#📚┃suno-school
1 messages · Page 16 of 1
I got a question, according to the voice of the A.I. Are they protected ? If we want to put those music on youtube ?
See #⏩┃continuation if you’re using the Discord version. The web app has a different method but I’m on my phone at the moment so bear with me while I find the link for you.
See here for the webapp: https://suno-ai.notion.site/FAQs-b72601b96de44e5cacd2cd6baa985448?pvs=4
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
See the FAQ I linked above for an answer to this question too.
Okay, thanks alot
Same. Trying [End] or [Break][Silence][End] after previous part and is works. Sometimes I just add 1-2 chirps as instrumental.
Each next part is less similar to first - including deeper, metallic vocals, same instrumental. Parts 7+ are totally different than first.
Hey, I currently prefer using Suno via either Discord DM or a private server with the bot.
In part because this makes it fairly simple for me to organize and navigate generations, but also mostly because I've never really used the web version.
I noticed on the web version there's the option to download mp3 instead of mp4, but I'm having trouble figuring out how to find my finished tracks generated via Discord on the web version.
I also can't figure out how to extend tracks when creating via the web version, it seems like all I can find there are single clip generations?
It would be cool if I could figure out how to get the mp3 or better for my full tracks though. Not sure what the best filetype would be but would like to run the Suno generations through RipXDAW for further processing before publishing and it won't accept mp4. So I currently need to either convert the .mp4 to something else, or find a way to get something other than mp4 as my source file.
Anyone have any advice? I don't yet know much about audio filetypes.
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
/Hi sir can i create sinhala voice song , your website can create but discord its not working can you fix it?
Discord ha v1 version, v2 us only available on website so far.. it will be updated soon but... no one know when
I'm about to subscribe the monthly plan. Will Suno delete all of my previous songs (the ones created with the Free plan)? Thanx
It did with mine.. so I advice a different account
but what if I want to get the rights on my song? I know the subscription can't give you the rights in a retroactive way but what if I want to continue the first (free) version of the song in a better way AND to get the rights of the result (because I've subscribed the plan in the meantime)?
Technically you created that with the free credits, so you are not allowed to monetize them.. but I am not a laywer.. better ask them directly.. I believe you can mail them on support@suno.ai
thank you!
Please! anyone, I am newbie here, can anyone tell how to find all my generated songs?
Is there any way to get more gainular on the style prompts? BPM, specify two-bar rhythm for drums, etc Just seems limited in # of characters for the style.
or is there decent flexibility in meta tags (more than what is listed in faqs) added to the lyrics section to guide any changes to the music?
Try here: https://app.suno.ai/me/
and the previous pages here:
2
I'm trying to generate a song, entered the text that I translated into English chatgpt and SUNO writes that try a different text. What is the problem, if there is no such song in English?
If the song is under copyright I'm not sure it matters what the language is.
I'm going to assume this is a glitch though and the text is incorrectly triggering the message. Have you tried the lines of indeterminate length trick? Just add several lines of spaces so the profile shifts and Chirp thinks it's a new song.
Another example:
And I added [throbbing vein] to that one because the trick wasn't working, but with the addition of a new metatag it went through. So ....experiment?
¯_(ツ)_/¯
is there a way i can use paypal to pay for my suno pro?
i cant use my credit card for online purchases since im under 18
are there allready solutions for some of this possibly functions: - to use the allready created beat for another text/song soon
bcause if the text dosent fit the beat is gone, dont matter how wild it was
- or just change some textparts in the exsisting allready generated chirp
- or create a part in front of the selected one instead of the next following part
Is there a way to keep the same voice for different songs? I tried to use "Continue" with a Part 1 and gave the prompt "Change vibe and theme but keep the same voice" and it didn't work 😟
Just a question, is there a way to FORCE the app to use a specific Language?
For Instance.
if the Song Started in English, If you Use Japanese Characters it does do Japanese.
But in the next part if you want type the next part in English, it Refuses to switch back.
It takes the English and Auto translates it into Japanese instead
Hey everyone, new here. Quick question: I'm sure this has been asked many times but is there a prompt to create vocals only (no music) thanks in advance
@weak geode that is hard to do. Many tried to find a consistent way to do that, with several methods. I suggest you try the search function here and look for: acapella in: 💬┃general
its impossible for me so far lol
yeah.. it's more a happy accident when it happens
Is there a way to choose male or female vocals or even a duo?
try ur luck with [narrator]
you can use [Mary] and [John]
This is just brilliant Cat! I’ll test it tomorrow. I was forcing to end by [Silence][Stop][End] and it usually works on ending part better than just [End]. 😉 I’ll test yours, looks great!!
No, sorry. No seed number. 😦
LOL
I ran into something after I tried that one there and it was doing the same thing so I just added [throbbing vein] at the end on its own and it worked. If it didn't fail for someone else who tried it I'd have thought I found a skeleton key of sorts.
If it’s glitch and f.ex. I’m 5th part copy of copy of copy 😉 of first part doesn’t sound similar, all you can do is to return to first part and continue from first with new text, and then download many parts and try to combine it in good steps using Audacity, Audition, etc.
If you wait too long, I noticed Suno forgets old song styles / seeds.
You can try:
[Verse]
[Mark] Lyrics words and so on
[Anna] Lyrics words and so on
[Mark and Anna] lyrics lalalal and so on
But you need to generate couple of times, it rarely works.
And Suno is so politically correct, doesn’t discover genders, and last time my Anna was male and together or Mark only, were metallic voice of another girl. :):)
Just test. You need to be lucky with that, especially adding second voice when you don’t put meta tags with names in first part.
I’ll show you my 3-lines-tags when I come back to office tomorrow, you will get crazy 😂
how do I tell a song that I do not want it to end?
i want to be able to generate the rest of the song without cutting it off mid-lyric, but it generates an ending to the song automatically
You tried continuing the song? https://suno-ai.notion.site/b72601b96de44e5cacd2cd6baa985448?pvs=25#baf0eb82963f467488f5b6f167ba8102
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
yeah
Is the song ending on a quiet note?
it ends abruptly after the third chorus
i managed to brute force one that did not end after the chorus
and continued from that one
the other version i generated always ended after that first line
You may need to go back up to the third part and start working from there.
Used to have to do it that way in the Discord version sometimes. Hated it but when you work yourself into a corner there's sometimes nothing else you can do.
yea idk, i just tried continuing like 10 times
one of them eventually kept on going so i used that one to add the little extra bit after the chorus
Yeah, going by that ending at the end Chirp thinks the song is over. It may never continue from there or you could try again and again and maybe get lucky. But it's probably better to go back to an earlier part and try again from there.
Hi, I am developer and would really like to implement Suno into my project. Are APIs/documentation available to integrate into my project?
No unfortunately I think.
What do u create BTW?
There is a new upcoming project in which I will take input from users and generate a poem and a song dedicated to the poem, basically for Valentines.
Hi, I have two questions.
- what genre would be good to create a company song?
- when I create a song, it cuts off before it's complete, is this because it's a free plan?
- company song.. like muzak or a jingle?
- happens. it generates certain periods of time (up to, i think, 1 minute 20 seconds), then you have to continue it from there
Hello friends, I made a perfect guide to SunoAI on Youtube. I hope it helps.
"🎶 수노에이아이(SunoAI)를 이용해 당신만의 음악🎵을 만들고, 수익화하는 방법을 단계별로 소개합니다. 챗지피티를 이용한 노래가사부터 SUNO AI를 통한 음악 창작🤖, 플랫폼에 업로드⬆️, 수익 창출💰까지, 이 모든 과정을 이해하기 쉬운 가이드로 만나보세요. 음치들도 쌉가능한 음악 제작🌟 꿀팁 시작합니다.
thanks
@cold zenith You're welcome. If you have any questions about use, please ask in the comments on YouTube.
Imma do it thank so much
is there any limit for using chirp like the studio?
Hi,Jesus,
Thanks for your reply.
- Company song refers to a song that is sung in chorus by all employees during company events. The lyrics usually represent the company's philosophy or spirit, and the first and second verses are usually sung in full.
2.Please explain in more detail how to continue the song when it breaks at 1:20.
It seems that after three or four continuation, the vocal's voice would start to degrade to artificial AI voice, what do I do to prevent that?
i don't know
See here for directions on how to extend your song: https://suno-ai.notion.site/b72601b96de44e5cacd2cd6baa985448?pvs=25#baf0eb82963f467488f5b6f167ba8102
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
reroll each continuation to be sure good quality voice is left in last 20 seconds of the clip, so new clip has better reference
Trying to back to an old version, found it on the site... How do I continue it? The link isn't there to keep it going.
All I see is remix. That's REALLY shitty if old songs can't be gone back to to continue. I've tried recreating a similar opening over a dozen times with no effect.
Unfortunately the old version of Chirp is not compatible with the new model. This extends to the chirps produced with those models.
clips made on discord (v1) are currently incompatible with the webapp (v2)
You can try to continue the song from Discord (for now) as the upgrade hasn't taken place yet in Discord and the older model is still available there.
Boooo. Exactly what I was wanting to do - the hook is perfect but the rest of the song on the old version sounds very auto-tuny.
Time to spend 1000 credits to get a similar hook.
I really wish they'd integrate a way to upload clips for a "similar sound." The tech is based off identifying song patterns, right? Why not implement that directly into the UI?
I have no idea how they do what they do. For all I know they simply generate artwork of waveforms and send it to be converted to audio somehow via OCR.
(Silly Disclaimer: The above is pure speculation, an "arsepull" if you will. Do not take it as authorative, please.)
That's fair. Question, what's it called when a song diverges a bit? Not like a short hook or anything, but a sudden change in pace, changes into a part of the song that will only be used the one time.
Is that the bridge?
yeah, that's a 'bridge'
here's an example of [Bridge]
https://app.suno.ai/song/04bd49c8-45e7-42ad-8e00-4d76d029e444
Sparse strut, sinister electric-blues, note-bending guitar, slow male vocal song. Listen and make your own with Suno.
Does the new clip only reference the last clip, or all the previous clips?
mostly last 20s of previous clip
Is there a way to change the vocal in the continuation clip?
All before (usually unfortunately). That’s copy of copy of copy of copy. Each next part it goes own way.
Yes, sometimes works female/make name.
[Verse]
[Anna] lalalala
[Alexander] lalalalla
[Anna and Alexander] lalalala
But you need to be lucky to catch that.
If a clip ends in the middle of a word, is there any best practice on how to continue with the next clip lyric-wise? If I start with the last word of the previous clip the results are weird 🙂
This is the routine I follow:
Try depending on where you are in your lyrics (if you have a lot more to go) just continuing and running the same lyrics you have. This can sometimes make it pick up where it left off. A large percentage of the time it will start over in the middle of the song. This is just the easiest thing to try to see if it works.
Next try removing everything already sung but the last good line, then try continuing. This sometimes works. It's how it always seemed to work in v1 Chirp for me.
After this try creating a line break (empty space) and continuing from the line that was interrupted. This has a good chance of working.
Then starting with the line that was broken. It sometimes will "remember" where it was and pick up from there.
Finally when all else fails, batten down the hatches, remove everything up to the word that was interrupted and hope for the best and throw the dice. You will likely encounter some weirdness and you will need to keep trying. But if everything else has failed it's this or...
...or you go back up a chain in the song and try again giving up on the good part. It hurts, I know, I know. But sometimes it's the only way forward.
...As you can probably tell this is a bit of a pain point for me and I've done these often...
Thanks a lot for the info 🙂
Good luck!
Something kinda odd that I've noticed when I generate music sometimes, there's a little hallucination where we hear "thanks for watching!". Anyone experienced that ?
thank god I'm not alone! 😂
lol.. I have it with a Dutch sentence.. with the words "van de video"... and it then continues that sentence the whole song 😂
Let me follow your guidance. thanks
That’s what you call dirty data, it’s clear they taught Suno with videos that included that line
I've had a few where it ended in applause, like it was a live recording.
Hello friends. I studied the correspondence regarding the issue of the open API and realized that it is not yet publicly available, only for some individual cases. I want to connect a telegram bot that will create compositions. In what ways can I implement my task and who can help me with this?
Since the Chirp Model V2 is already in the web app will the V1 model release like bark to Github to run localy o: ?
Hi there. Where is the most up to date list of functioning metatags?
nvm, I just found the wiki
which genres and prompts do i use to get a similar vibe to music by dayglow, i tried dream pop, dream-like, vaporwave, but didnt work
also what is a consistent way to get an [Intro]? Will something like this work? [Intro] (). (). (). ().
also to get vocals with specific things going on
like doubling or harmony on a line
can somebody from SUNO start a room for FULL finished pieces of music please? where you are only allowed to post full finshed songs?
also, if there are any entrepreneurs out there, why not code a spotify for ai generated music?
How can I reuse a script on you website?
I have a song I like but it is a bit inperfect between 2 lines and I want to change that
how can I do it
Does someone know for certain how the lyrics feeding into the model works for continuations? Does it concatenate it with previous lyrics?
start the next lyrics where the old left off. Don't repeat the lyrics (or metatag).
if the end of a word is cut off, Chirp will continue singing it, as it hears the end of the last clip and Continues.
Sometimes this can be a problem, especially if there's a lot of that word left to sing... (multiple syllables) it may hallucinate the end of the word or phrase, and then keep hallucinating....
Sometimes an ending just can't be continued cleanly. Until there is a away to continue from a different part of the song (before the cut) the only option is to go back and regenerate the clip being continued from, and hope for a better timed ending...
Is there a way to ensure the second chorus of a song is the same as the first? I haven’t had this issue in the past but recently I’m trying to continue a song and it’s reimagining the chorus everytime.
You can bypass iot if you really want for example with Selenium and write your own semi-api. Or use RPA automation. BUt I have no idea why would you need it for telegram channel xD
Today Suno webapp is all about no song created and "Credit refunded" 😑 Is anybody experiencing the same trouble? I used to think it's random but at this point I start wondering if it's linked to the lyrics I wrote in the box 🤔
no... it is working fully on my end.. I suggest you send a mail to their support@suno.ai
thank you!
Anyone have a trick to make child voices? If possible Girl and boy on same time (duo)
how do I add a product name into the lyrics? using regular prompt (not custom mode)
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
is the model open source? like could someone retrain the bark or chripmodel on other things?
is it possible to download the audio without the voices, or the audio without the lead vocal and keep chorus voices? i want to sing over the music or over the chorus voices
No, sorry. Everything you can do is to:
1 - Make instrumental and create vocal on your own
2- Use another AI to split to steam packs or remove vocals (AI vocal remover - it’s free), but quality will not be 100%
When daily credits reset? In 50 minutes?
FL studio has a stem remover you can use (not free)
bro this has so much potential but the damn static
https://app.suno.ai/song/851797fb-22e7-486c-997f-dc85bde1f025
I would love to know how to do this too please ❤️
be very careful about anything that has to do with kids lol
I want to be able to do duets between adults.
Any tips on what to do if I keep getting instrumental only? I have lyrics in the box and reference vocals in the beginning of my prompt, but keep getting these long instrumentals and if I try to continue from there hoping the vocals will finally show up, they never do :/
I've had that too. I rebooted and started again.
I did clear my cache and that seemed to help. Thanks
Once again the web app started telling me that it can't generate my song with original lyrics because "Suno is for generating original music", etc. Already tried changing style and name but with zero success. What's the point of paying if this is gonna keep happening?
it's a generic error message. ChatGPT doesn't tell Suno why the prompt didn't go through, Suno gives a generic message that covers the most likely reasons
Other reasons include:
- lyrics about torture, sex assault, or self-harm
- lyrics implying something inappropriate with children
- Un-intentional copyright lyrics or phrases
- names of bands or musicians ("style of...")
- a web glitch
anyone knows how to reliably force Suno/Chirp to hallucinate vocals? experienced vocal hallucinations for an instrumental track and thought the effect (almost ethereal) might suit an other instrumental idea I am going to experiment with.
some genres want to sing, so they are easy to get hallucinations just by not adding lyrics... 'Gospel' will (always?) hallucinate.
On other genres you might have luck adding a voice description to the style or lyrics prompt. Encourage it to sing, but don't give it any lyrics.
Does ever Suno staff answer e-mails? I sent them three messages (on three different addresses) a few days ago but no answer 😟
Same
lol, i have 2 kids and would like to "implement" the kid sound
Hello, for a few days my songs have been full of 'clicks' and distorted noise disturbances, especially in parts with fewer instruments. Can you help me? Thank you. I'm using Sumo from the website, not Discord
could you post one or 2 of the bad clips to https://discord.com/channels/1069381916492562582/1082775596767117312 ?
ok yes
ok yes
When I "Continue from this clip" to do Verse 2, it will not use the same (or similar) melody. It's as though it is just making up a new melody for each verse. Help?
Why this error occurs for some prompts ?
It usually means they think it violates their content policy.
And they in this context might be ChatGPT and not Suno
Youve put in some copyrighted lyrics or you've used a famous singer or band in your prompt.
or used some profanity or sexual term they deem bad. which is annoying 😄
Assuming none of the previous guesses are accurate, it could also be a glitch.
You could try placing some lines of spaces and maybe a random instrument at the end to change the feel of the lyrics and this could enable the lyrics to go through.
"Gospel will always hallucinate." There's a joke in there somewhere.
Is there documentation somewhere that lists the types of things we can put between [ and ]?
Or is it more like... "just try it and maybe it will work"?
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
Guide the lyrics and music breaks with [Metatags] to help Chirp transition from section to section. A small number of metatags were reenforced when training the AI model, but the rest have been discovered! The choice of genre and your lyrics will also influence how Chirp creates the song.
Is it known that Chirp sometimes just straight up gives you an existing song?
My prompt was "an upbeat pop song that has a twist ending" and somehow I got portions of Bad and Boujee as the song which is surprising on multiple levels:
(Explicit Lyric warning)
https://app.suno.ai/song/bf89197f-792f-42c9-9b45-64ab8bcaf923
I can't break my song out of an instrumental interlude. I've tried going straight to [Verse], adding a [Break], adding another instrumental interlude, build up... nothing seems to work. Ideas?
Also—I'm assuming as I've found nothing in my research to indicate this—if you like or dislike a continuation and proceed to remix, does the AI create based off of your likes and removes elements of your disliked continuations?
Hey, is there any known way to influence ai mid promt (suno.ai web version), like to say wich part should be singed via male and wich part should be singed by male vocals?
[4:06 AM]
I got some results, but it appears to be by pure luck
I looked at the notion, metadatas and voice metadatas dont work at all, like [Mary] [Female narrator], or [John]
Have they said anything about the fact that songs usually end mid-sentence?
That is just you i think
You are probablly giving it too much text
If that is not the case, add [Break] to the end
or edit your last song section, example [Chorus] ==> [Outro]
Are updates being deployed on a regular basis, even though they aren't being announced?
{Fade Out] rarely works. I usually do a double [Song Ends] with a blank line between to get an actual gradual fade out to the end of a song.
Okay, so you are saying that metatags are actually mini promts?
How do they get processed?
Like is it before generation or can i say for example [change of flow] and it will change it
because i already tried to do a lot of shit with male and female vocals, and I cant do anything to influence it
I think its completely random in that case, i had duets when having no tags for male or female with no explanation
and when i try to use it it doesnt work, really weird
Wish I could get some infro from actual mods and devs
Nah, they're still metatags because if you do a sentence or whatver in brackets it'll just be ignored in most cases. My biggest pet pieve is not being able to create instrumental intros to songs consistently, such as an instrumental lead to start off the song, followed by vocals. Either I get a completely instrumental clip, intro metatag is ignored and vocals start immediately, or I get a good intro where a part of the vocal isn't synced correctly or something.
Duets are hard. No real way to do them consistently.
Thats probablly because of intro meta tag, for me it works consistently
They said not to use [Intro] tag
You can use [Hook]
or [Refrain]
At the beginning of the prompt? To get an intro? So I could do:
[Hook}
[Verse}
?
I'll have to experiment with that. Refreign will just repeat a Chorus right? So I doubt it could make an intro.
Same thing with chorus, if i want to use chorus twice, i can change them and it wont double up
But its probablly better to use [Hook]
In my country you usually call the intro to the song a refrain
Yay, hook actualy worked for this. Glaring mistake with the pre-chorus in there twice and it messed things up, but I'll fix that with a new version some other time. Turned out well and [Finish] tag worked to end it. https://app.suno.ai/song/ba352166-87a8-44cf-af33-dc582e4557e2
piano female clear close perfect song. Listen and make your own with Suno.
@light ether Any info on longer duration coming soon?
Cool, glad I somewhat helped
Aiming to be a part of next major update
Is there a way to ensure that generations meant to be purely instrumental don't have random lyrics added to them? I'm speaking about the Web App. I'll often describe an obviously instrumental song and it will still add lyrics. Adding the word "instrumental" results in a warning about not using artist names for some reason. Any help would be appreciated.
Adding the word "instrumental" results in a warning about not using artist names for some reason.
That sounds like a bug. Can you submit that prompt to this form?
https://docs.google.com/forms/u/3/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/edit
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
I went ahead and submitted the form. In the meantime, are there any other ways to increase the likelihood of the generation being an instrumental?
Any descriptive words that are generally used to describe non vocal music more than vocal music will trend it that way. Also repetition, "folk, instrumental, instrumental, instrumental". Are you in Simple or Custom mode?
I'm in simple mode.
Switch to Custom Mode where we can control the lyrics separately
https://www.notion.so/How-do-I-make-an-Instrumental-352abcb0bf6a4f77b8fcf76a768f18af
Oh great, this is exactly what I was looking for. Thanks!
hey @zinc crow , thanks for submitting the form! I couldn't reproduce your error, though-- could you post a minimum working example of a lyrics prompt that doesn't work? (feel free to DM if you'd rather not post publicly)
I think it was Simple mode
(same deal tho)
What about a "please just give me like 6 more seconds of this clip pretty please it was so perfect and it just cut off the last two words of the chorus and why does it have to be like this" button?
I have burned literally thousands of credits to this problem
Yeah, working on "continue from" so you'll be able to regen any endings of a clip
do you uh, need a beta tester for that feature? 🙃
I've finally subscribed the Pro Plan [applause] 😅 I'm sorry to bother you with this stupid question but I can't understand what "10 running jobs at once" means, as I get 2 versions each time 🤔 Thanx.
Is the generator down right now? Just bought credits and I have a bunch of prompts stuck generating for the past 10 minutes
Same here
I too am having things hang on me.
May be some site issue, yeah. This is the first time I've chirped in like a week in and site goes down 30 seconds later... just in case that isn't a crazy coincidence, I was continuing somebody else's dead-end-er clip, but from a second dead-ender segment, then continuing from that one which has a non silent bit of noise. (A last ditch way to rescue a song that won't continue, works occasionally with luck.)
So what you're saying is...this is your fault? 😉
lulz
(Added the z. The z is important.)
I'm curious. Since a lot of people say that the regular create mode is better than the custom mode, how do you create full songs with that mode? You can continue clips, but your lyrics might be out of wack...on a side note my stuff is stuck at loading.
Suno is Still hanging

has anyone said this 
Seems it's working again for new prompts
Did today's update change the algorithm? At least in Finnish the system is flipping out now in strange ways when continuin clips? After two minutes it starts to hallucinate weird stuff, and this happens now with almost all of my continues.
Nope no change in the continue models
I keep getting copyright warnings from gibberish I write for my friends. Stuff I know for a fact does not exist in any song.
Can you add those to the list here, whenever they come up. Sounds like a useful bug report! https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
I removed every sentence from a long song until there were almost none left, I kept getting copyright warnings.... Then when there were like 4 sentences left and I arrived at a sentence that was "kominn til að fagna með þér." which translated to "Arrived to celebrate with you" and once I removed that particular sentence, it started generating.
So basically, it's stopping the generation of a song because of a single sentence... I've reported this using the form. But Suno definitely needs to de-escalate its copyright filters, this makes no sense whatsoever. And you can't even find this sentence on Google. So it's not in any song.
Furthermore, I was able to pin it down to a single word. "Fagna"... which means "Celebrate".
Interesting, thanks for narrowing that down.
I've pinned this here and in #💬┃general-chat so it can be found easily and recommended to others so we can hopefully capture enough data on this to get it fixed.
Oh, I got the reverse. Singing like a chiptune!
https://app.suno.ai/song/b4aaaa7b-aa37-47c5-9236-2ecca2bca4c5/
Melodic Chiptune, female vocals, vocals, sultry vocals, chiptune indie folk song. Listen and make your own with Suno.
Ok, then it must have been something specific with those clips that set it on a bad path.
Is there any way that I can ensure that the last word in a generation doesn't get cutoff? I just want to generate a "song" with a Chorus followed by a Verse and perhaps a very short musical ending. I am using the web version.
which part of the second do you think sounds like Bad and Boujeee?
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
Is it possible to make a country song with a rap in it? I've been trying various things, but I can't do it.
hi guys, where was the info on using more than just [verse] and [chorus], what other terms can i use to include variety and texture to songs, or make it purely instrumental?
Guide the lyrics and music breaks with [Metatags] to help Chirp transition from section to section. A small number of metatags were reenforced when training the AI model, but the rest have been discovered! The choice of genre and your lyrics will also influence how Chirp creates the song.
No way to determine what get's cut off, unless you're very strict using a four line approach to each generation. I use eight but a word or part of a word might get cut. The best solution to that is to simply continue the clip and delete up to that word, leave the rest, and then add the next section.
Yeah, I had an argument with somebody about it.
Generated a nice house-like beat. https://app.suno.ai/song/8772ec54-5e35-45f0-8769-92f39f8b08f9
clear close perfect female song. Listen and make your own with Suno.
Also, there goes the aI with Neon again.
Nice, I got a Dropkick Murphies sound with this one. https://app.suno.ai/song/a995b1d7-b3c8-45da-9591-4e0c54af7180
Lots of meaningless words in my lyrics, today 😡
Even if I write my own lyrics, the voice keeps on singing: “Sottotitoli e revisione a cura di QTSS”, an Italian sentence meaning “Subtitles and revision by QTSS”
I googled it and I found this:
https://community.openai.com/t/subtitles-created-by-amara-org-qtss-etc/462561
but no one answered the one who asked about this strange bug 🤔
Very frustrating tho.
Many times into SRT appears strange lines: “Sottotitoli e revisione a cura di QTSS” “Subs by www.zeoranger.co.uk Diolch yn fawr iawn am wylio’r fideo.” “Sottotitoli creati dalla comunità Amara.org” “Sottotitoli a cura di QTSS” “Sous-titres réalisés par la communauté d’Amara.org” “info un libro pubblico su www.mesmerism.info” etc. 510 00:...
It’s because of dirty training data, the only way to fix it is to clean up the data which contains those sounds and train again.
thank you, that's interesting! I'll try to understand how can I do that in my case, because the in 1st part of the song the singer follows my lyrics until the end...when I click on Continue from this clip, the next part of the song keeps the same music but those words always show up!
Unfortunately there’s nothing you can do about it on your end except change the prompt and lyrics to something that doesn’t provoke it
Maybe you're right: I've been using other ways like different metatags ("End of song" or "Outro") or prompts like "Start a new song" but to no avail. Thank you!
❌
In the web app a lot of my generations are coming out sounding a way that reminds me of (for lack of a better way of describing it) a computer grinding to a halt. Like a digital crunching sound of very low quality. At times it makes me wonder if the server is overloaded, but I don’t notice any increase in queue times. I’d definitely rather wait longer to avoid it… any ideas?
Like is there a way to request that my account is put into a deprioritized queue that takes longer, but experiences fewer quality issues?
Why does chirp always add “THANKS FOR WATCHING “ in the song when you are extending a Clip , developer’s please fix
Funky
I'm sure this has been answered but I can't seem to search in this channel. Any tips how to get an instrumental interlude in a song after a chorus? I've tried [Instrumental] with "Instrumental" in the style... no dice.
I only have 10 creds left, so I gotta use it wisely. I wanna continue a song that sounds really good so far, but I'm uncreative and don't want to write the next set of lyrics myself. What's the best approach?
Less importantly, if possible I would like to insert the same chorus though because it goes hard, but I would want the rest to be newly generated lyrics
Actually, I guess I can have ChatGPT to make the continued lyrics
big brain moment
frick, there's a chatgpt partial outage
Neon, Echoes, Shadows, Journeys...over and over again. Echoes in particular drives me nuts. I tell ChatGPT not to use the word "echoes" and it uses it the very next song.
prompt engineering basics. You can only tell an AI what to do, not what it cant
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
I have a bug to report: whenever I input "twelve bar blues" as the requested style, it always ignores the lyrics and plays a pentatonic guitar "solo" instead
thanks ya on some genres somehow it decides to do instrumental 😂 working on more control there...
It was not meant to be a serious bug report, more like a joke on blues guitarists 😄 Good to hear, though.
How do I delete my Suno account?
Please contact billing@suno.ai
ok. thanks
I made a mistake in the text but I like the music generated. Is it possible to change the text to the same music?
Unfortunately not at this time.
Save the clip though, they are working on a feature to allow a manually set resume point, which means you might be able to resume before the vocals start and try again once this is available.
ok thank you 👍 it's a shame I liked the chorus but I can't reproduce it
I will wait
Is Suno planing to add other payment methods? Im from EU where debit cards or paypal is usually preferred over credit cards.
Please make a post about this in #💡┃suggestions-web so that team members see it. It may not seem like it at times (backlog is a thing) but the team members do read suggestions in there and it does inform their priorities a bit. Just...delayed. As I said backlog is a thing and we have credible rumors that work on v3 has begun.
\chirp
Is it possible to release a full sound only on part 2 ?
Not sure I understand the question. Could you rephrase?
Are you talking volume? Instrumentals? What do you mean by "full sound" in this context?
no singing? no words?
when I do "continue From this Clip" I go to part 2 but when I want the "get whole song" it gives me both parts but I'd like to be able to generate "get whole song" only on part 2 (sorry for my English I'm with the translator)
Still no word about the out of tune issue? 2 weeks have passed, spent about 7000 credits on out of tune results.
if you press generate all song it gives you all parts together. if you want to get only part 2 then just download part 2 not whole song
It was a cool surprise the first time, but its getting old lmao
I can't seem to create in the custom mode on the website. The simple mode is fine but the create button is grayed out and not able to be clicked in the custom create mode for several days. No matter what account I use. I emailed a support ticket and no response about this for a few days now other than automate reply.
The create button will be greyed out until you type something in the fields.
Is it not changing when you type in lyrics?
(or style)
I swear I haven't been trying to create from empty fields, but I'll be damned if it didn't work just now
I must be losing my mind thanks so much
We've all been there.
If im trying to generate (continue) it keep saying on that specific Lyrics "Credits refunded" and just 3 second clips with random sounds gets created, why is that and how do i fix that ?
the problem is the clip you're continuing from... it has probably ended the song within the clip, leaving a tiny amount of dead air at the end.
When you continue, it continues dead air, then hears nothing is playing and ends the clip (it's a know issue, that why we get the refunds)
a feature that will allow us to continue a clip from a specific time (before it ended) is coming according to Devs, but no eta...
uhm so cant i continue from that clip at all ?
I’m having troubles with the voice of a continuation.
I used “Keep the same voice” and “Keep the same vocals” (along with “Continue this song” or “Continue the below song”) but the result is always a different voice (and it’s quite frustrating, because the music is good).
Any idea of other prompts I could use?
Thanx and have a nice day.
P.s.: do you think the lenght of lyrics could have any influence on the voice? (yes, I'm desperate 😅 )
I'm trying to continue a song but swith from a guy voice to a woman voice, like a duet, simple comadns like "woman voice" didn't work, any idea?
The style field text isn't that sophisticated. Just music genres or terms, general descriptive terms, instrument names, moods, that kind of thing.
However, the voice of the singer should generally stay the same in any continuation, no matter what you put in the style field.
The length of lyrics could have a small effect (especially in the first segment, it probably makes rap more likely) but not a large one, and very small in a continuation.
When you continue a song, are using the menu option in the "..." menu to do it?
Just to double check we're on the same page.
In general, if you repeatedly get the same bad result (a changed voice, for example) in a continue that is a continue from the exact same clip, the problem is in the previous clip, and may not be related to what you type in any field in the continue itself.
Thank you...Yes, I use the 3 dots and then "Continue from this Clip"; sadly, this is a Part 6 and I guess Suno has lost somewhere any actual hint of the voice (even if he sang a few lines in Part 5 ). I tried many many times and I got the craziest results: foreign voices, disturbing noises and LOTS of awkward off-key "singers" 😭
The voice wasn't actually that bad in the previous part
Yeah, you may be out of luck... for now. In the future there may be some better tools to recover the voice. But at the moment you may have to back up to an earlier clip and go forward again. For very long songs you have to try and keep things you want to stay unchanged not too far in the past, reprising voices and melodies, so the model doesn't forget them. But sometimes you just get an unlucky clip that seems to have an invisible "new singer" command in it.
do you have a release date for V3 ?
Not at this time. All we can confirm is that there will be one and that it is being worked on now.
ok thanks for the feedback.
I'm looking forward to the functionality allowing a manually set resume point 🙂
You and me both!
I keep getting "An error occured generating lyrics" on the site.
This message: https://suno-ai-wiki.notion.site/Error-Couldn-t-generate-that-or-Try-another-prompt-d6e88752dd9144d6b2c14503ceca43aa
Or a different one?
OK, are any of the things on that list applicable to your lyrics?
I have had this too
I've only seen that one when my lyrics are too long...
(like I pasted in a full song, or tried to [Remix] (copy full prompt) from a full length song....)
Yeah, try shortening the prompt or switching some words. It worked for me
hmm, never had that issue, usually starts a new beat for me
i dont know what your promt is like, but usually when it stops for me its because of this issues
use of [Intro] tag, multiple chorus and verse tags or uneven flow that makes no sense
Continuing from here: #🐣┃suno-showcase message
https://app.suno.ai/song/7521ed97-13db-4980-acf7-43cd77b75986
https://app.suno.ai/song/9055c88b-5256-4acb-af4a-a9a179a52ef0
Still no vocals...
Tropical House, Reggaeton, Calypso, Catchy Hook, sweet velvet male dance vocals song. Listen and make your own with Suno.
Tropical House, Reggaeton, Calypso, Catchy Hook, sweet velvet male dance vocals song. Listen and make your own with Suno.
Might just be bad luck.
It does happen ¯_(ツ)_/¯
Tropical House, Reggaeton, Calypso, Catchy Hook, sweet velvet male dance vocals song. Listen and make your own with Suno.
https://app.suno.ai/song/9bc8cebd-a30d-4e3a-8483-cf1ec71d5766
I've instructed it to take the first two lines first verse and prepend [Verse]... Sometimes it does this, sometimes it doesn't... The more I badger it in the training dialog, it seems to improve - but if all else fails, bringing up the lyrics as the opening line seems to help retain the lyrics.
But I copy and pasted those lyrics, and asked for them in the style of the Poolside band
Poolside Band Style, Dreamy Synth-Pop, Laid-Back Groove song. Listen and make your own with Suno.
Oh shit. If you don't spend all your credits, they disappear at the end of the month.
Well, there goes almost 5k credits. Here I was thinking I could just have a bunch of credits, and spend them at my leisure, because I paid for it.
That really burns
Yeah it's a use it or lose it situation - i just upgraded from Standard to Pro and im going to make sure i go ham towards the end of the month if im not already out if credits.
how do i prevent songs ending so abruptly at the last word? is there a list of commands or terms that affect the music?
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
Hi!!
I need to know about the datasets on which suno/bark has been trained. Can i get some info related that...
Love how the text on screen is almost part of the content. Self documenting music.
I wonder if complete sentences like that (that don't look or sound like song lyrics) are almost seen as if they had [] around them.
It's less relevant now, but the overall shape and form of the lyrics can have an effect. For example it would probably be more likely to sing formatted like this:
[chorus]
The track starts with a funky synthwave groove,
Capturing the unique atmosphere of 'Viewpoint'.
A combination of electronic beats and rhythmic bass lines
Sets a lively tone, while synthesizers mimic
[bridge]
The bright, 3D polygons of the game’s visuals
The vibe is upbeat and engaging,
with a hint of digital innovation
[Verse]
Transitioning into a club electronica style,
The verse features soft, jazzy organ tones
Layered over a deep, thumping bass
reminiscent of a Friday night at a local nightclub
Removed punctuation, lines roughly look like normal lyrics, capitalized first letter on each line, etc
I've also tried style prompts like [verse], [chorus], melodic vocals, female vocals, sultry vocals, verse, chorus jamming in vocals and even chorus because it's associated with sung lyrics
You may be able to avoid that by paring it down to just the musical keywords, something like:
[Opening - Funky Synthwave]
[funky synthwave groove] [electronic beats] [rhythmic bass]
[bright synthesizer instrumentals] [upbeat]
or just spitballing, variants like:
[Verse - Funky Synthwave]
[funky synthwave instrumental :: electronic beats :: rhythmic bass]
[bright synthesizer solo :: upbeat]
(untested)
Though having a lot of words, just in general, is probably good for musical diversity
why it shows me message on subscription account "manage your account"
I haven't seen a warning or popup for that before...
Do you see anything suspicious when you look?
multiple 'active devices' maybe...? A request to connect your acct to Discord ID...?
I don't see the point of the explore page. I don't believe mistletoe is still number one on the trending list. Until we have an actual way to find new songs by people, why not just remove the trending page?
And great... Now I can't select custom mode... Does this mean a V3 or what?
what Remix means after i generated a song ?
It's very misleading. You can't really remix a song, it's more like recreate the whole song, different. You can't remix full songs because there's the character limit which is annoying.
Longer it is, higher probability is to be singed.
[Verse - Rhythmic, Drums, Upbeat] is usually ok.
But:
[Verse - Soaring vocals, happy, enthusiastic, building (…)] is too long
Custom mode broken for anyone else?
What actually happens when you report a bug? Thanx!
You feel better and get +1 hope
🤣 ...because I've reported A LOT of them so far (empty file, drunk singer, random notes and so on)
any previews or hints to what v3 will sound like?
Is there anyway to repeat the verse and chorus melody from the previous clip. The melody always changes.
if you want an exact duplicate, can always just download the clip and dupe it in audacity (or any other daw) where the beat aligns
also saves on credits tryinna do it
Thanks. I'm looking for the same melody and music, but the singer to sing different lyrics.
oh, then that's luck. if you used metatags to specify a verse and chorus in the first clips, use the same tags in the second clip and it should help out
Thanks!
We're all chomping at the bit for that one @hollow pollen 😩
Yep! It will be a game changer!
It's the one massive weakness in how clips are generated... if you're working with a continuation clip and you don't get what you want you just rewind one clip and come at it again.. I've lost track of how many times the first clip was perfect in every respect except one line of my lyrics was a syllable off, or a word was misprounced, and there's no way to fix that with the current system.
Right!? I have so many good first clips, but hardly any good second clips. It's a shame.
I blew something like 5000 credits last month trying to randomly get the same style on the first clip that I got previously to try and fix a rhythm problem with my lyrics because I really, really, really liked the vocal performance and the clarity of the backing instruments, which isn't always easy to get. Eventually I had to settle for something I liked less but there was no way around it
Any idea on when we'll get this?
I think the last I heard was it was a "future feature" that would probably be coming at some point... but I never heard any kind of timeline
This would be the one thing that would make me quit until they get it.
I do my best to compensate by trying to carefully count my syllables in my lyric lines to make sure there's nothing obviously broken that's going to fatally screw the rhythm... is the more frustrating stuff like common words that get mispronounced. I had "chicadee" (i.e. the bird) in a lyric ... the correct pronounciation is chik-ah-dee and I got ch'eye-kay-dee ... drove me nuts
we're gonna have cropping continue very soon, would that fix things? like take seconds 4-9 and continue that
that would sure help with this glitch I'm fighting against right now where a word changed between the clip and the "full song" right at the continuation split.. if I could have cropped backwards and repeated the entire line it might have fixed that... I had to give up on the song entirely and go with a completely different rhythm and vocalist since there was no way to fix the glitch, even with painstaking audio editing to try and overwrite the glitched word
Ugh sorry yeah that should fix that. We already have that feature in beta so should come very very soon and should work for old clips too
👍 🥰
I have so many clips to do this with...
that would be a bit of a game changer, especially with being able to crop out clips that think they're the end of the song and lead to silence. (assuming it would otherwise continue as normal just with bits cropped out)
probably made 150-200 clips yesterday tryinna get around that lol
@untold briar @graceful condor I'm kind of stuck here with the same on instrumentals. My questions are: 1. Do I put the description of the entire track into the custom mode? With lyrics it's OK to find out where it stopped and then continue from that, with instrumental much harder. 2. To control the length of a section it seems better to put those into separate continuation sections. However that does make it harder to progress out of that section. Like if I start on a [Intro : Solo sting quartet] and in the next continuation change it to [grand orchestral theme] most of the time it will just continue with the first section and ignore the custom mode. 3. Or should I use the style of music box to change the style for the next cotinuation? Thanks.
Kinda off topic 😅
Suno generated a song that starts with a few words by a voice-over that are spoken in a foreign language I can’t recognize!
I could keep that sentence because it’s actually mixed with the music – which is actually the intro – and it fits nicely into the rhythm but I need to be sure the words don’t have an offensive meaning 😅
I tried to find a website where I can get some hints but they all ask me what language I want to translate…but I don’t have any clue about that!
Thanks!
It's most likely not using real words (95%). In the cases where it isn't, it's most likely a super common phrase used in most songs. Worst case it's random sounds that happen to form some offensive word, but that's pretty rare.
Thank you, I'm gonna keep it!
Usually if it hallucinates real words, the words are in the same langauge the rest of the song is, so it's probably 99% not real words even.
Perfect!
Pro tip: Suno may create low-quality or distorted (can be in drums) songs, like this one. But it is still an amazing generation though.
Hi guys, have you ever successfully created a song with multiple voices (male, female, choir)? And if so, do you have any tips?
If you label your lyric sections with [name], chirp will usually switch to a new voice, but that voice is as random as always and may not be the right gender....
`
[Linda]
I am singing...
[Tony]
now it's my turn...
[Linda and Tony]
There's a chance we are singing together now...`
It's fun when it works, but being so random it will cost a lot of tokens to re-roll until you are happy with results...
Hmm, interesting, I hadn't thought about using names. I had tried with [man voice] [girl voice], but it wasn't respected. Thanks, i will try !
I tried to put the different song sections on continuations. But having only one section in the lyrics part seems to always make it think it's the end of the song. And then it stops the continuation with a refund of credit in dozens of times I have tried it
I'm trying an Orchestral track, a movie theme song but I'm out of ideas on how to control the sections and breaks. Some things come out really well but no way to control it. Maybe @untold briar has some ideas
Philadelphia Sound is not as easy as it should be... 😓
https://app.suno.ai/song/fd1c3094-3e01-4895-935a-a2b452fd8bfc
summer disco hit as Philadelphia Sound song. Listen and make your own with Suno.
breezy summer hit as 1976 Philadelphia Sound, groovy, jazz horns, whistling, fun song. Listen and make your own with Suno.
(still not Philly sound)
hey all , so i went ahead and signed up for pro suno plan... and credits didn't transfer to chirp? telling me no credits? are the two different?
Hey guys. Any idea when Suno will be releasing an API for Chirp?
email
billing@suno.ai
with your discord username and (google?) login email. they will get it sorted.
(they just need to connect them manually if it didn't happen automatically).
Have not heard anything about API 😕
so?
honestly, all I've heard is it's being worked on. 🤷♀️
no specifics.
ah, ok. thanks for the info tho. i would love to hear what it sounds like when its out
I would like to know the models names please 🙏🏾
that is just making up names... but re-using them consistently...
it could be
[dog]
[cat]
they will not be animal sounds..., but like a script with actors taking the different voices.
(those actors being selected at random)
I haven't heard much from the Devs this week... We haven't been told any updates, or why Bark stopped working, or plans for near future...
I would love to betatest it. I would focus on Finnish primarily, not English. I know Finnish is difficult for AI. I would send feedback on how to improve it, but if that’s not possible i’ll do it when it’s out in 2 or 3 months. Keep it up!
V3 I mean
yes, i like beta testing, i do that with a bunch of ai's
i do beta testing every once in a while
What are the commercial terms on the pro plan?
with paid subscriptions you get a commercial license and all rights to the song...
there's a FAQ here:
https://suno-ai.notion.site/FAQs-b72601b96de44e5cacd2cd6baa985448#f0f1441180484d6094206e84e334ba36
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
For some reason suno decides hey were gonna force end this song sense your on part 9 now :/
Songs only 6 Minutes and 23 seconds long
As much as I dont want to leak it may have some insight ||https://app.suno.ai/song/f46e2b0e-b2d2-4733-9d26-cd77d5536179||
Where did you get this
Was that with V2? Metatags were so long and complex
Yea V2 also tends to be me Adding multiple into one tag
Hey, how do you make your song trending and make your song(s) show on the Trending tab?
I made this song Public.
There is a planned Explore page that lists creations that are public others have made
how do you stop songs ending abruptly after the lyrics
[Instrumental Interlude] Or smthing similar typhically
Yeah but how do you make your song show up in that tab?
It doesnt exist yet, and by getting the song popular enough
Similar to YT trending for example
thx. how about a list of terms we can use
Ok iguess
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
Just use logic
i am not up on music lingo
Oh wait not tos 🤦♂️ one sec
Guide the lyrics and music breaks with [Metatags] to help Chirp transition from section to section. A small number of metatags were reenforced when training the AI model, but the rest have been discovered! The choice of genre and your lyrics will also influence how Chirp creates the song.
There is a ton of models like that out there. Just focus on improving chrip and fixing issues and don't waste your time on Bark... Just my 2 cents.
Cool wiki, Thin Creator.
I meant to reply to the website
[Instrumental Outro, Fade Out] usually would work
[fade out] sometimes works as just ending the song instead of actually fading out
I mean like [fade out] sometimes just suddenly ends the song instead of the fade out you expected.
Hense why its a combined tag
More likely to fade that way
Maybe it's just in me tho
Do you find using time period works for instance 2000s hip hop, 198ps hip hop would produce different styles
Absolutely
Wrote a verse that suno refuses to turn into music. It did it once and after that it gave me the denial message for unoriginality as if I chose a popular song lyrics or artist
Hmn, maybe add it here if you rememeber the exact verse, could be a bug https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/edit
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
Thank you for this
If found the content filter to be very very hit and miss. Several of my works (my own lyrics and designs) are flagged repeatedly but if I keep trying to submit it it will eventually just generate it. There's definitely something wrong with the moderation filter system.
I'm off on holiday ATM but when I'm back I'll run more indepth tests and post results for the team.
i found a v3 song.
georg sent it
I was wondering about those lyrics, but now I understand....
not fully done yet.
I think it’s V 2.
georg is a dev and said it was v3
It doesn't sound different (not like the difference between v1/v2)
But v3 might involve...:
more training so the output is somehow different
it might have a larger language model (making the prompting better)
It might be able to exploit the context memory in some way v2 can't
It might solve some issue we haven't been told about
It's also not clear how much of v2 was based on actual v1, or if it was started over from scratch (same with v3)...
Has anyone noticed that the first continuation clip is VERY much louder than the initial clip?
Would be nice to try V3
V(-1) (Nightmare edition) is my favorite.
NEGATIVE ONE
the 'Dire Chirp' from before time....
🦕 🦖
Humm... has the process for contining from an interrupted clip changed? I just had a chorus hit the 1min time limit on the last 4 words of the chorus. Usually I just repeat the last chopped off word and then continue with what was missed, but when I tried that it seems like the point it's trying to continue with is 2 full lines back in the chorus now.
I do a lot of songs that are slow ballads, and the 1m limit on continuation clips tends to bite me all the time...
This is more like some Beatles Revolution 9 madness than a nightmare... hmn. incoming
Mistletoe… trending?
OK... so it's not a continuation change... it's that glitch again where the word gets swapped out... the word it should have sung at the end of the phrase was "born" and instead it used the last word from two lines back in the chorus "moon"
tried to get Bark (Chirp V-1) to hallucinate more verses of the v3 clip but as usual bark is a free spirit. All clips are pure bark, the only prompts were original lyrics and [music] tags.
wait. i have one for you to try.
It may have to be later, but sure send me something and I'll try. (Already spent most of my morning on that one)
oop. the ending
let me fix
there
Accepted terms of service.
I'll post a few in bark-techincal channel, mostly they go off the rails immediately, sound like tv clips
xdd
does Suno have an API for Chirp?
Not at this present time.
v2 song
VERY annoying the restrictions on lyrics, especially with lyrics generated by the AI itself, it won't let me extend the song!
i can show you how to extend the song.
are you using site or discord?
I know how to extend, but when i click "continue" it gives me error "Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names."
yeah, you can't copy famous artests, bc suno would get sued
you can't put their names in
chirp
i'm not.... i'm copying and pasting the exact lyrics generated by the AI itself lol
i mean, you prob asked like "in the style of (famous singer)" and the bot was like "nope, do your own style not other peoples"
no it is a children's song about brushing your teeth.... and the lyrics it wont do are "[Chorus]
Shiny teeth, shiny teeth (shiny teeth, shiny teeth)
Gonna brush 'em twice a day, oh so neat (brush 'em twice a day!)"
hmmm...idk, did you add style?
yes try those lyrics and i bet it fails....i noticed this happening a lot more today than usual they must have updated a lot more restricted words which im tyring to say is very annoying....
idk whats happening since i don't work here, haha. i wish i could help
no problem did it fail for u too?
but do u know how it tell the staff?
no.
the staff will see the chat, and they might help out soon
maybe go to #🪲┃bugs-web
thanks
any plans for an API?
It's a bug from ChatGPT... they don't tell Suno the reason the prompt didn't work, so Suno gives a generic error that is meant to be the most likely reasons.
The work around for this very annoying bug is to type extra spaces at the beginning, and maybe some linebreaks at the end... anything to change the token count on the lyrics prompt so chatgpt will process the prompt a little differently (but not effect Chirp's song lyrics)
that is helpful there is a workaround, thank you
I had this happen to me earlier today. There was no issue with the lyrics for 6 continuations, then I clipped two lines out of my Chorus to use as an Outro and suddenly those same two lines of text blew up saying it's not allowed... jamming a bunch of empty lines before the [Outro] solved it
Please do not spam the chat with the same question, especially when it has already been asnwered.
Take an hour's time out while you give the #📜┃rules a look over.
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
So, I'm trying to continue on a song on the website. However, the song is on page 1, and when I click on "remix" on the song I want, I get to page 3 and can't continue on the song I want. Please help
(on desktop browser)
make the song play, but turn off the button that allows the next song to play.
the audio player at the bottom of the screen has it's own [...] button
if you do [Continue From] first, and play it from the thumbnail to load it in the bottom player, then you can get the [Remix] from the bottom player's [...]
hope that makes sense... I don't know how it works on mobile ....
Thank you!:)
That button was quite hidden, lol
How do I make the song continue from part 2? It wont put together more than 1x Chorus and 1x Verse from the first part.
I have part 1,2,3 but how do I put them together with the same "tune/vibe"? lol
you need to update the Continue From clip each time...
continue from clip 2 to make clip 3
continue from clip 3 to make clip 4
and so on...
if you don't update the continue from... it keeps making clip 2 over and over
some peoples suno songs sound better than others...is it luck or a process I'm missing? ( mean sound as in SONICALLY)
As for me, both luck and choice. I mean: if I decide to continue from a clip where the singer does a good chorus, it's more likely to get a good chorus or a nice sung part in the next clip...but in fact I get an off-key part quite often anyway 😅
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
/crip
The ai is random 🎲
but there is definitely a 'skill' to prompting it.
You haven't posted any specifics about your songs, maybe try to figure out what other people are doing on the ones that sound better...? Are they doing a lot of instructions, or not a lot, compared to yours? can you find tracks in your genre that sound good?
Thank you
If you move to a free plan, but have credits left on your account, when you create a song, what licence type does it use first?
if you have 'extra' credits (the kind that don't expire) they will stick with your acct until all are used up.
But the subscription credits will not rollover or transfer.
When your renewal date comes up, the acct will auto-switch to the new setting and you will get the credits that go with your current acct type (ie: switch from Pro to Free: at end of your month any unused Pro credits will disappear, but you will start getting the daily free credits....)
The main quirk of the billing is that it keeps your original renew date.... There is an option to change the acct immediately BUT it means you may get a full-month of credits but your renew date is only 5-days away and you don't have time to spend them all before the month rolls over....
Okay, thanks
Hello all, is there a way to add a few seconds of instrumental as an introduction before the first verse? I've played around and in many cases, the songs jump immediately to the first verse within the first 2 seconds.
To add, have any of you experienced an issue in which the output simply becomes an instrumental without the melody?
Hello all, I got a music style I LIKED from the bot... now i want music in that style only BUT it keeps giving me different syles... can i make it to get music like that only?
I am big fan of suno.ai
It would be nice to get some honest feedback for my latest development. I added an ai music generation using suno and also included a voice cloning.
Limited generations could be done for free but you can contact me for more credits for a testing.
https://www.aimastercraft.com
weirdly aggressive. first q was whether there's currently an API, second q was whether theres plans for one, a natural follow up
The answer still applies. Not at this present time.
And while it is possible that you were merely following up on the previous question, I asked a fellow moderator at the time and to them too it appeared as though you had forgotten you had just asked this question.
If this was a misunderstanding on our part, I sincerely apologize.
I want to make music in turkish arabic style by chirp how can I create dataset and training chirp?
new avatar 🤯
Yeah. I used a GPT in ChatGPT Plus (while I still have a subscription) to make a new logo for me. Tried it out as a avatar.
wee
Seems friendlier than the one I made for christmas...
At the moment you can't. You can only prompt for a style, which will generate a random vocalist, rhythm, instrument selection, etc.. .every time you make an INITIAL CLIP. Once you have made the first clip, if you like everything about it, you can continue from that clip and the next clip it makes usually follows the initial clip pretty closely. See the #⏩┃continuation channel for help
Sometimes using the tag [Intro] will do that, but it's currently unreliable. It seems to depend on the style that gets selected by your prompt. Some of the style prompts I use seem to consistently generate an intro of variable lengths. For example, my songs with "heartfelt country ballad" as the prompt seem to honor [Intro] fairly consistently.
I've had the most success with (Intro). Do an (Intro) tag, followed by two blank lines, and then your [Verse]. Keep in mind there's a chance it won't work but I've gotten the most consistent results using this.
In (parens) not [brackets] ? Interesting!
Yeah it is indeed interesting. In custom mode, I'd get the AI to generate lyrics with an instrumental intro. It'd do [Intro] followed by (Instrumental) on a new line. When the community wiki said that [Intro] was largely unreliable, I tried stuff like (Intro, hook) or (Intro, build) and results were way more effective. Seems you can do that with instruments as well, such as if you generate a verse and in the next verse you want drums or something, do (Drums) before the verse tag and that also seems to be effective. Wish I could find the song on Suno site where I first did that. Seems like anything not lyrics related can go in parens outside of the song structure. If you put parens inside [verse] content the model will think adlib.
So is the prompt
[intro]
(intro)
[verse]
Normal stuff
I've tried that and while I've gotten it work, it seems to just default to the [Intro] behavior and assumes there's going to be lyrics so just not effective. Maybe the devs will enforce that in the model though.
What I do is
(Intro)
[Verse]
lyrics
Doesn't seem to work without a style. I thought it would given on the backend it just prompts for a style anyway, usually rap or pop or something.
The big issue for me right now is getting the song to end the way I want. If you do [Fade out] whether brackets or parens, it will usually make a long repeating instrumental where you can fade out in a DAW or something. But I have gotten it to do a nice fade and end.
I usually do [Outro, Fade Out] and it's worked pretty well for me most of the time
Ah, okay. I like doing an [Outro] tag on it's own and then do [Outro, Fade Out] to basically make an extended outro and hope the AI doesn't cut it off midway.
I usually resort to that when it doesn't seem to want to end the song...
As an example anyway, I'll have to try [Outro, Fade Out] to see if I can get good results with that.
Which people want an [intro], how many seconds long is ideal? What's the ballpark you expect/
Probably 10-15 seconds. I've tried (Intro, 15 seconds) but that doesn't seem to work very well. I've tried 4 bars and that only really worked once and I'm not sure whether it was four bars...I counted and it osunded correct. It's in showcase buried there somewhere.
I've tried telling it the "bars" and I've never had success with that... I think the times it came out right were probably more accidental than it actually paying attention to my prompt
I just spent 500 credits testing
(intro)
[verse]
Honestly, I want an intro because I hate how songs can sound like they start in random places without one, as though the band already started playing and the producer edits the recording to start where the vocals start.
however I only used the same lyrics and style so this may be not reprentative
(intro) [verse] by itself almost never worked. But there was a tweak that worked quite often and I liked the results a lot.
Adding it also to the style field like
https://app.suno.ai/song/0b97dc69-1ad1-45ce-b44f-a2a5c8c0c853/
https://app.suno.ai/song/493e1792-9ffb-408c-8287-031116cc1826/
(just two random examples, not picking best quality or anything, just two that had an intro). worked both with , and without comma.
pop acoustic intro song. Listen and make your own with Suno.
pop acoustic song. Listen and make your own with Suno.
I want an intro of at least ten seconds, depending on the song a longer one. But I tend to go from feast to famine with those.
Too long is probably better than too short, since you can often chop off the beginning to shorten it
Is there to make them long?
As long you can still get the verses in without crunching the singing speed
very tentative result of this test was the best results from:
Style: blah blah, intro or blah blah intro (not sure which)
(intro)
[verse]
And optionally adding some descriptive words into the (intro) in the lyrics.
Still fails quite often, but closer to 50% on this particular prompt
Here's one that worked for me, I'm not sure how many words can go into the style box before the aI is overwhelmed, but I got a clip with an intro and an instrumental piano clip fir the second one so it didn't count. https://app.suno.ai/song/008f414c-a3c7-4ed1-a790-1bead91e00fa
clear close perfect female song. Listen and make your own with Suno.
If you stack "intros" it seems to work more often but changes the music genre a lot, so not without tradeoff. Like it makes it less like what you asked for.
Guessing that is a no
Maybe putting it in style prompt, or repeating (intro) in lyrics on multiple lines?
I will test
(Intro, hook)
by itself. Maybe that works without having to use intro in style prompt. IN general I find having to put things in style prompt is a bit of a tradeoff
I can get the lyrucs go past a tiny bit
If you want a very long intro, probably best method is to look up techniques for prompting instrumentals and just start with a whole verse of that, and then only one verse of lyrics.
You will run into the case where sometimes it never starts to sing 45 seconds in, and fails that ways, but should work sometmes
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
On my prompt (Intro, hook) worked 3.5 out of 12 times. That is better than (intro) not quite as often as also putting it in style. Intersesting.
I'm giving .5 to 3 or 4 second intros like https://app.suno.ai/song/7f3325a4-ecb2-412e-bd62-60044e59e252/ - it's better than starting from nothing.
pop acoustic song. Listen and make your own with Suno.
@untold briar Build might work as wel, seems to create a bit of a longer intro when it does work. I basically went through the metatags for intros on the wiki and tried as much as I could, E.G. Refrain.
Followup to intro in the style prompt: Doubling both the original prompt and intro as in pop acoustic, intro, pop acoustic, intro, pop acoustic, intro, pop acoustic doesn't seem to trigger intros nearly as often as pop acoustic, intro does. Interesting...
Could be RNG, I already have 100+ mistletoes in my library and not testing as many at once now
Here's one with a longer build. Probably not 15 seconds though but putting a random seconds number seems to reenforce a longer build in any case. https://app.suno.ai/song/1a4afb65-163a-455c-b08a-beab8cc015cf
clear close perfect female song. Listen and make your own with Suno.
|vocal only| (Intro, hook) is what I will try
Here's a fascinating hallucination. Chirp hallcuinates the word "yuletide" for the (intro), even though this is NOT a continued clip where that was song in the song history. The word yuletide is only used in the later part of the song. But Chirp is moving the word to the front of the song? That doesn't fit my mental model, weird.
https://app.suno.ai/song/3c70eb10-d055-4189-90e2-7ba112d28fdd/
pop acoustic song. Listen and make your own with Suno.
For vocal only use A cappella in the style prompt or something... intros are usually classified as intrsumental beginnings but can be vocals only. I noticed that pop-punk as a style always would do vocal intros followed by a heavy guitar riff.
Thanks I will try that
I have issue where it keeps cutting off the lyrics
and stops around 40 seonds in to 1 minute
Yeah, if you're new, each clip can only 1:20 long.
I have noticed
but these do not make it to that point cutting off early
of it ends up sing rubbish before it starts reading off lyrics
hello i was pay to site.but bot say me you dont have enought credit
even if you pay you get a limited amount
So, if I got a song that's almost done but it keeps crashing generating the last extension is there any point in keep submitting it and hope it goes through or is it just corrupted and I need to go back and regen from the 2nd to last?
If it won't go through after several times, it's good practice to just run back up the chain and try again from there.
ty. I was hoping to be done, lol. This one's been fighting me and I was finally seeing the light at the end of the tunnel and then a string of "Credits refunded". 😛
We've all been there. I had to do it earlier today myself when the vocals went thin on me.
I think its going to be crazy time till the marshmallows fly he's a crazy ma a a an oh he's a crazy man
But it looks like I still have credit on the website.
This is ridiculous. Part 4 of a song. I got one generation already. All I added was [interlude].
That may be a bug. If it's reproducible can you add it to this form? https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
Thanks. submitted.
In case anyone else has a similar issue, changing it to [instrumental interlude] fixed the issue.
Same here!
???
Eta for continuing from anywhere I like…
/chrip
any chance of the chirp models being open sourced in the future?
Please do not post the same question in multiple sections. It is a violation of the #📜┃rules you agreed to. #3️⃣, specifically.
criar musicas on line crie uma letra de música instrumental em inglês com 4 frases no verso 1 e 4 frases do refrão
@marsh berry English
Boa
I have four songs in my browser-based account that are still "generating" after 12 hours.
If you're on a free subscription, and your credits have refreshed, I think you should re-gen those clips...
they may never finish....
(not technical advice, I don't know why sometimes the gens hang like that, it is usually only when an outage happens...)
it's the paid subscription and I don't think the credits refreshed...not that I care about that.
Have you been able to gen since?
standard answer is to contact
billing@Suno.ai
maybe ask those clips to be refunded...
(thanks for reporting the issue, some people mentioned chirp being 'down' yesterday, but it seemed like maybe a scattered issue... and I didn't hear if chatGPT went down)
The lyrics were written, but the song didn't generate. And a while ago when I clicked one of them I got a "server-side error" but I wasn't able to replicate it again.
Hi guys, I just used up all my credits for pro plan, really suggesting to have at least some extra free credits distribute per day as it doesn't feel quite fair that free basic plans have 50 credits per day that makes up 1500 credits a month, pro plan is only providing the free credits at once with 1000 more credits...
Please add your voice to official https://discord.com/channels/1069381916492562582/1154587937611190354
I think this would be fair.... imo, Pro subscription should have everything free subscription has, plus perks of subscription.
(I'm not Suno employee, and I buy top-ups 🙄 so my opinion doesn't matter)
thanks man
It says I don’t have enough credits using the chirp on discord and I have a ton, I need to finish my song….
Check that your Discord account and whatever you used to do the subscription login are linked. If that doesn't work for you please email billing@suno.ai to be assisted in this matter.
is there a way to reset suno?? mine got stuck, already for more than 13 hours
I think the "main" thing you are paying for is ownership of your creations, not the extra credits.
Of course, I'd love it if they also granted the 50 free credits per day to anyone who pays.
And the right to monetize your songs, not attributing Suno on sharing them, and so forth if I understand correctly. It's not only about the credits.
How can I make a mal and the a femal voice in one song?
Sorry?
What is a mal.
Wait, did you mean male and female voice...
I must be tired today.
...
Duets are not easily done. Some have had successes with prompting in their lyrics and giving the characters names in [brackets] like metatags.
I've had very inconsistent results with it personally and the few times I succeeded the genders of the vocals were often switched.
men and woman 😉
I mean
thx, what are good names for men and woman and have I to add there [Verse] [Peter]
I've been tryna call...
Let's say in a duet, you want both singers singing, but each are singing different verses at the same time. How can this be accomplished?
It really can't right now with any level of consistency.
I ve seen it on the wiki . I tried it and it did not worked for me
[Mary]
Singing my part
[John]
Singing different thing
[Mary & John]
Singing together
Did not worked for me
Any clue?
I've had it kind of work once.
The one that I know who seems to get a lot of mileage out of it is Jonathon Fly. You'll have to see if he comes in later tonight and ask him about it.
That’s rough. Works is 10% cases. Try also my favourite [Rap Verse], sometimes man sings (if previous vocal is female).
I've emailed billing earlier today, and I haven't heard anything, are there delays?
I am still unable to end a song in a clean way. It always stop suddently. I have tried [outro]. What is the recommended way to end à song? Any idea?
Listen to We Believe on Suno! 🎵 https://app.suno.ai/song/689b12dc-6db2-49a8-b666-aa83dcb4e88e
What exactly does the remix option do? It doesn't seem like it does anything.
Many of us think it shouldbe renamed "Re-Prompt" for now. All it is supposed to do is copy the previous text in the lyrics and the music styles to let you try again. It does not guarantee the same melody or vocals.
Indeed. Basically it's the same as not having that option at all if it's not the same style or vocals. 😵
I've had some luck with
[Music Fades 0:05]
[Song Ends]
The 0:05 absolutely is not understood, it never does 5 seconds of music fading but for some reason it seems to respect the music fades more if I put it in. It's still very hot or miss, sub 50% but my best success so far.
Does anyone know if it's possible to vary the melody in a [Verse] with Metatags? I get so many songs where 4 lines of verse all have the same melody. It's a bit boring 🤠
Have you tried adding instruments? Describing the instruments? Their playing style?
Do you use the style prompt to its fullest potential with things like catchy hook or other stylistic modifiers?
How about solo? Do you use those in the lyrics to prompt it to add those?
No, I haven't tried any of those, but I will. Just like to make it more like a real songwriter would do 😁
Understood.
Have you used metatags much before?
Do you use custom lyrics or use the automatic feature aka "Simple Mode"?
I use standard metatags, , verse, chorus, instrumental, etc. I mostly write my own lyrics.
@untold briar Keffiros and Farafina have some questions about duets if you can help them?
That's up here: #📚┃suno-school message
Have you looked into the list from v1 Chirp?
You can find that here: https://suno-ai.notion.site/Using-Metatags-34944efe09ec41d693e314f13af44695 Just make sure you click on the section for Chirp v1 because the ones for bark don't work.
Metatags guide prompts but aren’t verbalized themselves. Like stage directions or parentheticals in screenplays, you can use metatags to describe a specific part of your prompt (e.g. [laughter] and piano interlude) or to influence your entire prompt by describing speaker characteristics like [whispering] or music genres like rock.
OK, thanks, I will check it.
Glad to help.
With the right metatags and some careful selecting of parts to continue on and some various tweaking I'm sure you can keep things interesting.
Is there a metatag you can use to have certain lines\lyrics to be halftime or double-time ? I manage to get it once but it seemed to be random when it happened
I do not know of anything specifically. I do apologize.
I've got a great first chorus. but none of the subsequent choruses are the same as the first. Is there a way to make sure that all choruses sound the same?
I often use "keep the same chorus" before the lyrics, then click on Generate, then I delete the generated lyrics and paste mine. Hope it helps.
Thanks, I will try it. You enter it as a Metatag [ ]or in parentheses or as free text?
I enter it as free text and it disappears after I click on Generate but it probably "stays" in memory.
Please remember that it's not 100% guaranteed, because you can get something different (usually off-key), but on average it works.
I think that's just a coincidence - if the song doesn't show the words it should be having an effect. (If it does work it'd be a bug actually). If you're talking about making the music melodies of the chorus the same in all the repeats, the best you can do is make sure you use the exact same lyrics, same text formatting, and same label like [chorus] in the lyrics. And then try and make the chorus be towards the last half the each audio clip repeated so it's not too far back in the memory/history.
Actually, I made a song using my "method" and changed a few words in the chorus every time and it worked: the "singer" actually recognized the chorus anyway.
Yeah it should work with just the label. In fact I often change every word in a chorus and it almost works fine. But if you are having trouble with consistency then making sure the same may help.
For some reason, the neural network believes that two verses in a song are enough, so in the second piece it simply generates an ending (takes the sound to zero). In such circumstances, the third verse refuses to be generated, as it seems that the silence at the end of the previous part is a model for it. How to defeat such a phenomenon? If it were possible to trim the unnecessary ending yourself, I think that the generation would work as it should, but, unfortunately, we do not have such a tool, what should we do?
from my understanding, the feature to clip that silence off (as well as other portions of the generated clip) will be coming soon™️, according to a team member the other day
Can I use Suno's ai generated songs for games or videos without any profit?
See the FAQ: https://suno-ai.notion.site/b72601b96de44e5cacd2cd6baa985448?pvs=25#f0f1441180484d6094206e84e334ba36
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
so i can use it for games?
(without profit)
If the FAQ is in any way unclear you should do what it says and email them at the address listed in the FAQ.
Just understand that the emails may take some time to get a response as there have been an unusually large amount of emails recently, which is an issue the team are working to address behind the scenes to ensure a more timely response time for everyone.
'non-commercial' isn't really the same as 'un-profitable'... (plenty of commercial projects are unprofitable, lol)
but using the music within another non-commercial project should be fine, just the same rule about giving credit to Suno.
"A song dedicated to my best friend" triggers the filter in the song description
ok
try 'best-friend'. 🤕
Still triggers
Might be a straight up bug, can you add it to this form? https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
Submitted
/
hey, wondering why it primarily churns out male type vocals? do we have any hierarchy of input for the AI such as pop, vs female pop, vs pop female? Like what would be more dominant? Lastly, it often returns 28-36 seconds without having finished lyrics I gave it, what's this about?
Please be mindful of the #📜┃rules here, especially rule #3️⃣,
Your question has been answered at the other section you posted it.
Take five minutes to read those rules!
Hello, in the pro version, is it possible to export in wav format or the same as free version?
same.
Thanks, it would be awesome if they could add it in the pro version
Not yet, likely in the future, maybe soon?
FWIW there is little to no practical benefit at the moment to higher bitrate (the audio problems you hear are very likely in the raw audio file even before compression, from what I understand) but as the AI gets better it will be more important. (And there still may be some edge cases where .wav would be nice, even now.)
Yeah, for example most of the pop songs has good output, but for hard rock songs, its very noisy
Just a suggestion / request, hopefully in the Pro version, these features would be nice:
- different export format (mp3 / wav etc)
- some basic commands to understand [male/female changing, harmonizing, adlib, acapella]
- prompt wise output sample [so user will know which command to put]
- instrument selection
- a wide range of voice genrewise
- a special feature [user can sing in a tune and later, it will be kind of same way; just like we do for the image geneartion, we can give sample images and later, it gives output similarly; here it would be nice]
Hopefully there will be better ideas by Suno developers, its really awesome and loving it!! A huge shoutout to you guys!! In coming days, we will see more development.
I was wondering how does the ai generate songs, is it like a stable diffusion type thing or what
Anyone know how to make the bot stop missing [Melodic Instrumental] and [Bass Drop] ? it never does it. only once in a blue moon. it ignores it 99% of the time.
I keep getting songs without vocals, what's up with that?
too mixed genre, probably
at least I've found out in past that if you can't make chirp sing, it's typically coming from the genre you ask it to perform, some combinations just confuse it and it never gets there
getting songs occasionally without vocals still happens, that comment was just if you constantly get songs without them
Can we invite chirp bot to our discord servers?
Yes, just bear in mind that the Discord version is behind the webapp version and that when the Discord version is updated to have parity with the webapp version any songs made with the earlier version will not import to the new version.
This document should help you get started: https://suno-ai.notion.site/Add-Chirp-to-Your-Discord-Server-997a3121b373431695b953ce164800b3
Great 👍 Ty - awesome project guys!
Any way to get the light-sounding pop rock sound (like this https://www.youtube.com/watch?v=mOAUfsHlFK8 )?
https://app.suno.ai/song/382948a8-c9b0-4704-8f3b-13f6757fe15f (yes, I made a whole song out of this) does not sound rockish.
https://app.suno.ai/song/d32a0c22-c6aa-4bff-a532-51bb2d13d3d8 sounds too rock-oriented.
Provided to YouTube by TuneCore
Year 3000 · Jonas Brothers
It's About Time
℗ 2006 Jonas Enterprises
Released on: 2006-08-08
Auto-generated by YouTube.
Fast-paced, upbeat, energetic, happy pop rock song song. Listen and make your own with Suno.
2000s pop rock, energetic, female vocals song. Listen and make your own with Suno.
good luck lol, rock is hit or miss with the bot idk why.. metal is SO good but rock is dookie with the bot
https://app.suno.ai/song/c29e968b-ca4e-48f3-b01c-b18534c076e0
https://app.suno.ai/song/db7e6d42-c7e5-4829-8397-29eee47fca80
think i got pretty close, used the same theme but not the same lyrics as your reference song
Pop Rock, Punk Rock, Guitar, Drums, Bass, Keyboard song. Listen and make your own with Suno.
Pop Rock, Punk Rock, Guitar, Drums, Bass, Keyboard, Male Vocals song. Listen and make your own with Suno.
Hi
I am having part2 better than the beginning of the song.😀 Now i would like part2 to be the beginninng of the song. This seems impossible to do but if anyone has a method please share. Thanks.
Is it just me or is getting upbeat, feel-good EDM (like this old favorite of mine https://www.youtube.com/watch?v=iaKgF1Vf5bQ ) a hit-and-miss?
Subscribe to NoCopyrightSounds 👉 http://ncs.lnk.to/SubscribeYouTube
NCS: Music Without Limitations
NCS Spotify: http://spoti.fi/NCS
Free Download / Stream: http://ncs.io/entropy
▽ Connect with NCS
Facebook___ http://facebook.com/NoCopyrightSounds
Twitch______ http://twitch.tv/nocopyrightsounds
Twitter______ http://twitter.com/NCSounds
Spoti...
everything is, not just edm. need to gen enough times to find something that's decent
Not sure if this counts or not? And I was focussing on the feel and not the vocals so if those are annoying I apologize in advance.
https://app.suno.ai/song/93f57b48-9882-4a05-ba0e-50011dc54641
Words are (C) 1919 Joeseph McCarthy and in public domain.
edm, electronica, upbeat, euphoric, house, catchy hook, warm soulful vocal song. Listen and make your own with Suno.
What the hell is this 😭
LOL i had very low hopes for this after reading the lyrics but once that beat drops it goes pretty hard
I honestly don't get how people get the standup thing work with this 😄
I already gave up, it clearly ignores every attempt I've given it, with exact same genre prompt and trigger words that worked for others
would have been fun tho
could still do the audience on post but that kinda beats the fun of this
14 failed tries is enough for me 😁
spoken word comedy, storytelling, quiet environment, audience laughter, no music song. Listen and make your own with Suno.
It's not consistent but then I've not been doing it for very long.
This segment refused to do any applause or laughter really...
https://app.suno.ai/song/55c43d76-1f57-425f-95bb-4c22c2dc63a6
spoken word comedy, storytelling, quiet environment, audience laughter, no music song. Listen and make your own with Suno.
That bit in the beginning was it.
spoken word comedy duet, quiet environment, audience laughter, no music song. Listen and make your own with Suno.
Wouldn't let me finish it out...I assume because of the Dr. Robotnik reference.
For those who have never had the pleasure before:
http://megalomaniac.com/~andrew/funny/bloodcyber.html
what does remix do
it re-prompts the lyrics and style of music instead of having to copy and paste
Thanks
I like making suno music with the worst lyrics i can imagine
Hello. How can I unlink my credit card in suno? because I don't see any option
What does prompting (...) do?
Sup guys, I have a question: How do I explain the ending of a song to Suno? Example: at the end of the song, the singer prolongs the last word, and the volume of the song decreases until it ends. How do I do that?
bro where are my lyrics??? 😭😭😭
I don't think Suno can take that kind of prompting exactly.
Im asking it because on Suno's official website, in the "Explore" part, most of the songs have something like this at the end. Or at least an ending without lyrics
To see what one of those songs did, you can click on one and hit remix to see the exact prompt including style.
I wish we could have profiles on Suno so we could browse creations. I'd be curious to try random stuff @untold briar has done to push the AI to it's limit.
oh thx
I would try [Outro] before the last lyrics.
Here is an example
https://app.suno.ai/song/7f101110-4310-45a7-aa2b-cd92f19484dc
Swing Electro-Folk song. Listen and make your own with Suno.
This one didn't have a very good finish, but the rhythm was very good! I wish I had some knowledge about notes to do something like this
i think the outro will form based on the song.
If you have GPT you can try this
https://chat.openai.com/g/g-ldbohmO2c-compose-me
it will add the chords for you
I hate being poor
I don't know if it's just luck and repeat generations, but I swear by [Music Fades 0:05] [Song Ends]. This is probably the best closing I've gotten out of it. https://app.suno.ai/song/3133f0df-22a7-47b6-a015-3a0590a20188/
girl rock, 80's metal, ballad, haunting song. Listen and make your own with Suno.
thats awesome
exactly what I want to get
I will try to use this, thank you very much
heavy metal, catchy hook, playful, female song. Listen and make your own with Suno.
I just had the exact same problem!!!
mine doesn't even have music lol
and the audio sounds like it was recorded in a church with a lot of echo
sometimes it gives space, so things don't happen all at once.
Hmmm! Getting half of someone else's song today! "Feelin' it, feelin' it" if anyone is missing some lines, I've got them!
I kinda got the duet thing to work. ChatGPT generated the metatags; I use it to try to help me prompt. https://app.suno.ai/song/96e9218e-d365-4169-9fdb-ebbb74dd3b13
That is so cool! How did ChatGPT generate the metatags, btw?
In custom mode, instead of writing lyrics, prompt with an instruction such as make a duet between two singers. It's ChatGPT so it really only knows what metatags Suno trained it on, [Verse] and [Chorus], but I've gotten it to create other things such as intros etc. Hit generate lyrics and yeah. Now I'm trying to prompt it with the tags that don't work etc to see if it'll eventually come back with something that does work consistently. I don't think it will just because the music model just wasn't reenforced on the concept of duets.
ChatGPT seems to love the system instruction "You are a master song writer. Help the user complete their song:" LOL
/chirp 老家墙上挂历又翻一年
春夏秋冬是一年跌跌撞撞再一年
抱着妈妈的叮嘱慌张了一整年
你看镜子那个又老了好几年
/chirp 老家墙上挂历又翻一年
春夏秋冬是一年跌跌撞撞再一年
抱着妈妈的叮嘱慌张了一整年
你看镜子那个又老了好几年
Rhythm: [Syncopated] - Controls the beat's pattern and pace, bringing rhythm to life.
Melody:
[Ascending] - Shapes the sequence of notes, forming the song's main theme.
Harmony: [Gm] - Underpins the melody with chordal support, adding depth.
Dynamics: [Crescendo] - Adjusts the music's volume, from soft whispers to powerful crescendos.
Texture: [Polyphonic] - Layers multiple sounds, creating rich musical landscapes.
Timbre: [Warm] - Colors the sound, defining the unique quality of each voice or instrument.
Voice: [John] [Chorus] - Specifies vocal characteristics or identities for song sections.
Outro: [Fade Out 0.5] [Song End] - Concludes the song, guiding its closure with a specified effect.
Hello Chirp and Suno TEAM!
Do you guys giving API's for websites ?
there's no API... Haven't heard any plans or announcements about an API...
That would be an awesome addition, people could build SAAS applications that tap into Suno.
I've been told by management it's on the roadmap...
Dear Suno team, I'm a blind musician trying to get to grips with app.suno.ai but I can't handle the login as my screen reader thinks the page is completely empty. So here I am using the Discord version of Chirp. Could you please tell me how to access the stereo mode?
unfortunately, stereo is only with the new v2 model on the website.
Discord is still using v1.
I will pass the information along to Devs, that the website needs better accessibility.
I'm sorry it isn't useable for you.
Discord will be updated eventually, although I don't know exactly when.
Hey, great that you'll pass it on. I have already sent two emails to hello@suno.ai and didn't get so much as a reply. That address isn't monitored?
I'm sorry about the email silence...
I think the Devs are working through an email backlog... I don't mean to make excuses or answer for them...
That email is the 'front reception' for business inquiries, but Suno is a small team – I assume they have seen it but have no formal answer or solution yet to share (?)
I know they have been actively working on technical aspects of the website behind the scenes, to bring the same functions to Discord.
I don't think there is a specific email for website issues... I've left your details in our Mods area where the Devs will see.
Thank you for bringing it up.
Thank you so much! 🙂
Hello, I created a beautiful intro that I wanted to advanced into part 2, however I noticed that in part 2 it never generates vocals (the intro of part 1 is only electric guitar rythm), is there a way to fix this issue somehow??
Thank you...
I find it's hard to get Chirp to start singing again, if there was no voice at the end of the clip before...
I have forced Chirp to start singing on the 4th clip, and again later... by spamming the style prompt:
Sultry Female Vocals, Sultry Female Vocals, Sultry Female Vocals
but on other (most) songs, I was not able to get Chirp to start singing at all.
In theory you could also use it like a metatag:
[Sultry Female Verse]
If you are able to copy and paste javascript code into the Chrome developer "console", this line of code should allow your screen reader to read the login page popup (link to earlier Discord message with the actual code): #💬┃general-chat message
Yooo you’re a life saver thank you
Thank you! That helped a lot, and I was now able to sign in with Google. Took some tabbing around but now I'm in and already going wild with it.
Hello to you all,
I have a problem, maybe you can help me. I have a complete song text that goes over six verses and I would like to create music for it. When I enter the text, it stops after 1:20 in the middle of the text. If I want the text to continue, I have to enter a new prompt. But I only want my complete lyrics to be output. Is there a solution for this?
This may be of use to you: https://suno-ai.notion.site/b72601b96de44e5cacd2cd6baa985448?pvs=25#baf0eb82963f467488f5b6f167ba8102
Hello! Please email billing@suno.ai if you are unable to find the answer to your question on this page.
Do I have to rewrite in the middle of a sentence?
Yes, depending on where the vocal stopped.
for Voice: how to adjust the volume of. the voice, and how to remove chorus ?
Have I secretly uncovered Lady Gaga's next album?!
Nah, it's the REM thing. The content bot is stupid enough to think REM owns the words "It's the end of the world" as we know it (but I feel fine) which is exactly not how copyright works.
https://app.suno.ai/song/7cf9809f-dfd3-43b0-a0cc-46cc40c70706
https://app.suno.ai/song/76259d1d-1baa-43aa-8c8e-4fc7a7f246e4
That's very far fetched but I appreciate the explanation
Funny thing is the second one says "it's the end of the world" in the lyrics it created but won't let you prompt it.
Can you please submit this to the form: https://docs.google.com/forms/d/1I6B1GyBSMV3apIaxJMvgZ19nHwFUzUFieCmOoL2s274/
Error: "Couldn't generate that. Suno is for generating original music. Try different lyrics, and use genres and vibes rather than specific artist names." This means your song was blocked and credits refunded.
Songs on Suno.ai are moderated for:
Non-original lyrics (copyright & artist names)
Being overly explicit (in general or in the use of s...
Mmmmmm
If you think it's a Moderation Mistake, you can submit your prompt to this unofficial Google Form to help improve the moderation system. (This isn't an official Suno report, just a form setup by Discord mods.)
Says a lot about the state of pop music nowadays 
No matter how I reword it
Ok this one seems to work
Ah! Could it be this?
https://genius.com/Beck-super-cool-lyrics
Beck apparently owns Super Cool?
I mean, if that's the case, at this point, we shouldn't be able to gen anything
everything's been pretty much done
They're working on it....
https://app.suno.ai/song/f83f71ab-8f0c-4d7b-b551-9aecc36643ce
https://app.suno.ai/song/efc1f8e2-0361-4964-88f1-ddb166a6f990
I changed it to: super awesome pop song, girl forgot to paint her nails, girl thinks it's the worst thing ever
It's actually really good
I'll go back to writing my own lyrics
it works better and it's hilarious too
Thanks for your help @neat nexus
Glad I could be of assistance.
Sometimes it works ok
Just had to swap it to world's end for me...
Thank you very much. Unfortunately, the transition is not smooth. Is there a solution for this? Possibly with a Promt? There is a 5 second break, which doesn't sound natural.
"ear down the walls of yesteryears, Forge ahead, conquer ...... " 5 second pause then " in shadows, chasing" I would like to have this more fluent? Although, it did not appreciate me teasing it...
Although, it did not appreciate me teasing it... 😛
It can be a bit of a challenge to get things to line up just right. Have you tried getting the full song from one of those parts to see whether it sounds unnatural when stitched together?
This is the routine I follow:
Try depending on where you are in your lyrics (if you have a lot more to go) just continuing and running the same lyrics you have. This can sometimes make it pick up where it left off. A large percentage of the time it will start over in the middle of the song. This is just the easiest thing to try to see if it works.
Next try removing everything already sung but the last good line, then try continuing. This sometimes works. It's how it always seemed to work in v1 Chirp for me.
After this try creating a line break (empty space) and continuing from the line that was interrupted. This has a good chance of working.
Then starting with the line that was broken. It sometimes will "remember" where it was and pick up from there.
Finally when all else fails, batten down the hatches, remove everything up to the word that was interrupted and hope for the best and throw the dice. You will likely encounter some weirdness and you will need to keep trying. But if everything else has failed it's this or...
...or you go back up a chain in the song and try again giving up on the good part. It hurts, I know, I know. But sometimes it's the only way forward.
...As you can probably tell this is a bit of a pain point for me and I've done these often...
I redeemed a lot of credits for a perfect transition. I know that AI can only fulfill my wish to a limited extent, but I would like to have my song created "perfectly" according to my wishes.
@neat nexus I have little experience with AI, but I know what I want. I think your tool is really cool and I'm happy to pay more for it, but I want "my song" to be perfect. I have a lot to tell and would therefore be happy to pay more. Even if the AI has its limits, I would prefer it if the entire lyrics were transferred into a song. What prompt would I have to enter for this so that I don't get a pause between the passages?
First of all, it's not my tool. I'm just a discord mod here who also makes music with Suno, just like you do.
Secondly I gave the advice I could above in quotes. There is no perfect prompt. All you can do is try, try again and test things.
Y'all seem to be forgetting this is all very new technology lol
Give it a couple years
For what it is, it's quite incredible already
Unfortunately, the "whole song" doesn't really work well. Is there perhaps also a prompt? I am really very satisfied. but the "small" details are still missing. ok. i understand that it's not your tool, but i see a lot of potential. That's why I don't see it as an "attack", it should rather serve as a suggestion for the developers. I would like to exchange ideas with you. Maybe we can help each other. I don't come from the music industry, but from the financial sector. But I know a bit about IT and have a clear idea of what an AI needs. But most importantly: I am happy that you are sharing your experience with me.
I would like to avoid that.
About the developers: Unfortunately, the transition from part 1 to part 2 doesn't work very well. The text transition from part 1 to part 2 doesn't fit very well. Is there a clear rule here? No matter what I insert as part 2, it doesn't fit to part 1. Just because I have a subscription, I would like it to work properly.
Unfortunately, the transition from part to part is very faulty. Pitch, style, text and also text breaks are very faulty. I would like to support you. What do you need?
Anyone have any tips for song endings? I have a couple of songs that turned out really well but just "end" after the last lyric with no transition at all. Is there any workaround for this?
[Outro] and [Fade Out]
THANKS!!!
Very helpful. Appreciate it.
for Voice: how to adjust the volume of. the voice, and how to remove chorus ?
Have suno turned PC on us? Can't use some words and sentences
I ran some numbers on this Duet format with two tweaks. 10 chirps each.
I think there is a likely improvement in hit rate if you add a full line break before each singer switch. 10 chirps is a small sample size but it was 300% for the second format. Worth a shot.
Version 1: (Just add line breaks)
https://app.suno.ai/song/1bb4f99b-73bb-4ccf-8e83-bc1d22da8615/
Version 2: (Add line breaks, remove typical [verse] and [chorus] tags
https://app.suno.ai/song/c3b7bf99-117b-47d4-8104-8ecbea4141dc/
Version 2 seemed to work as often as 1, but also resulted in more semi-duds where it didn't finish the song.
TLDR: Whatever format you are trying for a duet, maybe test a blank line before each singer switch. (Note, the video removes the line breaks from the text, so the prompts look the same.)
epic song. Listen and make your own with Suno.
epic song. Listen and make your own with Suno.
cc @neat nexus Maybe a concrete Duet tweak to improve the odds. Just one test, but worth exploring.
I'm going to recheck the OG prompt, just in case I got a bunch of bad rolls actually.
Nope, holds up. TLDR, adding blank lines:
Original:
[Verse]
(johnny)
I've been waiting here for you
Since the day that you went away
(sarah)
I've been dreaming of your smile
Hoping that you'd come back someday
Potentially works more often:
[Verse]
(johnny)
I've been waiting here for you
Since the day that you went away
(sarah)
I've been dreaming of your smile
Hoping that you'd come back someday
Nice. What about songs where a whole verse is sang by a different singer? Having trouble getting this scenario to work even with line breaks. I guess this one partially worked though sometimes I can't recognize if the male is just higher pitch in this instance. https://app.suno.ai/song/d99fabdc-80fd-4dc4-894c-1ba33920b825
Is there a way to combine accounts? Just realized I created a song on my Google account which was not subscribed. Or is there a way to recreate a song on another account?
Is there a comprehensive list of the prompts that Suno understands anywhere?
This is an old list (v1) but these still seem to work, Make sure you're in the Chirp part and not Bark as they have incompatible sections:
https://suno-ai.notion.site/Using-Metatags-34944efe09ec41d693e314f13af44695
Metatags guide prompts but aren’t verbalized themselves. Like stage directions or parentheticals in screenplays, you can use metatags to describe a specific part of your prompt (e.g. [laughter] and piano interlude) or to influence your entire prompt by describing speaker characteristics like [whispering] or music genres like rock.
Hey guys, aside from the basic stuff in the notioin docs....is there any written description offering a coherent clear explanation of the processes suno uses to generate audio. I know they dont reveal specifics, however, Id like to create a working set of parameter guidelines that can help us create more focused and accurate results. Anyone with any input would be appreciated.
No official talk from Suno Devs...
But if you've ever used an LLM for chat or roleplay (as opposed to instruction/coding), Chirp's Continue From This Clip seems to be following that model – it retains some memory of what has happened within the session, but is literally continuing from the end of the last clip – the way a chat/roleplay session is just continuing whatever was last in the thread....
Thank you. That's exactly what I was looking for.
yep im defining a comprehensive command list through prompting a framework library in gpt, having more of an understanding of the specifics of sunos diffusion model helps clarify. so i get the general terms...but its akin to telling a nuanced joke about cullture to a frenchman, using only a tricolour school book from the 70s
is it already clear if one's own voice (say Elevenlabs API) can be thrown in to a chirp?
no... not through Suno.
You might be able to replace the voice with RVC-Replay or similar.
Question - Am I tripping, but intro does the non-extended version sound better than the intro in the "full song" version - These should be identical, but I feel like like initial snippet (not the full song) has a better sound quality. Is this a known thing that we're working on, seems like extending the snippet may decrease the sound quality a little bit?
First snippet (is the better quality)
https://app.suno.ai/song/fb2b2e83-9dab-496e-afeb-c525995a72f6
Full Song w/ Two generations combined (it has less punch in the begining)
https://app.suno.ai/song/e7d7a40a-baaa-4b89-a4b4-4175400995ea
Boppy, Retro, Funky Jazz, Energetic Synth song. Listen and make your own with Suno.
Boppy, Retro, Funky Jazz, Energetic Synth song. Listen and make your own with Suno.
It does degrade over a song, love the genre on this though.
I confirm I am hearing what you are hearing.
(you might also try comparing the mp3s)
Curse you guys and your duets. This is the only one that did anything... https://app.suno.ai/song/cfe50baa-732f-478c-aa78-780302ef7dfc
[Verse]
(kyle)
Hello, my love / I miss you so
The way you flip bacon / makes my heart go
[Verse]
(sarah)
That sizzle, that snap / that pop and smell
I love you more / than I could ever tell
[Chorus]
(kyle+sarah)
Breakfast with you / couldn't be better
You're beautiful in / your ugly sweater
Hi, Could anyone Tell me if i can find out what female Voice was used in the Web Version. I found a nice German female Voice playing with Suno and want to reuse it
Unfortunately I don't believe this is possible yet.
LOL I have enough trouble getting a female voice when wanted. Let alone a specific one!
I tricked it, using continue from for new Songs, then cutting it in Garage Band that works for Style and Voice
Even that's not 100%. Last night I was trying to complete a song that was already 3 minutes of a nice female voice and she became a man for the outro but only one time.
yes i had a Male Rap Part in a female Pop Song but it sounded cool
Are full clips not working?
They work but if you want two songs with same Voice you need to Trick
Yo a heads up - i didnt request these songs... they have nothing to do with what i just sent through (playing)
Please report this in the #🪲┃bugs-web channel so a team member will be more likely to see the report.
Please, could anyone tell me why when I continue a song the singer sings words that are totally different from those written? Thank you
Not cool. First off we responded to you in the other place you posted, #💬┃general-chat message you may not like the answer but it was an answer. Secondly the rules are very clear, do not post the same question in multiple places, so take five minutes and reread the #📜┃rules you agreed to. Thanks.
sorry, thank you
You only psoted it twice. Some people post the same question everywhere and then it can be hard to make sure they got an answer...or worse the automod kicks in..
Interesting insight into hallucinations. Right clip is a continuation of Left.
Right is just repeating/hallucinating the text prompt in Left, but the music and vocal style is quite different, almost like a full song reset. The words as content are reasonably distinct in the model.
https://app.suno.ai/song/f8e85430-fbae-4a82-809d-8c22ddf1cb90/ https://app.suno.ai/song/0a93bfcf-054d-4839-8e0c-7267a577a9a2/
Edit: Continued from same source, but same voice and hallucination: https://app.suno.ai/song/d6a55111-4c2f-47c7-9daa-7612d672a68b/
close, sentimental, emotional, spare, heartfelt, sentimental, perfect, close song. Listen and make your own with Suno.
country, indie, spare song. Listen and make your own with Suno.
[ A perfect day, a perfect life, a perfect ] song. Listen and make your own with Suno.
I can’t get a full clip all of sudden everything just keeps saying “getting full clip” but it never produces. It’s been this way all day
Hi all, I am a newbie to Suno and wanna ask a question
What is the best way to specify male or female voice? I am using Style of Music section but sometimes works, sometimes doesn't...
I check other people's work but cant really find good example
Mostly the style of music section is what I use.
If it doesnt work, just create again then?
Basically.
It has been happening to me too. I wrote a song via the bot's DM, but when I tried to press enter on the heart of the last part, it keeps on saying "Getting full clip...", but nothing else happens!
Glad I’m not alone hopefully it’ll be fixed soon still isn’t for me yet tho🤞🏾
I am having the same problem for about a day now. "Getting full clip ..." just gets stuck.
In the Discord, heart button seems to be stuck. It does work in the web version, but I don't think you can use the web to stich together Discord clips. I do see some Discord "getting full clips" that finished, off and on, over the day.
im having some troubles making the bot understand what style I want. When prompted with "describe the type of music", I found that giving too many prompts is typically not helpful either, as it'll easily confuse the bot.
This begs the question, how can I balance the detail of my description and the accepted range of metatags to give?
For example, if I want a rap with very clear pronunciation, which is decently fast, something like em, what can I send besides "rap, fast rap, hard beat" to give more accurate results?
I only now realized you can generate lyrics on custom mode too
having the lyrics box as input box as well wasn't all intuitive to figure out
I've known about the random generation there, just didn't realize you can provide input on the same thing
It might help to use lyrics which are a little over the top in the style you want, in the first clip. And then you can continue the song with your normal lyrics. An example:
https://app.suno.ai/song/56f50163-fc10-467c-a94b-702a39a21f5b/
https://app.suno.ai/song/820d017e-f2ff-4cf1-b431-603180bf05f1/
You can also just use in adjectives, emotions, vibes, in the style prompt. angry, sultry, fiery, just try things.
lyrical rap, fast rap, rhythmic, angry hip hop song. Listen and make your own with Suno.
lyrical rap, fast rap, rhythmic, angry song. Listen and make your own with Suno.
thanks. I'll give that a shot
my lyrics might be too long too? would that really impact the bot a lot
For rap you may want to use longer lyrics in general (if you want it fast) but if it's not finishing the lyrics you give it, then that's too long.
okay, tysm!
Did you try any full clips inside a DM with the bot? Thanks!
The Discord "heart" has been working off and on (if you look in the chat channels and scroll back in time, there are some that finish, it seems quite random) so it's not completely offline, but it should work the same in DM as the public chat.
Anyone I could contact to warn about this problem?
I recently pinged Suno just to double check they were aware it was still a problem. And just in case you didn't realize, the web version of Suno music is actually higher quality at the moment (and is getting most of the traffic I think), Discord is still waiting to catch up and match the versions. So you may prefer to use it anyway, at the current moment.
I prefer the bot version for accessibility purposes. It's more screen reader friendly, so for a blind person, the web app is a bit messed up. Anyway, if the bot doesn't get fixed, I'll try to tab around the web version. Thanks anyway!
Yeah, the Suno web UI should improve for screen readers in later versions, but it is known to be rough right now. (You can't even sign up as a new user without a javascript workaround!)
Actually, I could make a quick Chrome plugin for screen readers on the Suno web site, a few extra bits of text here and theree might go a long way. But will wait until after next major UI change so it doesn't break right away.
What I can also do is to try and get the parts of the clip and glue them in Audacity, lol.
They may not glue together as cleanly, I think when Suno does it, it merges them at a lower level than pure audio, so it's a bit smoother. But it coudl work fine.
Well, tomorrow I will decide what to do: new credits, new options and even another chance at the web version!
If you try the web version, feel free to post the biggest screen reader problems (in channel #🪲┃bugs-web or #💡┃suggestions-web ) and @ me. I don't know what's changing in the next version of Suno or when it rolls out, but when it does, I can try to highlight the lowest hanging fruit (smallest tweaks that help the most) with the Suno folks.
Hey, I'm trying to make guidance for my projects. I want to make sounds like Depeche Mode and New Order. Any suggestions with the prompt or style of music?
Hi Suno, I updated the payment on December 31. I want the credit term to be until January 31
Today all credit will disappear
Hi DSSX, for billing issues, you can email billing@suno.ai which goes to people who can fix the accounts directly (and creates a ticket).
Hi. look DM
I didn't get any direct matches with a few generations, but it's not far off. Used https://sonoteller.ai/bKVbB_qZQcU for hints. Roll those credits, you might get lucky.
Understand what your favourite songs are about. Get accurate and in-depth analysis of songs with SONOTELLER.AI, a music analysis tool powered by artificial intelligence including a comprehensive summary about the song. Analize your song, analyze genres, analyze moods and enhance your musical understanding. Use SONOTELLER to autotag your music fi...
Joanathan. Could I DM you please?
You can DM me and I can help with information or guidance (or non billing/account issues, of course), but Discord mods are just for Discord. The people who receive the emails at billing@suno.ai or support@suno.ai can actually fix your billing problems.
first thing I try is just Google "What type of music is X" -- usually it's a pretty basic, but sometimes that's all you need.
Suno recognizes 'synthpop' (without the hyphen).
'Synth-pop' is some kind of vanilla pop (to my ear). 😫
For Depeche Mode try style words like: ominous, sinister, sparse, pulsating...
For New Order try: melancholy, melodic bass, energetic bedroom...
the only synthesizer I've found in Suno is 'synthpads'.... cold synthpads, gritty synthpads....
"fm synthesis" sometimes worked for me. Though sometimes not exactly as expected: https://app.suno.ai/song/d13bd85d-0efa-4179-8dd4-1e7e412247d4/
fm synthesis, fm synthesis, fm synthesis, fm synthesis, fm synthesis song. Listen and make your own with Suno.
I believe Discord (chirp V1 )hearts should be working now.
Yes! Working for me now.
So, I'm having this issue again where the background instruments start off super clean and clear for the first couple continuations, then they seem to get dirtier and muffled with each continuation until by the time you get to the Outro the instruments sound like they are being played at the bottom of a well filled with mud.
As an example, I've got a country music style song where there's a fiddle in the instruments and it's crystal-clear and beautiful in the beginning. By the time the Outro is reached 4 mins later you can't even TELL it's a fiddle and it sounds like cats wailing mixed with fingernails on a chalkboard
please can anyone tell me I generated a song beat ,music, composition everything is perfect but the voice is like slurry voice kind a distortion in voice means not quality all i can say is not good voice quality like robotic voice feeling how do i fix this
You received a response in #💬┃general-chat where you also posted this. Please take a day to re-read the #📜┃rules you agreed to when joining this Discord. Particularly #3️⃣.
How to use the same music already generated, with others lyrics ?
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
You mean reuse the melody and vocal with a new song?
Exact
Unfortunately that is not something that can be done easily.
The only way to get close to this would be to reuse the first segment of a song and on the continuation use [break] to try to get a clean stop. Then trigger the same metatags that you used in the first part and input the new lyrics. Then once you complete the new song and get the full song, you'll have to remove the part of the old song from the file in a DAW program.
Therre aren't seeds available yet so that's the only way that I know of to do what you want.
I'm trying to get something of a multivocal thing going on...not having much luck.
https://app.suno.ai/song/99ee3607-c7ff-4c74-a90f-f6cbbf9140af
https://app.suno.ai/song/02d83269-640e-4d0f-a80b-00dc608bb00f
https://app.suno.ai/song/0cac026a-d75c-4798-8194-4a1b3535baa3
https://app.suno.ai/song/3f5e3a05-7e5d-4ed2-a740-c00245d1fbe8
https://app.suno.ai/song/15e0fc76-465e-4bb4-a366-84f6c0e84482
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats, duet vocals song. Listen and make your own with Suno.
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats, duet vocals song. Listen and make your own with Suno.
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats, duet vocals song. Listen and make your own with Suno.
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats, duet vocals song. Listen and make your own with Suno.
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats, duet vocals song. Listen and make your own with Suno.
Some magnificant hallucinations though.
If you get a moment before five, I'd love some advice @untold briar, and if you're busy I'll check in again after nine tonight.
Hi! I used Sonoteller for the first time and it's great to get inspiring styles! Too bad I found out I can use it only 3 times. I subscribed the Free version but I can't understand how it works in the API version.
Just generally, I think best to make big prompts changes, cover the range, especially in format tweaks, like try:
TEDDY: Line of lyrics.
TEDDY
Line of Lyrics
(Teddy)
Line of lyrics
Names, try with no intro stuff, straight into lyrics (even if it's not what you want, if it's what's blocking it, want to find out fast.) style prompts to. duet in the style prompt is a wildcard too, sometimes it makes duets from regular prompts, other times does absoutely nothing
(And other duet formats, there are others I saw). [both] instead of [name1, name2] or [girls] could work. Though usually the names and voices are random, it seems to me it's more the presence of a name, in general, indicates "new singer" but less often a specific one. And in the chorus you often DO get both, even if you don't tag it that way.
I'll try without the intro and musical intrumentation, see what happens.
If you need a reference, this is what I'm trying to do: http://digitalcommons.library.umaine.edu/mmb-vp/487
Argh...this one it did it but got the genders wrong, which is usual for me. https://app.suno.ai/song/913f8cb2-4252-46d8-b68e-6b3c310afe91
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats song. Listen and make your own with Suno.
It also seemed to switch halfway through.
It gets genders wrong in any situation, but maybe try generic names too
(That was from before the advice. Am generating with beginning with the Verse now.)
Was originally Teddy and Girls. Tried with Teddy and Betty and Veronica. Now just Betty.
I think about this like tags in bark and voices. A voice in bark is just audio, like literally an audio clip. And sometimes a tag like [giggles] works perfectly with a specific voice, other times it's completely ignored, other times it generates horrible distortion. And it feels random, but it is consistent, for each voice.
And the only thing varying is the audio before the tag in bark. A lot of the less standard Chirp tags that sometimes seem to be perfect and other times do absolutely nothing feel like that, based on the audio (the bark voice) context you are in, they are quite different. But in a particular song/context they could work well and be used muiltiple times.
Have changed Teddy to Robert since I guess it could be considered a female nickname?
This one was odd, but good. It started in the middle of the chorus for some reason and if it hadn't had ladies in the verse I might've tried it any way..
https://app.suno.ai/song/2ca641f4-ca0f-4a1c-b4ac-65d11522883f
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats song. Listen and make your own with Suno.
Tentative success!
https://app.suno.ai/song/b0a075ce-ec6c-460b-83c7-e8332d850e3a
Melodic Disco, Soulful Vocal Disco, catchy hook, funky beats song. Listen and make your own with Suno.
It screwed it up half way through. 😦
Maybe add some style tags related to vocals, or even literally "vocals", to buff up the voices a bit. I have on occasion just spammed ",vocals, vocals, vocals" and it seemed to help lol
My credits renew tonight which means as a min/maxer I'm at a literal 0 for the moment, but can test later
If I do that too much I lose control over the style though.
I have 3800 myself and a rollover on the 29th.
Melodic Disco, Soulful Vocal Disco, catchy hook, male vocal, female vocal song. Listen and make your own with Suno.
lulz
Honestly my only real prompting trick is not having strong preferences, so I can let the style roam, which I relaize isn't practical for most people who are aiming for things
This may have been a bad idea.
Took a minute to catch hold, but it did
[Two people talking, no melody]
[Alice]: Hey, Bob, how was your weekend?
[Bob]: It was great, Alice! I went hiking in the mountains.
[Alice]: That sounds amazing! I've always wanted to do that. How was the weather?
[Bob]: The weather was perfect, not too hot or cold. You should definitely try it sometime.
[Alice]: I might just do that. By the way, did you catch the game on Saturday?
[Bob]: Yeah, I did. It was a close one, but our team pulled through in the end.
[Alice]: That's awesome! I couldn't watch it because I was at a family gathering.
[Bob]: Well, you missed a good one, but there will be more games to watch together.
[Alice]: Absolutely, Bob. We should plan to watch the next one together.
[Bob]: Sounds like a plan, Alice! I'll bring the snacks.```
I was just thinking it'd be fun.
Have you looked into any of the comedy clips?
Also, nice to see you around! 😄
You could also try to start the duet on the second clip, the style won't change that much even with drastic style prompt changes.
Likewise, time has not been my friend lately
spoken word comedy live song. Listen and make your own with Suno.
https://app.suno.ai/song/1fb6fa0e-f6a0-46cf-823e-4dfca0c1c4a2
https://app.suno.ai/song/d4dd3c74-9f13-445a-ba25-40e76ef58540
live comedy spoken word song. Listen and make your own with Suno.
spoken word comedy live song. Listen and make your own with Suno.
Oh my...