlier Chirps)

atomic bloom
#

yes they have. im still amazed and grateful regardless lol.

rose seal
severe basalt
#

Also:
-Trainable AI models, so you could copy instrument or voice.
-Library of public AI models.
-A little bit more of GUI, for example, some previews of styles combination, i guess.

limpid nymph
#

An AI editor feature would be a game changer: imagine
Being able to insert song tracks and ask the ai to regenerate just some parts or make bridges between different songs

leaden marlin
#

Would really be a cool idea to be able to have an AI Studio where you can generate the melodies and choose the instruments and generate the vocals, would give alot more freedom.

limpid nymph
#

current playback screen (on the left) vs redesign (on the right) for mobile devices

rose seal
fickle valley
#

Hey Team, great job with V2 web based but how I can't stop the "Thank You for watching" ? trying to ending any song but burning a lot of chips chasing one without it

young basin
#

Please - add, (on Site)
for example - so that I can write Promt "FULL SONG" of 4 verses and 3 chorus, the generation result will give me 1 minute, and if I like the result, then when I press the "Get Whole Song" button == I should give out the "FULL SONG" from the promt (!)
Please 🙏 🙂

naive moon
#

pls, add brazillian funk

severe basalt
#

And more options, such as:
-Teams, a union of numerous creators, having control over joint songs (for example, i create joint song in one genre, with one style, and they can continue it with their own). This could bring lot more creativity in songs.
-Inside editor, just a small feature, allowing to erase some parts, for example, too long instrumental part, and or glue parts, or remake deleted part.
-Little visualisation (that for the end version) of song's lines.
-Equalizer.
-Second vocal adder (as a different feature, as there might be people, who would like to choose themselves, how second vocal should work).

naive arch
#

Translation tool. What I thought was if a song was in another language as an example and you wanted to translate it with vocals and the original track, it would basically replicate the original version but with translated vocals.

idle ember
#

Hello again, I have another suggestion. In this case it is about song management, I don't know if it is a bug, but the storage is limited, the first songs created have disappeared. To partially solve this problem, a delete song option could be implemented; there are times when the result is not satisfactory at all and it is not worth maintaining that result. I hope it helps you for future changes, greetings.

rose seal
ashen creek
#

A feature for song generation that would be great would be to be able to add duration to the start of a song, before the vocals begin in case the song starts with instant vocals. makes it more realistic

a feature to add a loop point back to the intro at the end would be great too

bold cedar
#

Ability to generate longer songs!!!

#

I don't know...But I know that I love this site and it frustrates me that it only generates songs up to a minute and a half long, after all, it's kind of a scandal.

serene mango
#

Ive generated songs 4 minutes long on the site, just use this: continue from clip:

image.png
bold cedar
#

It doesn't work for me... The song is incoherent then

serene mango
#

Hmm might be glichy. So far doing that as many clips as I want then "get whole song" works OK for me

bold cedar
#

Good For you

serene mango
#

...wasnt a brag, or belittling you cannot. Saying it probably glitchy, and less scandal.

bold cedar
#

Hey, it wasn't ironic, I was really happy that you didn't have this problem

serene mango
#

Ah ok 😄

sick zenith
#

make generation chats by style. for example, I like to listen to this style and want to listen to generations from others in the same style.

night turtle
#

Congratulations. What I saw that could improve the app a lot is if it were possible to model audio alongside the lyrics. Example:
1 - Audio of Sung Music
2 - Song Lyrics
Based on the way the song was sung, the AI ​​looks for patterns and creates its own sound and style. It would be very interesting and very economically viable. If possible, extend the time at least for the PRO version to at least 3 minutes.

idle ember
idle ember
kind flume
#

after using, I like it, obvious improvements that would make things a lot better are 1) longer songs 2)fix the start of songs where words can be messed up 3)more variation and details in instrumentation - for instance if i make a funk song or hip hop it creates close to a loop without enough variation and details from other instruments; also batch downloads would be a huge time saver (where you can batch download all of your favorites or songs that you select)

last fiber
#

will we ever be able to create instrumentals/beats only?
The potential is crazy

fringe knoll
last fiber
fringe knoll
last fiber
#

oh there ty!

steel meadow
#

please add paypal support!!! (we brokies wanna giib money)

random tundra
#

this may have been requested before, but... the "my clips" should have a bit of sorting

1 section for "whole songs", songs that were clipped together

and the other section might be a bit more difficult but... might need a tree view for the clips.
it pains my heart when I produce some banger by accident which I don't need right now, but when I want to use it later...
it's a bitch to navigate through to find it again... and once I do, I don't know which clip that one was based on either

that's especially difficult when I'm in a 6min song, where I may have clipped a few 30sec clips togethers, and I am on segment 8/14 or something

snow bane
#

make its so that you can keep the same tone of voice throughout all your song making so that its sound the same for every song and it doensnt change the pitch and tone of voice.

delicate cedar
molten orchid
fringe knoll
tired yarrow
#

Would be good having a way to cancel the generation, this way if someone reminds their choice they will get their points back

tacit solstice
# fringe knoll replaced by the right of 2500/month 😆

But it's not 2500 chirps/month. It's 250. Which means that free users get an average of 150 chirps/month (30*5) and "pro plan" paid users only get 100 more. It's a pretty silly setup, and I fully assumed that the 5 chirps/day would still be added with the paid plan when I just signed up.

fringe knoll
molten orchid
fringe knoll
#

you also get commercial license.

delicate cedar
#

And you can buy extra credits too.

#

Those don't expire month to month.

#

That's a nice benefit

fringe knoll
delicate cedar
#

lulz

azure flint
#

with pro plan, can use sounds for commercial use ?

fringe knoll
delicate cedar
#

All I know is that I switched to the largest plan I could as soon as I could. I spent so much money on credits...

If I'd been smart I would have cancelled and started a new plan with the higher credit package. Even then I wasn't sure if it would be enough. And then they increased the amount of credits which was a Godsend! (Even though I still wasn't sure if it would be enough...)

azure flint
tacit solstice
#

Since this is the suggestions channel, I'd suggest that it be stated explicitly when signing up for a paid subscription that the paid tiers don't include the 5/day free credits. Maybe it's thought that this is so obvious it doesn't need stating, but in my case, the converse was true. I thought it was obvious that the paid tier would still include them, considering that the "pro plan" sub is only 100 chirps more without them (alongside with the other benefits, although I'm not sure I'd dub the privilege of being able to purchase further credits a benefit per se).

From a consumer psychology standpoint, it would probably just be better to let the paid tiers keep the 5/day, but I guess from a business standpoint things might look different.

autumn pilot
#

It would be killer to have an intro feature , that is a "continuation" but from the beginning. Maybe not possible but it would be really nice.

gilded cedar
#

Suggestion 1: Export to WAV/FLAC or OGG at least, mp3 has some drawbacks and lower quality

Suggestion 2: Create the singing on a different channel, if this isn't easy to do by default, consider making it so the instrumental is generated first, then when the AI vocals are generated and "glued" onto it, you can subtract the instrumental by inverting the polarity of the waveform

Suggestion 3: more space in "style of music", because i found the more detail i put in there, the better the music gets, and the more "consistent" it gets

Suggestion 4: I guess be able to change the language of the random lyrics instead of it defaulting to english. just a minor nitpick because i go to chatgpt to generate more specific lyrics with nuance anyway, but yeah.

Then, I just want to share my appreciation for this platform... This is truly life-changing for me in more ways than anyone could imagine, and I am so glad something like this finally exists. It helps creators more than you guys know 🙇

gilded cedar
desert charm
#

I'm just curious, is it possible in the future to have the functionality to upload midi(gp5, gp) to make it sound real with this service?
Like

  1. Upload midi
  2. Type text
  3. Type genre and sound expectations
    magic
    Track is ready
autumn pilot
rose seal
shell quest
#

We need a seed number or some visual indication of which chirp is a continuation and of which other chirp.

kind flume
#

suggestion: allow for bpm, key, scale controls, and instrumentation in addition to general music style and lyrics controls; also add ability to create custom collections of clips, beyond just being able to filter them by hearted clips; add more styles like IDM, drill and bass, and more intricate complex styles like this (seems like complex drums and bass where theres repeater fx and glitch fxs are not very possible as a music style yet); add more film music and epic music styles (seems like its pretty limited when I use this as a music style compared to others)

kind flume
#

I noticed the lyrics seem to override the musical style sometimes, I think there should be a "style lock" feature so if you don't want it to change music style by the lyrics it will not. This would be helpful to add more control over final result.

balmy zealot
#

Suno, partner with EmVoice App 🔥

agile badger
#

animation caption select word by word, on video render

slender onyx
kind flume
#

make continuation songs connect automatically so you don't have to do it manually in audio program, or just make ability to generate full length songs

delicate cedar
rose seal
umbral silo
#

would be nice to have some status channel like Luma has for Genie, so we'd get one centralized place for server reboots and other service issues

#

they do it like this, but main idea is just that you'd have one place to look at if there's some note why the service isn't up at the moment it gives you an error

image.png
#

right now if the chirp servers are getting rebooted, you kind have to find the messages somewhere on this discord

#

(last time I saw message on chirp-beta-1 but I actually have these gen channels muted on this discord so it wasn't that easy to find)

rose seal
umbral silo
#

ah, I missed this. it's even better if you use the site version

rose seal
umbral silo
#

one thing that could be neat for the web version is if you could have genre tags for your songs.. and you could just filter your clips with those

#

I dunno how easy that is to pull off, considering some genres are multi word

rose seal
umbral silo
#

this could just be custom search filters for your own generations, with some preset filters for common genres

#

I dunno how many of the users gen all different genres here though

feral aspen
lament verge
leaden marlin
#

seed number would be great =^.^=

odd shoal
molten orchid
steel meadow
#

cus like

#

sometimes i have an amazing song

#

but it's like the last 30 seconds

#

and when i continue it just gives me a couple seconds of silence

feral aspen
#

It is possible but I think still requires some research and dev to make it work. I am also quite curious if you could train another chirp model but use all backwards audio files and that would work. I think it would but it's hard to find any example of somebody trying this. Then you could add an intro to your intro, over and over again...

#

The general purpose version of this is basically inpainting and outpainting, which can do all kinds of stuff including that

sage mirage
#

It feels like sometimes the model is trying to complete a clip within whatever time it has

steel meadow
sage mirage
#

Like it seeks and ending within the 1:20 or 1:00 window

#

I know it's been mentioned by one of the team that they want to allow a section to select with then to generate a clip from

molten orchid
sage mirage
#

That too yeah

#

Both, idealy

feral aspen
#

Oh I thought you mean you wanted to generate an intro, basically generate backwards from the start of your song, the clip goes in front

sage mirage
#

The 1:20 to start and then only 1:00 chunks after that really fuck with my timing and song structure sometimes

molten orchid
toxic haven
limpid nymph
#

How does the algorithm select the genre? Are all genres avaible?

copper plank
#

I've got an suggestion:
What about making a tool that users can send a song and SUNO describe it for us. Doing so users can start learning how to create better prompts, as they will learn how SUNO is interpreting things

runic shell
#

pls make ability to generate our own voice and tone

vital estuary
#

sometimes my song loads without lyrics can i get my tokens back almost a 3rd that have music with lyrics printed but no singing

median kernel
#

Suggestion though I don't know how difficult it is to implement: What about being able to better direct what key the song is in.

rare basin
#

Api?

bright yacht
#

In Discord, it would be helpful if the file download included the clip's token name, similar to the web app, rather than just naming it 'chirp.mp4'. This is important, among other use-cases, since the 'Continued From' reference link is not available in the web app's user interface.

jolly meadow
#

Would be nice if the machine kinda learned the sound and flow of stuff we like from favorited clips.

somber ice
#

Would be fun to have an AI analyze output for chords/melody, whew output to lilypond (TeX-based music typesetting)

jolly meadow
#

Why don't you guys set up a poll with some popular suggestions and let people show their support for what they want to see next?

limpid nymph
#

a feature to cut the generated segment and reganerate a middle part would be awesome

bright yacht
runic shell
#

plz allow delete item on My Clips

mighty briar
#

SUggestion is there a way to take API to integrate with Telegram?

obtuse knot
#

This has probably been suggested... But it would be nice to have the ability to set a key and tempo for a clip.

#

Also, how about a VSTi? That's probably asking for waaaay too much, but it would be stupid powerful if you could sequence a melody in a DAW and press a button to have that melody directly chirped into your DAW.

amber violet
#

what all tags have meaning? [chorus] [verse] [bridge] [intro] [outro] .... [instrumental] [break] ??? and it does appear most of the time, it will repeat the chorus in style / phrasing / melody if words don't change.

#

suggestion: we are writing songs. we should be able to put the whole song in and the app know where generating stopped so it can keep going. (and I did see someone mentioned knowing the clips used to make the song, and tracing).

#

suggestion: would love a 'redo' with the same style, but starting fresh. and we create 'templates' of styles to reuse later (and singer vs song style).

noble mural
#

Suggestion: Can you provide an API to generate music? I want to try to provide a web page to generate

balmy zealot
#

A way to link progressions of a song together would be very helpful. Some way of organizing..
Like ripoff ComfyUI, if you’re familiar.
Users could benefit from a progression tree .. does that make sense?

balmy zealot
kind flume
feral aspen
kind flume
molten orchid
rose seal
copper plank
#

Suggestion: active members that help answering questions here in the server get free credits, to keep the ball spinning. 😎

compact lark
#

😎

#

Suggestion: let's give a whole day of free use Suno , for example Sunday, and all other days only premium users.

molten orchid
compact lark
amber violet
#

suggestion: use [end] [fade out] or something to actually end the song without being so abrupt. sometimes I get lucky with adding ah, oh as the last prompt by itself, etc.

copper plank
#

Clarify Terms of Use: legit question NOT COVERED in the Terms of Use
What if you make a continuation of a free song you made, with a new continuation that you now have the rights to, can you sue Suno for derivative works? And also the other way around: what if you want to use free chirp and add it to a paid chirp, the song can't be partially your .

harsh fossil
#

(I'm sure all of these are in the pipeline already, but wanted to ask about these):

  • Export/download the raw .wav file, rather than the hyper compressed mp3
  • Presently there are autogenerated genre tags based on the prompt. What about tags for additional dimensions like: Key signature/pitch area, desired instrumentation, tempo, vocalist gender.
  • Toggle on/off parenthetical expression (ad-libs) in autogenerated lyrics. These are very strange when injected into classical music etc.
  • Specify instrumental only
  • Documentation on what [section] tags are meaningful in the lyrics section.
regal dragon
#

SUNO...on this site you said that every day you give 50 redits/day....for 2 days nothing is happening....what is happening? https://app.suno.ai

delicate cedar
regal dragon
#

The Hepcat thanks for the reply, I sent it via Gmail...I hope it will be read? 😄

mental horizon
subtle glen
#

where i can list all the music style for version2 ?

long robin
#

A visual way to follow which parts you are continuing from which (like a flowchart) would be awesome if that's ever possible 😄

hard minnow
#

Make so when we use continue from this clip but input changed first verse it'll replace first verse

#

If it wasn't added already

sick zenith
#

I really need a voice choice to generate a track

lofty kayak
#

Hi, it would be great if there is a GPT from GPTs (openai) with all the parameters what can enter in lyrics and style section.

fringe knoll
lofty kayak
fringe knoll
lofty kayak
#

just a bit more info about usage. It is a great model and app you have here (paid customer here )

sick hollow
rigid spear
#

I'd love if Suno gave us an option to download the stems.

coral heath
#

Would love to download credits on a per use basis (like a 500-1000 credit package)

sage kite
#

I would love for the songs to be longer and not just 1:20 on the d suno website

compact lark
#

**Where is v2 in Discord ❓ **

delicate cedar
fervent zealot
#

an option to choose to get an audio file

rose seal
low dirge
#

Is there a way to output MIDI and/or maybe separate the vocals with the instrumentals? Mainly looking for this option so I could remix. The tracks are simply too good. There was another AI audio engine that could output midi but I forgot what it was.

I know that I could use plugins to separate them (potentially) but it will never be as clean as having separate stems (just two needed)

amber violet
#

suggestion: better file management online. maybe a song title to track / folder and group... and search. 🙂 I'm usually creating 2 songs at once and getting lost. going back to 1 song at a time.

zenith gorge
#

Update this to provide info via a Clickable button called Why was this flagged?

image.png
brave terrace
#

I am super impressed by the quality since v2. Here's a some things i would love to see:

  • ability to re-generate v1 chirps in v2.
    -ability to upload low-res audio clips and "upscale" it to full frequency 48Khz.
    -Improved file management in web interface; ability to delete chirps, and move them into subfolders
    -download as uncompressed WAV files.

I've also noticed that the vocals are frequently missing not generated, or the vocals start just at as the 1:20 is coming to an end. It's frustrating if you are spending credits but not getting the vocals.

But on the whole, wow! amazing stuff. Please keep it weird! 👏

leaden marlin
#

It would be so great if you could train your voice on Chirp v2 then you could use your voice that would be really great

velvet vector
#

A search function on the website would be very cool and also if you could maybe work together with other users.😁

leaden marlin
#

Because not many people like the voices that Chirp sings, there are some good voices, but if you could train a model of your voice and put it in Chirp to generate music with your own voice, that would be great.

#

Chirp developers could do this or try

#

And another thing could also have a function to choose Chirp's voices, for example if you want a male or female voice, which you want to use because there are times when Chirp changes voices.

silk egret
#

it would be great to be able to check how long until the credits refill ;D

rose seal
blissful current
zenith gorge
warped whale
#

I wish there was a feature to access single instruments of a recording and for example change the volume, the pattern, the details of a instrument. Example I have a track that is nice but the bassline is not loud enough, now I would like to keep this part of the track but change some details of the one instrument (make bassline louder, drums volume down)

#

Second wish: download function for a wav file, not only mp3

feral aspen
zenith gorge
zenith gorge
silk egret
young basin
#

add:

  • negative promt to exclude unwanted elements, for example - rap.
  • expand the positive promt in “Style” == clarify the rhythm, several styles, and using the symbol “(!)” == focus on the style.
  • indicate the number of bars for the Intro (0-1-2). with "0" == the song starts from the very beginning, along with the music. When "1" - the intro sounds 1 bar (4 bits).
  • a pause of 1-2 bars between choruses, when, for example, the chorus is LONG and consists of 8 lines. you need to add a pause between lines 4 and 5. Pause length = 0 - short, 1 breath, 1 = 1 measure (4 beats) ...
unique pine
zenith gorge
unique pine
#

But one reason for testing is that I wondered if the AI would recognize it and make a cover. 😸

#

Something I would have found absolutely hilarious.

silk pebble
#

I'd like an undo button next to the generate. I handcrafted some lyrics and it went poof when I pressed the wrong button. But teaches me to work in another editor.

warped whale
#

Another suggestion: there is currently no real way to organise the outcomes. I basically "organise" the different results of the prompts by adding some of them to the list of favourites. But a better, maybe in the logic of "folders" or project numbers, way to organise the results would be great. Also deleting a result would be good, some outcomes are rubbish and I would just not like to see them on my list.

spare orchid
#

Idea: A optional instrumental version of the generation you made (also benificial for getting the whole song)

unique pine
spare orchid
unique pine
unique pine
#

Not that I do not like your suggestion, it's a good one. Only mentioned it as a hint what to try meanwhile. 😸

spare orchid
indigo kindle
#

Full track option uses more credits i.e. intro verse chorus verse 2 bridge chorus

young basin
#

need to add SEPARATE GRADES (0-5, or 0-10):

  • separate evaluation of music,
  • separate assessment of rhythm,
  • separate rating of the song (!)
    to “train AI” == show with assessments what is good and what is bad (!)
  • and this is a “heart” - which already exists (your favorite track) == leave it unchanged (!)
  • improve == improve the composition, improve the music, improve the rhythm of the song, change the ACCENT (proportion of sound, emphasis) on other words
silk egret
#

Add a function to change the written lyrics before downloading the video (to get rid of wrong/missing/duplicate lines)

potent crane
#

A way to force it to create a fully instrumental track (adding "no vocals" or "instrumental" only works intermittently)

wraith cedar
#

Is there any way to increase the minutes of the song?

silk egret
zenith gorge
steep palm
#

Truthfully and honestly, bark suno only needs to add stems download and the ability to use custom audio as background so AI can sing on and or rap and that is literally it 🤠

ionic matrix
#

I wish there was a way to tag the clips so that when stitching them together the order of the clips or sections could be rearranged to play with possible iterations of the song faster.

#

like a general way to stitch the clips so that if I want to rearrange the order or redo one or more of the sections I could do it, while keeping the rest of the clips intact or close to intact.

full citrus
#

How can I reuse the previous music? I use reuse prompt but the music is different. Thanks

austere escarp
#

Can we get a button on website, extra button for "decent" like where it's good but you aren't sure. I have way too many favorites and when I listen to it a week later some iof it is not great, but others are really good, but still wanna keep it.

#

Also I would love to have a way to force add previous clips before completing song, instead of having to edit in software on the PC

full citrus
nocturne hearth
#

A few wishes for the future:

  • possibility to choose the cover for the video
  • possibility to choose the voice of performer: male, female, children
  • increasing the number of musical genres
  • a guide explaining what commands can be used to make the composition more major, faster, etc.
final mountain
#

Posibility to select from which second to continue a song. Because sometimes it cuts the song in the middle of a word or in the middle of a verse.

zenith gorge
# nocturne hearth A few wishes for the future: - possibility to choose the cover for the video - p...

Voice is already possible genres is technically more than in #1154790975240155247 , there also is that #1151542418991239168 and I can link https://suno-ai-wiki.notion.site/suno-ai-wiki/4e24aa838bf84fffaf383b1f9ed50748?v=755ec8954abd4e3f9f5c899950f0f315/. First one would be cool tho

Suno AI Wiki Notion on Notion
Suno AI Wiki [Home-page]

This wiki is Community Made so there might be things that are wrong here if so just report them to the Suno Discord: https://discord.gg/suno-ai

zenith gorge
dull sandal
#

At the moment when you chirp with a random genre, a generated result also shows "Style: random". Would be great if it would show a style that it ended up using. More helpful, IMO.

image.png
dull sandal
silk egret
#

just put in the same style again

dull sandal
silk egret
dull sandal
#

here's what I get there

image.png
leaden marlin
#

Hello, developers! Thank you for your magical music generator! It's really cool, I've been up for 4 days in a row making music🙂 .
I would really like to have a function
Separate the vocals from the song with separate tracks.
And it would be great if you could download a song by individual instruments: guitar/bass/drums.

This would allow you to experiment with music in professional programs.

It would also be cool if there was a function to improve raw tracks for musicians. I mean, if a musician has created a raw track, for example, and needs to improve it with professional mastering/equalizer, etc.

It would be cool if you throw in a raw track and the AI gives you an improved one, it would be very popular among musicians

#

It would also be great to be able to delete songs from your favorites

rose seal
weak elk
#

Web is superior, definitely. Except on mobile 😉

leaden marlin
#

I'd also like to be able to edit the lyrics in the created song, because right now, when you edit the lyrics, it creates a separate song.

frozen iron
#

I know I've seen this mentioned elsewhere, but it would be fantastic if after a clip was generated that there were some kind of flags that showed the settings that ended up getting used, and a way to reuse those same settings on a subsequent clip. A lot of the AI image generators do this. This would solve a big problem that I'm sure everyone is hitting where you start a clip and the style is perfect, the vocal are prefect, the rhythm is exactly what you want... but some error in the text or a mispronounciation ruins the clip. If it was a continuation, then you could just recontinue from the original. However, if it was the original clip there's no way to redo that clip without getting a completely different total result. Credits to @light oyster who posted this and a bunch of other good ideas in another channel

lone talon
#

I have a question that can also be a suggestion.

#

Is there a way to find out what seed we get from a certain Style version from the AI?
Like the beat, instruments, and voice the AI uses without having to click on to continue from the clip of the previous song?

#

I mean not here on Discord but on the website

analog apex
#

kohi karachi ka hy

woeful tide
#

I didnt see it anywhere on here, but how do we create an outro? I know what [Verse] and [Chorus] work, but I am getting very abrupt endings to my generated songs.. .is there a way to do something like an [Outro] or similar?

#

Intro would be nice too

#

ok haha found it myself rather quickly, to help people out there is a post up above that has a wiki, there is a lot in there... my bad for not checking that first... can we make it sticky? ( message ID 1181188750081007646 )

plush scroll
#

I’d like to be able to create an intro for a song for that matter, if a song I like begins abruptly

woeful tide
#

you can use [Intro]

#

but it appears you have to make that your first entry before lyrics I am guessing

#

So an example could be something like this in the lyrics generator:

[Intro]

[Verse]

[Chorus]

[Verse]

[Chorus]

[Refrain]

[Chorus]

[Outro]

In the verse and chorus section you would put your lyrics

frozen iron
#

I would really love a delete option on the clips in MyClips. I keep generating the same intro over and over trying to get the style and sound I want, and I’ve got a ton in there now all with the same lyrics and same picture making it very hard to pick up where I left off. A lot are junk with scrambled lyrics or stuff that came out with no lyrics for some reason and it would be really nice to be able to just delete these.

wanton leaf
#

Hey, commercial use include Spotify / Itunes / YouTube and etc?

frozen iron
# lone talon Is there a way to find out what seed we get from a certain Style version from th...

I would pay $50 to get this feature. I made a clip that was exactly what I wanted, except there’s a mispronounced word in the text. Because it was the first clip there’s no way to fix it without losing everything about the style. I’ve tried doing continuation clips with the corrected text figuring I could just audio edit out the wrong part later but the continuation clips keep messing up and I can’t get anything I can edit that will fix the problem

wanton leaf
#

Is there also export stems option ?

lone talon
delicate cedar
wanton leaf
# delicate cedar Yes.

When I will buy commercial subscription, can I also use for commercial songs generated in past on free plan?

light oyster
#

Reposting this from #📚┃suno-school message cc @dire frost

Had a few questions if the team can answer:

  1. Any way you can expose the “seed” in each generation so we can get consistent generations? Many times I re-roll with the same lyrics/genre and get wildly different outputs. I’m looking for something similar to Midjourney’s style tokens

  2. Any way to instruct the model to edit the resulting audio? “Make the vocals less harsh”, “Add drums”, “Make the tempo faster”, etc

  3. Does specifying key/tempo do anything in the prompt? Would love to be able to specify this at least, in addition to genre

  4. Any way to have the model output stems instead? Right now I have to manually use a splitter, but if the model can already output stems then it saves me the trouble. Often I just want to isolate the vocals, and recreate the other parts in real MIDI synths if possible

  5. Any way to do “audio-to-audio” or “outpainting”? Like, for example, a VST plugin that looks at my current song in my DAW and then “inpaints” vocals, synths, whatever I ask for, on top of it as an audio track, keeping to the same BPM and key

delicate cedar
zenith flax
#

would like to have the steme files multitracks

hybrid crane
#

I would love the option to cut a clip and then continue from there

mint thorn
#

bring back the old chirp ? 😢 now all don't work.. is so hard to understand one things.. clips with no vocal, clips with other language.. some word like "Hyperpop" style of music banned on the prompt... wtf are you doing on chirp x')

shrewd pendant
#

Website needs a volume bar (and if there is one it needs to be more visible)

delicate cedar
subtle glen
#

PLease, train your model with more music kids

#

kids music

#

it just give us a silly guitar in each song i created

#

Thank you in advance, excelennt app

silk pebble
#

Wanted: Generate an intro to an existing song continuation that happened to start midway. Or well, I guess I can try to force it into openings somehow.

atomic bloom
#

I'm not sure how this could be done and or coded. But... I wish there was a way to create an outro from the first initial generation and also get the whole song from the rest of the liked or hearted generations. I hope that makes sense. For example: I loved the intro of a generation. I continued from that and it made a great track. Then to the outro, I want the same feel as the intro but as is we cant do this or it just creates the song from that intro only. Unless I am missing something?

#

Better I share the final output lol. The intro of this has a cool intro. However, I would really like to get that feel in the outro. As far as I know, there is no way. I could generate the outro from the intro and hopefully stitch it myself lol. but thought I would suggest this just in case there is a way and or I am missing something.

misty blade
#

Game changing features:

  1. Inpainting: Modify a portion of a generated song, including those recorded outside of Suno.
  2. Outpainting: Extend an existing song, even if it originates from outside Suno.
  3. Inspiration Reference: Allow users to provide an external recording as inspiration, similar to using an image prompt in Midjourney.
  4. Melody Guidance: Capture melodies from external recordings, including human vocals or whistling, to guide song creation.
feral aspen
#
  1. Album Art Guidance. Allow users to provide the image of the Album and Chirp makes music in a match style.
leaden marlin
#

it would be cool if i could add this bot to my server

cosmic quarry
#

Love your work! I'd love to see these features:

  • specify Tempo as a BPM value
  • Specify the Key the song is in
  • Negative prompts - specify elements or style that should NOT be generated (i.e exclude Female Vocal, or Bongos..)
  • Download separated parts as uncompressed WAV.
    ✌️
plush scroll
frozen iron
# lone talon I tried to make an [Intro] from a continuation clip

I tried that once. It seemed like it was going to work, but then the first continuation or maybe the second afterwards got stuck in “outro” mode even though the total song length was less than 2 mins. No matter what text I fed it, I kept getting “end of song” type content spitting out

zenith gorge
lyric sedge
#

Delete function pls!

oblique elk
#

Suggestion: Volume control for web app. I noticed @barren radish posted this in technical but haven't seen it in suggestions. "would love a volume control on the web version. 👀 (and have it remember the volume. Most websites do not remember the set volume for some reason)"

mint thorn
#

There should be an option to lock the voice. For example, once we have the first sound that suits us, we need an option to save the voice model and reuse it in another generation.

nova escarp
#

Pleae fix this

#

If that didn't always happen, I would pay

delicate cedar
# nova escarp

Are you trying to use copyrighted lyrics or calling for a band name in music style?

nova escarp
#

No I dont

#

writing own lyrics

#

and saying Rap as music style

delicate cedar
#

OK then definitely a glitch. (Unless ChatGPT or Chirp are objecting to the content of the lyrics themselves?)

Try adding several lines of spaces at the end of inconsistent length and try again.

nova escarp
hybrid crane
#

I personally hope you guys can find a way to increase the "bang for your buck". I think the subscription models should have more credits.

delicate cedar
#

Are you really running through 10,000 credits in less than a month?

lone talon
#

Well, I already know how to make an [Intro] from Custom, but that's if I'm written my own lyrics. Some of the generate tunes and vocals I would love to repeat, if there is ever a way to get that seed number.

hybrid crane
delicate cedar
#

Yikes!

#

You may need to look into some extra credit packages. Those at least do not expire monthly. You can use them as needed and they will still be there when the new month rolls in until that is gone and then those extra credits start being used.

hybrid crane
delicate cedar
#

I use them myself.

Screenshot_at_2023-12-06_11-53-30.png
#

The idea isn't to replace your monthly subscription, but rather to supplement it until the month can roll over again.

hybrid crane
#

Pricing is a bit hefty. I understand it takes resources, so that is why I said I hope they could give us more in the near future. 😉

zenith gorge
delicate cedar
hybrid crane
timid wasp
#

Suno should create an affiliate link for me to use in my newsletter 😀

timid wasp
#

I'll send an email

lyric sedge
void panther
#

I want to make derivative work above suno's bark using serp's bark, since it's license is MIT i understand as: "do whatever you want with bark", so, if i'm wrong about that, tell me

steep palm
# delicate cedar

pricing makes no sense, you can make another account and buy 30$ for 10K credits

delicate cedar
steep palm
#

i have a label so wouldn’t mind spending 10s and thousands if i could train and do other stuff

#

currently this suno chirp seems to be just for fun no business commercial

delicate cedar
#

Well then we'll continue to have fun with it. 😄

odd hull
#

hello, i would like to buy a subscription but i only have paypal, is another payment method than credit card planned?

pale wedge
#

Maybe it's already there but

  • having the possibility to regenerate from music or from melody. (Often one test got a very good melody but the beat is 'crap' or vice versa)
  • Having more consistency when the continue option is use. One of my test the voice and melody started great but after the third continue wasn't able to find anything better than a somewhat good imitation of the first...
  • in the online app, being able to organize our song (folder or multiple tab)
  • in the online app, being able to delete tests or unusable sample.
shy pawn
#

Add more genres of music, like flamenco styles etc

pale wedge
#

Add more metatag for the lyric.

  • for duet (voice 1, voice 2, both voice)
  • fade in and fade out
  • solo (piano, drum, guitar, etc.)
  • speaking (with or without music)
severe basin
#

"DELETE FUNCTION" PLEASE, that should be very simple to implement 🙂

severe basin
grim sail
#

Quick idea, sounds easier in theory than most here:

When generating intros to a song, as the entire prompt (to then continue the song from that intro clip)

Sometimes the length is appropriate (10-30 seconds), sometimes it goes all the way (120 seconds).

Even on the long ones, there is a perfect beat in the first 30 seconds before it starts repeating.

What if we could crop generated audio? Take an intro of 120 seconds, crop to the first 30, continue from that point (second 31)

languid snow
#

Hello. I bought a membership with money. But I cannot use it. It keeps giving errors. Can you help me please?

zinc gyro
#

A suggestion.
I don't know if it's possible.
When generating the mp4 version
Give movement to the lyrics, similar to karaoke (word for word), as in the SSA/ASS subtitle format.

dull sandal
#

make it possible to create songs in users voice

void panther
#

but for users without a good GPU would be a nice feature

dull sandal
void panther
#

take a look about using a custom voice over chirp vocal there

hidden bison
#

PLEASE add a delete so we can remove the failed attempts

loud wyvern
#

Feature request: The ability to share a clip with someone so they can also carry on generating from that clip

lavish cairn
#

How can I view my past content?

cyan flint
#

I would love to have a slider to set the amount of time on my instrumentals

plucky yacht
#

It appears you guys do not have the option to delete previous songs. I tried to generate a whole song and it added lyrics from a previous song from before into it. It would be helpful if you guys added the option to delete songs, especially to keep everything organized, which you can even add a folder, so songs don't get mixed up.

plucky yacht
#

I'm talking about the website btw.

delicate cedar
plucky yacht
#

Completely unrelated. It was 2 different songs.

#

And I pressed clear but it did nothing.

raw bramble
#

Will suno studio look into speech to speech generation is this something in the works or planned for the near future?

feral aspen
#

Suggestion: Make "My Clips" pagination a little more obvious on all screens and devices. It's often missed. Maybe put it on top instead of bottom?

thorny mural
#

Anybody had good prompt for beat with something like this?:
https://www.youtube.com/watch?v=cXTgrHruvoo&ab_channel=ArmadaMusicTV
https://www.youtube.com/watch?v=F4eccPBFEjE&ab_channel=Raulieee
https://www.youtube.com/watch?v=RScZrvTebeA&ab_channel=piano77man77

YouTube
Armada Music TV
Mason vs Princess Superstar - Perfect (Exceeder) [Official Music Vi...

Mason vs Princess Superstar - Perfect (Exceeder) [Official Music Video]
▶ https://ARMAS1013.lnk.to/PEYA

Listen the Armada Music Top 100 playlist: https://ArmadaMusicTop100.lnk.to/PLYA
Subscribe to the Armada Music YouTube channel: https://armadamusic.lnk.to/YTSubscribeYA
Click the 🔔 to stay updated with our new uploads!

Connect with Armada Mus...

▶ Play video
YouTube
Raulieee
Daft Punk- Derezzed (OFFICIAL TRACK)(FULL SONG)(HQ)(2010)TRON SOUND...

Full New Trakk

▶ Play video
YouTube
piano77man77
Blade Runner Blues - Vangelis - 1982 - [HD]

Feels almost like a dream......

Composer: Vangelis
Soundtrack: Blade Runner
Year: 1982

▶ Play video
severe basin
#

Seed number + Delete Function please

young basin
#

Above, I suggested adding “GRAIN” to generate new songs with a similar rhythm and style.
Another suggestion = add a VARIATION to the created song:

  • regenerate with a different song style (sing the song differently, with a different emphasis on different syllables, with a different accent...)
    (for a different style/rhythm, generate separate Chorus and Verse)
serene swan
dull sandal
#

would be great if you allow extending of the song in both directions, so not only a continuation, but ability to add some instrumental intro for example or acapella onto already generated piece

formal bear
feral aspen
# formal bear Go back in time? 😄

I have a standing personal feature request for Suno to train a second Backwards-Chirp all over again on reversed data just to find out if this would work. There is maddingly little information about whether it would, but it seems plausible (if inefficient...) 😄

prime rock
#

Seems like you could make a specific error message to tell people when the error is because of copyright violation, it’s not immediately obvious there’s a rule against using names and it seems to be one of the most frequently asked questions, why the red error message

formal bear
formal bear
lone siren
#

I want random to display the style (instead of random). It should be text in latent space closest to the the random. At least the closest full word or phrase in that n-dimensional space. I've been liking some of the random styles I hear but do not know what they are called.

lone siren
# formal bear Go back in time? 😄

Yes please. Most songs I generate lack a clear beginning. Starting at the verse and then trying multiple intros after I like the meat of the song seems better.

#

Cropping an instrumental can also help with making an intro to a song that has vocals.

formal bear
#

btw you can try to prompt with [intro] right now

lone siren
formal bear
#

it will get better

lone siren
#

It's already better than I thought an ai could do without plagiarism.

leaden marlin
#

hi, where can i send suggestions for the beta? i made a nice song but i cant "change" clips to create the same song in another row. for example i wrote a 2nd vers but i cant create a new song with same clips and additional clips again. im using the web edition not the discord. downloading the clips and cutting together sounds awful when i do it manually on audacity for example

delicate cedar
leaden marlin
#

that doesnt work when u already have clips. i have 5 good clips but the software dont allow to recreate a song and put the old clips together. coninuing mean in that case "its the base for a new clip." but i already have clips. just want to re-render the song again with different and additional clips i made

delicate cedar
#

I see.

#

Sorry.

leaden marlin
#

maybe there will be such an otion where u "choose" your clips and press "rerender"

#

in the future

#

hope for it

#

still a beta 😉

junior hazel
#

I stopped using suno because I'm frustrated by not being able to find what metatags work. Guessing sucks.
Just saw [intro] works, and of course [verse] and [chorus] but... how hard is it to include a suggested list of ones that are known to work decently...
Same for music styles.
Doesn't have to be a comprehensive list, even...

dull sandal
weak elk
#

Suggestion: find the source for "Thanks for watching!" and DELETE IT

tidal wharf
#

guys, is there any way to cut off the end of a generation and continue from the middle? I've got a near-perfect minute but it then goes south

tidal wharf
#

I created Jukebox Web UI, if anyone remembers it, and it was SOOO much fun writing songs bit by bit (not to mention the insane amount of creativity you can put into it)

oblique elk
#

mp4 downloads to include text available in all languages. I created a song in Thai but the text does not display in the mp4.

obtuse kelp
#

i would really apprechiate a feature where i can DELETE my clips. Its kinda frustrating to find something on the website between hundrets of unused clips, even if give good stuff a heart

junior blade
#

any suggestion on how to make Suno add a bass to a generated song? I ve tried a few prompts like [bass instrument] etc. with no luck. thx!

wraith osprey
#

having the ability to search or filter by the lyrics in the My Clips view would be nice. I have to scroll through many screens sometimes to find the song that I'm looking for, and it could be much easier to just search for a word in the song

pseudo surge
#

Can we get a delete option? Having to regenerate several times when making continued songs gets kind of confusing and it would be nice to be able to get rid of the ones we know are not good or don't want to keep to get them out of the way.

mental horizon
mental horizon
mental horizon
pulsar flame
#

But I totally agree the "continue" option really needs improvement.

And another option for using exact same music and new lyrics would be really helpful. Or something called "recreate the song" so it gives exact same music but different voice/vocals.

#

Another good option would be an option to add a previous clip for intro/extended intro. (when you forget to add intro or the AI ignores the intro )

balmy zealot
#

Please make the download links on suno.ai open in a new tab/window

worn horizon
green scroll
#

Please don't forget to pat yourself on the back for a good work devs 😁

quick valley
#

Suggestion : ability to download separately acapella and instrumental

cedar stratus
#

Could we possibly have a stop generating button, that reimburses your credits for that song and stops generating it? I often accidently press enter twice and it wastes 10 credits making the stupid "country song about thanksgiving"

wind wraith
#

Yesterday it generated jingle bells, the original song with all of the original lyrics, which I did not ask for but then it made it sound absolutely horrendous

#

This technology needs serious help

wind wraith
worn horizon
oblique elk
#

Languages : Please add all available languages as font for mp4. Thai language is available to chirp but not as text in the mp4.

worn horizon
wind wraith
worn horizon
wind wraith
# worn horizon Well, I am not a musician myself, I wonder if I can prompt it better in that reg...

So most good songs are very simple with their phrasing. I got some amazing songs on here at first but not since then. Not sure why. But phrasing and melody has a lot to do with it and the melody needs to repeat in the phrasing. So the first two verses should be similar in phrasing and then of course the chorus changes up and then you have the bridge that is totally different but to get a bridge you have to continue the song like your options suggest

#

Is there a place I can send you the best takes on here? Do you see my account?

worn horizon
wind wraith
#

Okay well I’m sure it is a work in progress. Is there a way that we the users can tell it which chord progressions or phrasing? Do we just use commas?

worn horizon
supple lake
#

I have a question---do the monthly credits that are unused end up carrying over into the next month? Or are they lost?

exotic crescent
#

i love bark and chirp so much- how can I help promote in IT?

exotic crescent
wind wraith
wind wraith
hybrid crane
#

Here some suggestions from me

sunoi_product_Ideas.jpg
#

1. Restart the Clip While Maintaining Style In some projects, the style of the initial music clip is exactly what's needed. However, minor issues sometimes prevent its use. It would be beneficial to have the capability to correct errors in the clip and restart while preserving the original style.

2. Edit This Clip: Enabling Custom End Points Modify the feature to allow users to determine where the clip stops. This change would facilitate the removal of errors and enable further building or composing from that point onward with Suno.ai.

3. Rename This Clip for Enhanced Tracking Introduce a feature that allows renaming clips for easier differentiation and tracking of successful takes.

4. Remove This Clip to Prevent Clutter Add a function that allows users to delete clips to avoid clutter and maintain a clear overview.

5. Note Feature for Personal Annotations Implement a feature to add personal notes to clips, aiding users in keeping track of the reasons for saving each clip

silk egret
#

Options I'd love to see:

  1. Seperate vocals and music
  2. generate new vocals (keep the melody intact)
  3. generate new melody (keep vocals intact)

Would allow a more pick and choose type of generating, if you like only one part of the clip.

woven merlin
#

I would gladly pay more for better quality. WAW-file for premium subs.

split flower
#

Would be cool to be able to provide the instrumental and let Bark improvise on it. Game changer for me....

brisk grotto
#

Oh gosh... I wish there is an API access 🙏

deep tree
#

Include language designator, like [Language: English]. Having it in the style would apply it by default. Within the lyrics you could change it where needed. This would override automatic language detection, as it sometimes gets it wrong.
Example:

[Verse]
[Language: Finnish] Tämä on hyvä laulu ([Language: English] Oh yeah!)

umbral yarrow
#

Is there an API for this? If not there should be! I would sign up for a plan but my use case requires API access.

tidal wharf
feral aspen
hybrid crane
serene swan
#

Splitting vocals, drums, melody from another would be cool to further work with the generations and bring some own creativity into generated songs. I know there are stem splitting tools out there but it comes with a lot of smeary sounding sounds and artifacts.

tidal wharf
brittle birch
#

Option to import/transfer previous works from the free chirps to a commercial license once you decide to upgrade to a paying subscription. It would be really sad if this was not is possible as previous works cannot be replicated.

hybrid crane
#

Another thing that would be extremely helpfull is if the music can be automatically translated to sheet music. So you can let a real band play it in a studio.

untold hollow
hybrid crane
# untold hollow Even if a good idead, for me it seems asking too much of it.

That might be true, I also see some other suggestion that I think that might be really hard te realize like for instance: inpainting and outpainting (but I would really love it, if they can), training for second backward-chirp on reversed data, I think that is too complex too... But a girl can dream, can't she? 🤩

stoic nest
#

include an option to disable the enter hotkey when editing text, or in the case of editing lyrics, make it generate lyrics instead of the whole song.

untold hollow
formal bear
#

Inpainting is on the todo list

#

end will come out very soon

pulsar flame
granite cape
#

Suggestion: When continuing from another song it'd be great to allow specifying a timestamp as a point to continue from. That way you can revise/continue before a part it messes up in. Continue after a break to begin the next verse. Or allow more time to generate music that isn't already generated.

random python
#

Suggestion:
Taking samples from Chirps and just changing the lyrics (using the same instrumentals and voice but changing the lyrics)

tender gull
#

Suggestion #1: Add threading to the web UI so it's clearer which chirps inherit from which. Or at least make it clear with some sort of backlink to the originating chirp.

Suggestion #2: Add a way to delete/hide chirps that weren't successful, in the web UI. I have a lot of chirps with vocal hallucinations and also chirps with no lyrics in my collection that I'd honestly not miss.

sage mirage
granite cape
sage mirage
#

A few of us have been saving clips specifically for when we get the ability you're also after

delicate cedar
#

A channel for tools known to work well with Chirp or music files might be handy?

real crater
#

will be great make a hall of fame with best community themes using polls

tidal wharf
sage mirage
#

Damn same

autumn pilot
#

It would be great if the web player had a volume slider , so I can turn it down !!

valid topaz
#

radio mode would be cool if you could type style prompts into a playlist and have it use random selection from the playlist to guide content

feral aspen
zenith gorge
#

Lol

feral aspen
delicate cedar
feral aspen
delicate cedar
#

Multivac would be proud. 😉

ashen cedar
#

1: i would love to import my own instrumentals and choose my own section like the into chorus verse etc etc and create custom lyrics with a voice on it of any genre and create unlimited amounts of lyrics for any songs i have that would be awsome for ideas are to even use the lyrics that the ai used if i wanted.

2: the second problem will be what if the song lyrics that generates will be the same as another person, will the ai create unlimited amounts of lyrics without using any other lyrics that the ai generated?

3: it would be nice to pick the voice that you would like weather it be a singer a rapper weather it be spanish are english will the ai be train on million and million of songs and song lyrics as well from all songs that have been release for decades so it wont be repetitive ai as well.

tender dagger
#

making it free 😞

tardy night
#

Can we have a few voice channels?

valid topaz
#

Manual and automatic lyric partitioning

you could write really long lyrics and it will automatically partition them (or manually)
then after the first generation you have the next set of lyrics saved and editable
and when you choose Continue for a clip it will ask you if you want to continue where you left off.
if yes it will keep going.
you could write an entire song out and have it generate it without needing to worry about cutoffs

#

could look like this

image.png
#

it shows up under the song and you can read it in full by clicking it and make changes

#

would also be cool if liking generated clips gave u 1 credit, so providing feedback is rewarding; maybe add a daily limit or encourage liking generations you prefer to reduce spam

prime rock
#

Is anyone in here because all my clips are gone now

#

Please please someone help me

gusty bluff
prime rock
#

I’m already there

gusty bluff
prime rock
#

Nothing, I tried that first, I even tried on someone else’s computer in case it was a memory thing

#

My music is no longer under my account

formal bear
mellow pulsar
#

why did i get error?

tacit idol
# mellow pulsar why did i get error?

You'll have to be more specific than that.
If you mean a red error box that pops up when generating a song on the website...
That usually means there's something in your lyrics that the filter doesn't like.
(copyrighted lyrics from famous songs, inappropriate content, or sometimes just an honest mistake)

If it's some other error, you'll have to clarify.
(also, the suggestions channel isn't really the best place for a post like this :))

little smelt
#

Test

deep kayak
#

Would be nice to know the SEED for the generated piece - as sometimes you mess one word in lyrics and can't recreate with the same music you liked before

weary turtle
#

I tried a quick search and didn't see this as a suggestion, so hopefully it's not a duplicate. It would be awesome if we could specify "how" particular lyrics are vocalized. Perhaps bracketed directions or something similar. For example, if I want a particular line shouted ("We won't take any more" [shouted]) or some similar expression would be helpful in really nailing down the feeling we want to express.

Also, more control over tempo. If I want to down- or up-tempo a particular section, like a lead break or a bridge. That would be awesome.

Thank you for all your hard work. Loving Suno so far overall!

stray rapids
#

stable diffusion started (like this) and then (((like this))) or [this] and now everyone is much happier with the (We won't take any more:1.6) style syntax

high wedge
#

I want for trending section to see trending news feed , in which the news feed is scrollable for public songs, like facebook, i notice people in groups when chatgtp 4 came out and when bing image creator came out, the fact that people could comment on posted stuff under a post it helped make community stronger. so as such trending public songs needs a news feed for people to click on recent posted songs and for recent activity and for relevant or trending songs in news feed. also each public song needs a comment section under each post.

rose seal
high wedge
#

i want to put longer prompts into song lol in create song section ❤️

#

=]

rose seal
grim solar
#

I want the ability to edit the title or lyrics of a song I've already created

feral aspen
rose seal
raven hamlet
#

Can one delete chirps on web portal, I know I can here on Discord but can't seem to figure it out on website?

feral aspen
raven hamlet
leaden marlin
#

Hey @fresh shard @torpid jetty, would you consider choosing time of songs up to minutes and 3x longer lyrics and style prompts? It would be gamechanger. I'd buy 10 pro plans same time! 🙂

umbral silo
#

since we can give gens title now, could we have option to give the title to things generated without it too?

#

basically edit title etc, so you could rename the title or add it later

rose seal
valid topaz
waxen sand
#

Volume bar on the site would be good so that I didnt have to lower the whole browser/pc volume or keep multiple browsers open so I dont have to change the settings constantly

pulsar flame
rose seal
pulsar flame
high wedge
#

lol what if sono a.i could read and or tell stories with custom music background and natural exspression in voice and with custom voices as well =[] the ability to make stories using more than one voice in a story for characters would be neat and the ability to make characters with personalities with certain voices so that it can be used in stories.

pulsar flame
#

Loved that the Linked Email is invisible from this area now.. (could have been shared with public while taking screen shot for discord sharing etc)

But to make it user friendly, can we have a white outline around avatar and some text in front of it, something like "My Account"

image.png
high wedge
#

the ability to download only vocals and only background music or both in a file would be neat too.

idle ember
#

Congratulations on the website changes. I have a suggestion: Present the continuations next to the source song, it is complicated to visualize and track when a song is continued after having generated others, that is, a nesting. On the other hand, I still ask for the DELETE option, your servers will thank you.

rose seal
idle ember
#

A permanent delete is necessary, it will improve visualization and interaction, since we will be able to keep only the really good stuff. I hope I have explained myself well about song nesting.

sage mirage
#

So THAT'S what the thumbs down means

#

heck yeah

rose seal
sage mirage
#

Is there a place this was all announced?

#

It's not a huge deal its just I didn't know this all happened, now I'm right clicking and seeing new things

#

hell yeah publics

rose seal
sage mirage
#

No way to title previously made songs it seems like?

#

The new 'Title' option seems locked on creation

rose seal
sage mirage
#

Hope everyone enjoys the nonsense names! ultrafast

sage mirage
#

Now to have a karaoke-style highlight follow along with the lyrics off to the right when a song is playing

#

aaaannnd maybe an option just like 'Liked' and 'Public' with 'Full Songs'

#

I'm sure it's in the discussion just didn't make the cut for whatever reason

#

Since Full Songs is a natrual way to want to sort clips

rose seal
sage mirage
#

Also looks like the 'Public' option is more a test of how the UI will work? The 'Explore' tab seems like songs you guys at Suno have made, correct?

rose seal
sage mirage
#

Well just so you know all my public ones are S++ tier and my ego requires other people like them to so sarcastic

#

I'm sure you guys are going to get so much garbage though, dang

sage mirage
#

Along with sorting clips by 'Full Song' or not, it would be handy to see if clips are continuations from other clips when they aren't apart of a full song

#

Sorting by continued clips, basically, as an option like 'Liked' and 'Public' currently is

prime rock
#

Okay guys you tried but I hate the new interface, I’m on an iPad and it doesn’t work at all unless you turn it sideways, and even then it’s confusing compared to the old one. I can’t see which clip I’m playing and I can’t tell which pair of song were created together which makes it hard to tell what’s continued from what

rose seal
prime rock
#

It’s kind of not good on landscape either, it’s all crumpled up in narrow columns

#

Before, the clip you were playing was at the bottom and there were columns above, now there are four narrow columns side by side, which means your lyrics get chopped and you can’t read them as they were laid out

rose seal
#

The left sidebar can collapse, etc.

prime rock
#

Just fyi this is what I’m looking at right now

IMG_9268.png
high wedge
prime rock
#

Yeah well I am a full time caretaker for my mom with dementia, it ain’t happening

rose seal
prime rock
#

So there doesn’t seem to be a copy this prompt option anymore?

sage mirage
#

'Remix'

#

They renamed it

prime rock
#

Ah that’s confusing, does it just start right up or just add the prompt?

sage mirage
#

Exactly as before

#

just shifts the prompt over to the window

feral aspen
sage mirage
#

Having a volume slider on the Suno website is more missed now since there certainly seems to be more variance in volume between clips, even with the same prompt

ornate niche
#

Suggestion to specify re-rolls to specific areas of a song... say I created a banger, except there's one more thing that would need to be added after the song is done - the ability to go back and add to your output, albeit in a specific time region, it would almost be like Midjourney's vary region function, but for a song instead of a picture.

Also idea #2 - would it be possible to prepend a clip? Like - sometimes the lyrics start way too fast, and I need to add a 30ish second intro.

sage mirage
#

Adding intros would be pretty sick yeah

#

Don't know how the thing works temporally, since you'd have to be ending at a particular point instead of starting at one

prime rock
#

Yeah it would be nice to continue backwards from the beginning, maybe 30 seconds instead of a minute - I have a great song which just starts bang no intro

compact lark
#

I understand that Now the Suno begins to generate from the words. However, it does not generate musicality at this point. And if I want to start with music, what should I prescribe?

misty blade
#

Congrats for the new ui!

Screenshot_20231215_105144_Chrome.jpg
feral aspen
# compact lark I understand that Now the Suno begins to generate from the words. However, it do...

  1. You can try something like this https://app.suno.ai/song/2fba092f-d2da-4042-9863-882360b4760d/ where you put [instrumental] as the first verse. I often put one regular line of lyrics in front of [instrumental] so the instrumental isn't skipped over.

  2. You might get a better result using two clips. First clip 100% music only. Then continue it with lyrics. (You may need to trip the start of first clip manually if too long.)

manic shoal
#

Hello guys! Suno is most developed ai product for music generting thank you! Can i get you API access for using this technology in my projects in Telegram?

compact lark
#

where does my music appear labeled as Public ?

image.png
feral aspen
tardy night
#

I would like to provide feedback on the new website. I really don't like it. I'm on the laptop, and everything is so small and not as user-friendly compared to the previous version. The previous website had a very nice, clean, and intuitive layout. In the old 'My Library' section, there were big covers with well-visible lyrics underneath them, making it easy to quickly preview which song it is, rather than just having a small cover with a style description like right now. Even the text/logo 'Suno' on that website was better. I'm very disappointed by it. Can we please have the old version back?

rose seal
tidal wharf
warm jolt
#

Not a suggestion for the app itself, but a community thing. We should do a weekly or daily theme or contest for the discord. would be cool to see everyones different generations based on similar themes.

lyric knot
#

I've been using Suno in its browser version.
I've noticed that the recent update has shortened the generation time. However, I'm not sure if it's related, but the quality of the output doesn't meet my expectations anymore. In particular, I feel that the quality of the vocals and the progression of the tracks were better before. If this change in quality isn't just my perception, I would prefer a setting that allows the use of the previous AI, even if it means longer generation times.
If this change is only my perception, please don't worry about it. Looking forward to continued great development.

ember dock
#

Lets say I just want the Accapella and that I can provide my own Midi files/loop notes watever , would there be a way to have the AI Sing on my beats 😮 ? that would be Amazing ( I mean even more, bc it already is mindblowing )

hybrid crane
knotty dust
#

Why is it that after upgrading to pro we no longer get access to the free daily credits? In some ways upgrading feels like a downgrade. Sure it gives 1000 more credits per month than free users (from 1500 to 2500) but once the credits are spent I have to either twiddle my thumbs for the rest of the month or have to pay for more credits. 😭

Would be nice if upgrading didn't mean losing out on access to the daily free credits. <_>
IMO, the simple fix would be if the pro plan gave 2500 credits in addition to the credits from the free plan, instead of replacing the free credit system entirely.

rose seal
tidal wharf
zinc gyro
untold pasture
#

Hello! just a sugestion, would be very good in future we have access to layers instruments.... change drum beat, guitar, FXs individually for example or mute instruments

zinc gyro
karmic crypt
#

I will be a customer of the first high-quality AI music generation app that allows stems download. Trying to split a track into stems using a tool produces poor results.

blissful current
errant cargo
#

“Remix” mode = same seed?

tidal wharf
#

Would be great to be able to reuse a seed. (Provided it’s a diffusion model.)

dusk venture
#

When buying extra credits, it doesn't seem like such a great deal as you buy more. What about something like below that gives subscribers a little better deal to keep on chirpin'?

Buy More Credits
500 credits - $4.00
1500 credits - $8.00
3200 credits - $16.00
6000 credits - $25.00

tacit idol
#

Can you make it possible to change a song's title after it's been generated, please?!

rain kestrel
#

Is it possible to implement a feature that allows writing out lyrics with timecodes, in a format like TTML?
I'm currently developing an app that can display lyrics in real-time as motion graphics. It would be great if we could integrate this feature with the app I'm working on.

https://www.youtube.com/watch?v=l7UDlXBT08c

YouTube
8bit
〽 iori demo

Demo movie for "〽 iori" app.

iori has the capability to synchronize lyric data, including timelines, in real-time with music playback, displaying lyrics actively through dynamic animations.
In this demo video, we utilize Apple's MusicKit API to showcase the simultaneous synchronization of music and lyric display through motion graphics.

▶ Play video
delicate cedar
pure fog
#

Another suggestion for remix - id like an option to remix with same key and tempo as a workaround tothe struggle of making changes evolve rapidly in one clip

leaden marlin
#

Hello. I bought a subscription. 10 thousand tokens. Why can't I download the instrumental version of the songs I make? When will this opportunity appear?

amber violet
#

Suggestion: Please add some organization of content. Allow for keyword with search or adding to folders. 🙂 Thank you!

sage mirage
#

Custom tags alongside the current 'liked' and 'public' would be nice and match the current UI at least from a customer-facing POV
Then I could tag a song I'm working on since 'Full Song'; is a misnomer, it really means "at least two clips"

rose seal
sage mirage
#

I don't know what your guys' back end looks like but it would just 'fit' in the interm

#

The new compressed UI makes it faaar easier than before to keep track of clips so it's welcome all the same

pure fog
#

Suggestion : 1 "Define durations". Say you have a 30 video clip and want "short intro- motif- outro " to fit that specifically.

#

clips are limited to selected time and model understands how to structure it correctly

#

clean steam separtion is a missing link with all apps atm. i saw someone mention dawxpro just now and going to try it.

sage mirage
#

Ala MidJourney, how about a chaos slider? Weird slider? Something like that, so you can push the model more in an experimental sense if you'd like

#

I experiment allll the time so for me that would rule

#

Then when you find a clip you like, click it back down to keep the output more inline

pure fog
#

SUGGESTION : ability to weight the affect of each style.

#

or choose a primary style (model follows that structure] and secondary styles [ tries to maintain song structure of primary while introducing the musical elements of the other styles,

#

the more disparate you get the more it struggles, as youd expect. but is there a way to weight how the model interprets each metatag ?

#

So its like a "groove quantization'. heres the groove of the primary tag. [ apply and expand that using musical elements of the secondary tags

umbral coral
#

Yo, we need the bouncing blue bars back on the videos.

hybrid crane
#

Suggestion: be more clear what happens with left credits on a monthly plan when it renews. There seems to be some confusion about this.

tidal wharf
#

It’s not a given that the model produces stems at all, it could just as well giving the entire master track, like eg Jukebox

#

It helps to not think about music ai in conventional daw terms

tidal wharf
#

I’m describing how the model works (could work), not the perfect world 🙂

tardy night
#

Could I request a report on unintentional instrumental tracks? This would allow for an analysis of these tracks, potentially revealing the underlying causes. I don't know how the model works behind the scenes, so I have no idea if that would be helpful or not. I've noticed that cinematic tracks often provide predominantly instrumental clips, which is understandable given that this genre typically lacks vocals. However, it's worth noting that there are many tracks within this category that feature powerful choirs and vocals, especially in epic orchestral music.

dusk venture
valid topaz
umbral coral
#

train this thing with some modern rap/trap beats

sage mirage
#

How about the ability to pick what picture represents the completed song? Each clip gets it's own picture, so let us pick which one out of that sequence. If you do a naked [Outro] as your last clip it's always a bird

valid topaz
#

edit lyrics in-place to fix typos or mistakes, add or remove endings, or rework sections, maybe for free?

exotic crescent
#

i love Suno so much- I need Bark so badly.

tidal wharf
#

A small one: forward/backward by 10s

verbal geyser
#

Would be great to have the ability to upload an instrumental and have Suno put the lyrics on top of it.

eager fulcrum
#

Easy to implement:

  • Cut out x seconds from beggining and y secs from end. Sometimes it bugs out and the music begins after 10-30 sec of instrumental, or lyrics are too short and it sings nonesence and/or does solos. Can't give someone link to share, have to download and do it manually.
    Hard to impelement:
  • Regenerate parts of the song. Select part, cut it out, and knowing before,after and length, try to regenerate the missing part, giving us 4 variations, to chose 1 from.
woven merlin
#

The compression level is too high for most of the songs. I wish you could have an option for the WAW file.

lyric knot
#

I would like to request the following modifications to the outputs:

  • Reduce the frequency of generating vocal-free outputs.
  • Ensure consistent melody lines in [verse] and [chorus] throughout the song, even when using 'Continue From This Clip'.
  • When using 'Continue From This Clip', the lyrics should continue seamlessly.
  • I would like the ability to adjust the level of variation between verses and choruses to avoid monotony.
verbal geyser
#

Would be great to have the ability to delete songs we do not want in the web app

sage mirage
#

A simple editor to clip bits out of songs, even if just the end. You could clip off silence bits to continue clips, or clip out endings that suddenly surge with unwanted audio

cosmic night
#

I would like to combine the songs I made for "Continue From This Crip."

proven obsidian
#

Would like the ability to use the same voice model on different songs or implement my own voice and have it tuned to the music.

ruby magnet
#

I want downloads vocal stem data of them.

frozen iron
#

I'm wondering if the ability to add a placeholder structure at the top of the first clip would help the engine not generate wierd stuff out of place... For Example:

[Begin Structure]
[Intro]
[Verse1]
[Chorus1]
[Verse2]
[Chorus1]
[Bridge1]
[Chorus2]
[Outro]
[End Structure]
Then the actual lyrics would follow with adornment as in:
[Intro, Solo Guitar, 4 bars]
[Verse1, Male Lead]
Blah, blah
[Chorus1, Male and Female Duet]
Woo, woo
[Verse2, Female Lead]
Bleh, bleh
[Chorus1, Male and Female Duet]
Woo, woo
[Bridge, Guitar Solo, 2 bars]
[Chorus2, Male and Female Duet]
wee,wee
woo,woo
[Outro, 4 bars, fade out]

The structure would "pre-hint" the generator as to how many verses and choruses there are going to be, what choruses should be stylistically identical and which might be different, and where any bridge sections are going to be located and where the end of song is located... this would prevent the large number of generations we get where the engine randomly seems to put in alternate rhythms thinking that it's time for a bridge, or goes nuts sticking "endings" in the middle of the song making it nearly impossible to do a continuation from that point

wanton leaf
#

We need folders, to sort songs

#

and stems )

harsh spoke
#

I would like to suggest an api for this.

manic shoal
#

Hello! Thanks fro awesome product suno ai! Can i get api acces for my project ( AI converstion telegram bot 15k users ) ?

tulip fox
#

Changing the song title post-creation would prove beneficial. I neglected to alter the title, resulting in an excellent song saddled with an entirely inappropriate name.

grizzled cosmos
#

Stems, and longer song lengths

hybrid crane
#

Suggestion: beta-testers who earn credit for doing specific tests or actively achieve goals (bringing in audio clips with bugs consistently)

cyan flint
#

we need a channel dedicated to post suno processing so we can see what everybody is using and doing to get the best sound possible.

#

having vocal and music in seprate stems would be great of course

verbal geyser
copper hamlet
#

please add API usage in the near future 🙏

sage charm
#

переведи на мой язык

young basin
#

We need to EXPAND the “like/dislike” system (!)

in my example - I've been trying to generate 1 song for a whole month... the results are different, I can't catch the "best" (!) 😦
in one result - the chorus turned out well, but the verse turned out very poorly (!)
... in another result - the verse turned out very well only in lines 1-2-3, but the 4th line of the verse turned out VERY BAD ... therefore - it is very difficult and very long to catch the desired results (!) 😦

(!!!) Suggestion (for site) -->> add a RATING FOR EACH LINE in the resulting result (!)
To - by marking “like/dislike” on EVERY line (in the finished result) == change the “weight” of each element of the song (!)
So that, as a result, you can very quickly get the DESIRED RESULT (!)

Thank you ! 🙂

rose seal
#

Thanks all, keep the suggestions coming (we do read them)

tardy night
#

Would it be possible to have an option to replace lyrics while keeping everything else the same? Of course, the new lyrics should match the flow of the original ones to maintain the melody and rhythm, but in a different language, for example. That would be very cool and could help change some words or lines that don't fit well.

tidal wharf
rose seal
rose seal
elfin mortar
#

  • It would be cool if it was possible (in custom mode) to provide different style prompts for instrumental and vocals. An option to provide a prompt not only for "Style of music", but also "Style of vocals"

  • Would also be nice if generated audio were given a unique seed that could be re-used in future "style of music" prompts. This would make it possible to generate more songs that sound like they were "played by the same band". Sometimes I've accidentally generated something unexpected that I really would like to explore more by changing lyrical content/syllables, but reusing the same style prompt rarely gives me the same specific sound and style, only generally similar

  • Would love to see an implementation of spleeter (or something similar), for separating vocals, drums, bass and other instrumental parts into separate downloadable audio files. Could be a premier plan feature maybe?

wanton leaf
leaden marlin
#

This is the most awesome AI I've seen so far.
The biggest feature you could implement is : Keep the voice

Sometimes I would like to do multiple songs using the same voice but not the same instrumental and it does not seem possible as far as I know

median coyote
timid wasp
#

Suno.

sage mirage
#

A separate prompt box to specify what kind of image to generate for the cover art. At least this way I could avoid the deluge of birds. I assume it's a totally seperate thing that just reads the lyrics prompt anyway

sage mirage
#

When you 'Continue From This Clip' and the clip shows up underneath the 'Continue From' on the left side of the screen underneath the 'Custom Mode' creation part? It's great that you can play the clip from there without resetting any UI, but it would ALSO be great if it populated itself in the far right box that normally so I can see the lyrics it contains. Right now it doesn't, even though intuitively it seems like it should.

rose seal
turbid rover
#

Please help me!) how to download music in .wav format? or only mp3 and mp4? How do you download music in good quality?

loud kite
delicate cedar
#

Would it be possible to change the color of font where the verse was actively being sung? The line if the exact word is impractical? That would make it easier with certain things to know exactly where Chirp was in the song. It may even be illuminating if it turns out that some of these instrumentals we're getting are simply the vocals rendered as instruments, in letting people complete their instrumentals more perfectly.

tight jacinth
#

Boa noite 🗣️

#

Eu estou conseguindo fazer música, não sei o que está acontecendo com o Suno, alguém pode dar uma sugestão

west orbit
#

Hi all and congratulations for the amazing project!

#

My modest suggestion is this:

#

Please allow to delete a song from the list once created. It is very frustrating now as a user not being able to delete a song

#

I can only "thumb up" "thumb down" but not remove something that I created from my list

dusty dove
#

what about 5 minutes song lenght? Please, we need more lenght in our songs.

west orbit
#

@dusty dove you can do that by "continuing" existing song already

dusty dove
rose seal
urban sleet
#

I'm sure people already suggested it, but "continuing the song backwards", even just one time - ergo making an intro, would be a life saver!

twilit sphinx
#

hi everyone! Tell me please, how to make song longer? i have 48-60 seconds

delicate cedar
misty blade
#

It'll be nice if auto-generated lyrics will be also for the Continue option according to the start

faint hollow
#

I have an idea. If you write your lyrics and there are songs with the same lyrics, you can listen to other songs with those lyrics. An example:

Other_songs_with_the_same_lyrics.png
blissful current
#

When generating a song from multiple continuations, it can be difficult to tell how many lyrics to enter in each generation. For instance, if it generates a slower song, the added padding from instrumentals and slow vocals can lead to lyrics getting cut off at the max length; but if it generates a faster song, inputting only few lyrics leads to very short generations, and more of them, which costs more credits.

It'd be nice if there was some sort of analysis that would give us a message of "Too Many Lyrics: This will get cut off in most generations" without charging us credits in those instances...

misty blade
feral aspen
sterile pendant
#

gib api kk thx

quick olive
#

holy shit guys, i've been following this project for 4 weeks and the quality of the songs improved so much... congrats! whatever you did to reduce noise of the music and quality of the voices is working great! Only suggestion i have at this point is being able to change the lyrics of an existing song (i have some gems where the music is awesome but the text has typos or needs minor changes) instead of remix it...

tawdry dust
#

OMG!!! I am loving this!!!!!!!!!! when api PLEASE!!! this is legend!!!!!!

prime rock
#

It would be a nice addition to the interface if you could see when looking at a continuation what song you continued from, it’s sometimes very hard to keep track of and go back to an earlier step if the path you’re on isn’t working. That information has to be stored already somewhere since the program can put a full song together.

covert oar
#

Not sure if it has been suggested but being able to save singers and beats/music would be excellent

final mural
#

More meta tags to really set what part of the song you want to generate. Intro, bridge, solo, chorus, instrumental, ramp up.
So it can be used as musicians builds songs.

#

Maybe also to generate each instrument by itself. So we can generate drums for different parts, separate bass, guitar channels etc.

#

reuse generated parts later in song. For instance the chorus

balmy mountain
#

Can we generate only vocals yet without music or instruments?

delicate cedar
balmy mountain
lean zealot
#

Greetings guys, you have a pretty good music generator, however unfortunately it has one major flaw.
Many of the generated melodies have a song in the background.
Unfortunately, it is not possible to remove it even with special promts or filling the box with the desired text with spaces.
please make it possible to generate a melody without a song.

sterile solar
#

Hey guys just discovered your app today and wanted to pass along the feedback that there's a real lack of genre variety and detail of many "urban" genres like US House Music, Trap, Hip-Hop, etc. I think having someone with a more in depth knowledge of these styles of music would be very helpful in making your app as true-to-music as possible. Currently you're killing it with Pop, Country, Etc. I believe that's because your people have an intrisnic knowledge of these genres. In doing my trial I found all my songs to be amazing feat of technology but technically incorrect and disappointing. You obviously are on to something special and I thought I would share my feedback.

gentle wasp
#

I assume ChatGPT does the lyrics, but nearly every song starts with a setting-phrase of “In the _____”. I know we can customize our own lyrics, just saying.

hasty swan
#

Remix that actually keeps the music style. Is there really no way to stick to a style that I found?? --seed? Only on discord maybe?

#

Would be cool to have a MIDI file so I can start right away with recreating the song in a DAW. Probably impossible: Multitracking of the song so you can e.g. only use the voice and recreate the drums yourself

leaden marlin
#

option to make songs longer

tidal wharf
tidal wharf
dusty dove
tall jungle
#

Lets work

leaden marlin
shrewd socket
#

Would like to suggest a no vocals command

pulsar flame
# leaden marlin ???

for now you can use the "Continue from this clip" option to make longer songs.

but I agreee longer song creation would be really helpful. (3 minutes per clip maybe)

leaden marlin
#

thanks

#

it works !

leaden marlin
#

It will be nice to have extra African language support like Akan, Twi, Ga, Ewe and Hausa.

viral breach
#

Make a comment section so people can talk about songs when they arrive to a link.

#

Also while I'm thinking about it, a clip arranger would be awesome.

two rows, top is recent clips, bottom is a song. drag clips to assemble a song, let me preview it just playing the files in order. Also allow to fill blank space between with generation, starts and ends appropriate to blend

elder swallow
#

Hi, 1. You should accept Paypal for subsription payment 2. For customer protection its already industry standard to use Segpay (or any similar secure online payment processing) You should upgrade this as soon as possible.

radiant current
#

Hi, anyone know if its possible to listen more than 1:20 of the generated song? subscribing perhaps?...thanks

inland frost
#

Is there a way to alter lyrics on a song but keep the sound style that was produced? I don't want to extend, I want to fix some lyrics that aren't pronounced correctly.

blissful current
blissful current
honest flume
#

Super impressive. Please please please launch an api!!

rotund mulch
#

Clear dm history

tidal wharf
tidal wharf
#

Here’s one that’s super easy to impelement but will have a drastic effect on creativity:

Assuming it’s a diffusion model, show, and let reuse, the seed. Imagine the possibilities writing the same song with different lyrics, slightly different style, etc.!

potent heron
#

Loving the magic here, Let us add a deadzone. the songs end the instant the lyrics drop. let us add time to extend the instrumentals or a better way to define the end to a song.

dusty dove
winter zephyr
#

This has probably already been suggested and not sure if this may already currently be possible somehow? Specifying how music should be produced before it actually happens: It would be great to have the option to specify how exactly we want various parts of the songs to be produced. As an example, perhaps specifying the verse to be a male voice, and the chorus to be a female voice or having the option to change speed of a particular verse, for example double-time very common in rap genre.

fierce orbit
#

As a future Web UI Feature I would love to have the option to delete a song entirely.

gaunt creek
sly fern
#

Is there a way/plan for higher bitrates ? Orchestral music, metal or high pitched voices end up a with quite a bit of static, but at the current audio bit rate and compression levels, that would be unavoidable, even if the AI generated perfect sound.

feral aspen
sly fern
sly fern
feral aspen
gaunt creek
#

I wish I could specify that I want an instrumental song without lyrics in the web UI of app.suno.ai

feral aspen
gaunt creek
hybrid crane
#

I see people asking the same questions, over and over. I personally think an onboarding tutorial would help (in the web app).

delicate cedar
#

Some of these things are even listed in Discord when you join but people never read them....

hybrid crane
#

@delicate cedar If it is an active tutorial, where you have to click to see the next hint, it will help. I am convinced it will (due to my professional experience). And not everybody is on discord. 😉

delicate cedar
#

True. But the ones we see here are.

hybrid crane
#

I just think it would help. I have less patience I guess, I would rather see more indepth questions and not a record that keeps repeating.

sly fern
#

One WebUI improvement would be to have hierarchical display, so that continuations appear under the song they continue. After a few branches, it's kinda hard to keep track with just thumbs up.

delicate cedar
charred fulcrum
#

No suggestion, but damn what you've created... Congrats! 🔥

silk cargo
#

Suggestions:

  1. if the lyric is too long and it cant finish all of it, We be able to add the remaining on continue and it start from there.
  2. more controls, right now its limited to [verse] and [chorus] please add [intro][outro][bridge][drop] and [end]
  3. on [chorus] when lyric is same, Please keep it consistent with repeating those.
  4. the voice should remain same in all 3-4 mins song, at begin its crystal clear but at the end it becomes another voice.
  5. I know this one is probably not possible, but I wish we could use singers voices, atleast those old ones which are in public domain.
frozen iron
#

I don't know how hard and fast the 1 minute continuation timer is, but I've trying to crank out a really slow ballad and even with short verses and chorus I'm having a real problem that I'm hitting the 1 min mark before the chorus is done. I tried splitting the verse from the chorus as two separate continuations but then wierd stuff starts happening with the style and rhythm, it really works way better when the verse and chorus are kept together. Maybe there's some way that the time limit of the continuation could be tied to the BPM or words-per-minute so that really slow songs get more time and really fast songs get less?

frozen iron
frozen iron
stray rapids
#

with so many people dumping in generated lyrics from freshbots, there really ought to just be a regex to automatically convert (Verse 1) to [Verse 1] and so on

honest bridge
#

Previous suggestions I agree with (+1)

  • Higher quality exports (lossless would be great). On MidJourney it costs credits to upscale images, maybe a similar thing could be done here?
  • Tagging system that allows filtering
  • Allow cropping of clips, and the ability to continue from those. Someone else framed it as using a timestamp as a point to continue from, but regardless of the implementation details more control around the edges of audio clips would be wonderful.

New suggestion

  • Similar to the tagging suggestion, it would be great if users could create "stations" by tagging certain songs and then separately make that tag public, and then the tag URL could be visited reliably (a station, basically). I would use my different stations all the time depending on if I'm trying to work, relax, have a laugh, etc. (Edit: I also imagine some stations might become quite popular!)
dull sandal
#

would be great if it could take an instrumental composition (partially built) and allow remixing it by adding matching vocals or instruments or effects 🎸 🪇 🎹

honest steppe
#

Delete or hide music from the library

misty blade
honest steppe
#

Still appears though in certain places

#

It doesn't fully remove at all

twilit hull
#

The fact that pressing enter on the prompt (outside custom mode) submits the prompt to generate a song made me accidentally generate two songs 😭

silk cargo
#

Suggestions:

  1. Please let us know separate seeds for the generated voice, the song, and the style of music. Sometimes, a golden voice is discovered, and I wish to create a different song with that voice.
  2. Besides that, if this is technically possible, having separate seeds for the voice, the song, and the style of music would be great
serene compass
#

I really wish [Fade out] was a universally respected meta tag. Way more often than not, the song is finished, but that request is not respected when it appears at the end of a set of lyrics, causing me to have to waste credits on just a fade out.

visual dagger
#

My Suggestion Is That Suno AI Will Have a Remix/Cover Music Component: "Suno AI Remixer" The Premise Is That This Component Basically Takes MP3s of Movie Soundtracks, Suno AI Songs, Music & Video Game Songs And Then Converted It Into a Suno AI Cover Using Genre Styles That's Pretty much What It Says on the Tin, Be It Heavy Metal or Dubstep. Examples Includes In the Final From "M&L: Bowser's Inside Story" Being Converted Into a Japanese Anime Song and Infinite From "Sonic Forces" Being Converted Into a Post-Grunge Rock Song.

tropic cave
#

I'd really like to be able to continue from clips that end with enough silence that it trips Suno up. I have a first verse and chorus of a choral song that is absolutely perfect, but there's just enough silence at the end that continuing from there gives errors and refunds my credits. Either better logic by the AI to continue, or an option to trim silence from the end of generated clips, would be greatly appreciated.

unreal fossil
#

Hello, I believe this has already been said but:
I would definately pay more credits if I could do longer generation ~3min (or even more). Even if it can fail in the middle. It's just worth that much to be able to keep same voice, rythm, "feel" of the song

P.S. Blown away by Suno!

rose seal
red bone
#

I'm not sure if this idea has come up before, but having the option to select either a male or female voice would be fantastic. Take, for instance, the song I'm currently working on, 'Hold On'. The seamless shift from a male to a female voice in it really enriches the depth and feel of the song.

delicate cedar
#

Yeah, that really would be great if there was a way to do that on command...

#

(I just don't know that there is a reliable way to do it just yet.)

tidal wharf
#

More of a question than a suggestion: Is there currently a decoding capability in the model? Can it take a real audio and turn it into tokens that can be then continued, the way Jukebox can?

tidal wharf
delicate cedar
#

I didn't think so.

tidal wharf
tidal wharf
vapid basin
#

I was looking through the terms, thinking about signing up. But... I'm wondering if there is any provision that might allow for converting songs made in a free membership to commercial usage? Does anybody know off-hand?

Also, it's very unclear as to what happens if you start a song on a free account and then continue it with a paid account? Is the first part non-commercial??? Or, does continuing a song with a paid account convert the song to commercial usage?

#

Oh, SUGGESTION: Allow us to change the titles of existing songs! (especially the ones that didn't have titling ability when they were made)

blissful current
tidal lagoon
#

Probably something low priority for now, but having the ai understand more obscure and niche genres will be amazing. Like, metal music is a whole world and at the moment it understands things like "death metal" and "black metal" to be melodic modern metal. Although the results sound great honestly. Also an option for it to understand when you want an instrumental song, not just to make tracks to sing over, but also for obscure genres like dungeon synth or drone and stuff like that. But so far, this is mind blowing.

red bone
delicate cedar
jade cave
#

Please allow the ability to edit the Title of a track on the website interface. I often forget to change it (my own labeling and numbering system), and I can't seem to change the title once it's added.
BTW, I've been here since early August, and I'm blown away by the last month's generations. I have about a 90% acceptance (Like) rate with what Suno generates. Well done Suno!

crisp hearth
#

The word "girlfriend" is not allowed in prompts? I guess because somoene put in in a song title? I'm having a lot of problems generating music with innocuous prompts because it looks like a lot of dictionary words are blocked, due to being song titles. Not sure how to write songs about my GF now haha

potent heron
#

IDK how you guys could manage, but sometimes its hard to get a good voice and when you do different lyrics would compliment that voice better then your original plan, maybe a seed for the voice it generated

twilit niche
#

The title idea is something i wanted to suggest as well, but i also wanted to add the suggestion to be able to modify bits of lyrics in a song without remixing the whole thing, i want to keep the rythm and the vibes and the way the voice is singing, i only want to correct minor things in the lyrics, not remake the whole thing

vagrant rock
#

How can I do filemanagement in the web version? I want to delete useless files and rename some others and also be able to create folders for my projects...

#

It's a mess right now

velvet token
#

amazing. possible future developments: capability to create just the arrangement withouth voice (or part of the production like just one or some instruments) describing in input the mood and the style.

tidal wharf
#

a small but useful one: write duration above every chirp in the list

tidal wharf
velvet token
vapid basin
# delicate cedar You can ask but my understanding is that the answer is by default, no, as is sta...

That's kind of a simplistic way of looking at it if you're just talking about this part:
"If you are a free user, then you can only use your songs for non-commercial uses, so monetization is not allowed. If you would like to use your songs commercially, you would need to subscribe to one of our paid plans, Pro or Premier. For more details, see https://app.suno.ai/account:"

That actually doesn't specifically say that a song that was on a free account couldn't be used commercially. In fact, it makes it sound like if you would like to use your songs commercially, you would need to subscribe. It does not say that the songs from the free account are barred from being continued on a paid one. And, as I already stated, it is unclear what happens with continuations of songs. I wouldn't have posted anything if it wasn't totally unclear.

clever escarp
#

From the work in this link, continue from this clip
https://app.suno.ai/song/23c58fd3-b75b-4db9-b843-7572f7016041

I made the next piece, but the lyrics weren't made until the end

https://app.suno.ai/song/9c60b98a-862c-41ca-879c-9c41c4b217ae

This is the lyrics of part 2, but the work of part 2 has been completed up to the red part.
There's no work on the blue part

[Verse 2]
What they say, men do
The word that you break up right after receiving a gift
That kind of trashy behavior, is that really a man's behavior

[Pre-Chorus]
I'll say goodbye to my friends
You're so different
You've been a public servant for four to five years
You who put work before me

[Chorus]
I loved you for seven and a half years
The memories remain bitter
But we met a breakup
This moment when it all ends

[Outro]
It's time to say goodbye. I wonder what you've been doing for seven and a half years
I don't think I can wish you happiness

If you ask me to check the above, I would appreciate it if you could reply.

씁쓸한 이별 | Suno

acoustic pop song. Listen and make your own with Suno.

▶ Play video
씁쓸한 이별 | Suno

acoustic pop song. Listen and make your own with Suno.

▶ Play video
delicate cedar
vast terrace
#

This tool is amazing. It's a bit painful for trying to create an entire song, though, and now my library is filled with parts songs.

young basin
# young basin We need to EXPAND the “like/dislike” system (!) in my example - I've been tryin...

Add to site:
Considering the many generations that are very similar to each other, but still have some differences in different lines, I really want to be able to == “choose which lines I want to collect into a “complete song”” (!!!)

Because it’s very difficult to catch the “ideal” option (!)
All generations (remixes that continue the 1st part of the song) are all different! in some results = some lines turned out very well, and in other results - completely different lines turned out well (!)

Therefore, from all the variety, I want to choose exactly the “ideal” lines in my opinion (!)

pseudo surge
#

Can we get volume control?

#

Also great job with the new setup. Thanks for adding deleting option

wanton leaf
#

you mean unlike ?

paper echo
#

Hello. I just subscribed today and I've noticed that using my own lyrics is really annoying... I understand that copyright etc... but I'm just inventing my lyrics, and for sure there's no copyright issue, but it says "Suno is for generating original music"... It's really annoying, because I'm also not using any artist name or quote... Could you give Us a bit more freedom please?

#

It is already as annoying as censour in Roblox chat... I'm changing the words to different ones and still "Suno is for generating original music"... soon We'll be not able to even use any words 😦

#

I wanted to extend the song but now I just cannot 😦

pseudo surge
wicked pike
#

We need some sub folders to organize our projects. Like keeping the continues in a subfolder attach to what you are continuing off of.

radiant sedge
#

no lyrics option please

green meteor
#

Is there any plan to include an option in the app to download vocals and instruments separately?

frigid timber
vapid basin
frigid timber