#📚┃suno-school
1 messages · Page 74 of 1
Hello everyone! I'm new here and I'd like some advice: absolutely 100% of my music is now accelerating with SUNO, I'm sure the subject isn't new to you, do you have any tips to give me in the prompt so that this doesn't happen? Thanks in advance for your answers 😉
in fact you can separate the voice from your music for free with vocal remover (search on google : separate voice on music), it's pefect for this 🙂
could try to use cover and type in lyrics over the uploaded audio. That’d be my best guess, if it still won’t give u vocals maybe could try singing it and uploading, maybe with vocals present it may allow u to add a generated voice on top of urs? Or could try singing w guitar, getting the stems, and maybe the guitar stem would allow for adding lyrics with a new voice with yours stripped out
Ultimate Vocal Remover, i use it to correct bad words (i haven't try replace function yet)., Yes,a cover with lyrics is most factible. With extend. i haven't be lucky, it's like remixes and i can't adjust it.
Does Suno Scene support other than English language?
It's a good question. It would be interesting a screen capture to see the differents options. I hope it opens to the rest soon.
I’ve randomly gotten a result in Spanish for some reason, no lyrics came up though.
My subscription still has 2500 credits, but I am running out of the Cover Credits. Where? How? More Cover Credits?
Once the free cover credits have been used up, it'll use your normal credits instead
not sure if this is the correct place to put this question but does anyone know if suno can be temped to come up with something like this .. like what genre keywords would be required to come up with something like this .. youtu.be/I9qMm3Zp2g8?t=83
yes, I just made a french song, just type in your prompt something like : french female voice
I would like to generate a few thousands songs via API but i couldn't find anything. Is there any way to automatize the song generation?
No, there is no public API
ok, a pitty, thanks for your support @safe hazel
Thank you, I'll try it
The Cover-Function doesn't work really fine...
It seems the function generates a totally new song without ANY relation to the one that should be covered.
Maybe a Bug?
Hi,
May I get some help? I tried to extend one of my instrumental songs, but somehow it kept on extend only under 20 seconds. I also picked the middle of the song where it has more beats. Are there a specific tags that I can put in the lyric part for it to extend longer?
Extending it in the middle, where the music is still going strong, is a good idea. If you're still only getting 20 second outputs then try to put tags in the lyrics section like [Musical Break] and [Musical Interlude] and in between those tags put some characters that can't be sung, like ******!!!!!!!\\\///////......><><><><
There are no tags that are recognized that make the song longer, but your goal is to give the AI a bunch of unsingable characters to work with so it will continue with the instrumentals.
Thank you. Let me try it now.
Sometimes it is right but most of my covers belong to the original music
I got the totally opposite.
Yesterday it went perfectly
Today it is such a mess and I'm not really happy to be honest
Oh... I didn't try cover today, so you can be right. Maybe this new version could be the problem.
hopefully it gonna be fixed. Spent about 500 credits for absoloutlyi nothing 
Any suggestions on how to add "wordless choir" to the intro of a song? Suno just ignores everything I try to tell it unless I give it actual lyrics.
it works after 5 tries, I guess that's AI. Is there an instruction with all the tags somewhere that I can learn from?
There's a community made wiki (so, not official) that has a bunch if tips on various topics. It's linked here #1151542418991239168 message
What do you want the choir to say though? You can always just put the lyrics as "ooooh oh oh ohoooo" like that. A tag you can try is [Vocalization]. But it doesnt work every time (like most things with suno)
Yeah... I don't really want it to say anything. Just trying to get a "dramatic choir intro"
Zothmire, thank you very much for your help. I really appreciate it.
I've been trying "oohs" and "aahs" but they just don't really come across right.
Well what do you want it to sound like? If you want a choir of voices they'll be making some kind of sounds, what sounds do you want?
I managed to get one result just using [Choral intro]
(dramatic orchestral wordless choir stanzas and shouts) but the rest of the song isn't great, and it's also really inconsistent. I'm just looking for sort of the generic "oh" to the song's rhythm.
extend from that one you got until you have a song you like
but, use the good result you already have
Yeah... side note, any way to "reverse extend" a song? Like, I just generated another one I like, but the beginning just sort of cuts in halfway through a note/beat. I'd really like to be able to say "add 10 seconds to the beginning" or something.
the only way would be the new replace feature
Being able to splice songs together, or at least have Suno reference other songs you've made, would be great.
i agree, hopefully someday
I'll keep tweaking with extending, that's a good thought. Thanks once again!
Is there currently no good (working) way to change Lyrics? recreate or cver are not really usable if you want to keep the sond and only changing the Lyrics
This is the best way, whether it's good or not is open to debate: #📢┃announcements message
Yo! I said that I would come out with some comparison audio for the AT2020 and the LA-320 (Telefunk ELA M Clone) - I will get around to that real soon because I think it's valuable for anyone practicing their high end...but I just wanted to come through with a testimony.
ONE TAKE and I am sold. The LA-320 tames the high end SO WELL! I was haaaaating the results I was getting with my AT2020 - it forced me to manipulate my voice ALOT to try to get it to sound good, but it just wasn't ever what I was going for. I can tell you now, this new mic makes all of the difference. If you have a "sharp" or "crisp" sounding high end and the recording results you are getting sound harsh and make you wanna stab your ear drums, get this mic. Best $320 I've ever spent on anything that isn't a guitar. Highly recommended
Hi all. What do you think are the chances Suno will provide some kind of "project based" song creation? Or at least some "folders" like on pc to organize the songs. I think all users who use the platform a lot have a messy library...
They're interested in the idea, but no idea when it'll be implemented 🤷♂️
ohhh. ok. I think it would be a game changer!!!
I've been getting this "System Error" for 3 weeks now, still nothing
Have you contacted support@suno.com in those 3 weeks?
when i first got it i thought it was some sort of an outage, didnt really use suno that much since then, wanted to do something today but still had the issue so i did send them one today, still looks like more of a widespread issue
I'd be curious what percentage of users are effected, because there ARE users reporting it, but it's such a small drop in the bucket compared to users who are accessing the service without issue -- so it's not like it's a massive outage, but a vaguely worded error message that needs to be investigated in order to be explained.
looks like its been happening since end of 2023 to many different people and those are only the ones that actually did join the discord & reported it
Again, “system error” doesn’t point to one specific issue — it’s a vague error message that can be received for any reason that hasnt had an explicit error message written for it.
It’s as vague as saying “Something went wrong” — once the service gets big enough , there’s bound to be people experiencing issues, but it’s a RIDICULOUSLY reductive view to assume that every one of those instances is caused by the same root cause.
The only way an emerging issue can be confirmed is by direct confirmation from the staff, or multiple realtime reports like when the service goes down — this isn’t the same as either.
Email support@suno.com if you experience this error message or have any follow up questions about why you’d receive it.
Is this the correct place to ask about whether Suno may be improving their voices for music gen?
I've created quite a lot of songs so far but the voices always go completely messed up anytime they get 'extreme' or 'powerful'
everything's going great until the chorus kicks in and the vocals just turn into robots with throat cancer
For what it's worth, the cover feature at times will add intros, least it does for me.
Hmm. Interesting. I could play with it.
Have you tried "gregorian" maybe 4th or 5th in the styles line?
I’d been thinking about it but I didn’t really want that sort of sound. More of the dramatic Two Steps From Hell type of choral vocals.
I’m also almost out of credits so I’m running out of room for experimentation. ^_^’’
Often we're reaching for the nuances enveloping the style than the style itself. GREGORIAN is know for more of a mantra, seemingly non-spoken voicing. Placing later in the style might ignore the lower bass lines of gregorian chanting.
On that same note, then try "boys choir" , vocal ensemble, or since suno likes Latin "Schola cantorum"
Schola cantorum= boys church choir
Oh that’s interesting. Ok. I did get one iteration where the chorus basically said gibberish. It was kind of funny. But knowing those specific styles and prompts is super helpful, thank you!
Make up something like "drone vocals", or let's see..... "atmospheric voice"
Suno is generative, so I'd suspect , it's creative that way, and you might get lucky
What genre are you looking for .... churchy, evil, viking, Jacob Collier style?
https://suno.com/song/dc23a841-f8ee-433d-a3de-a72b18ce5c05
With each generation I'm getting a little different, some ethereal... oh... add that style too probably... and some pop sounding... only one of about 8 came across the gregorian chanty style
Ok. Good luck
Who should I talk to about account issues?
You'd need to email either support@suno.com or billing@suno.com
Who likes clarity ;p
#1Anti-pop#177606669@Y_R+P song. Listen and make your own with Suno.
Get recked
How in Replace Section I just want to change the content, but keep the old voice and style of music?
I think the Replace feature is having issues
whenever i log into suno with discord it just loads forever
it worked with google
though i dont have any credits even tho its the next day
plus i dont have apple, or microsoft
Hi, not sure where to put this as new to Discord. I went into the new iOS app, it made me make a name. I mistyped, went in deleted the new "profile." Now all my stuff from credits to songs / lists are deleted. No clue what to do as the help page doesn't address this and I'm not seeing anything that can connect me to help.
@gusty pendant this is what I got, extending from https://suno.com/song/2463c2ef-c7af-4f98-b104-2184af80051f
Some MF DOOM/Gorrilaz type beat if I may say so myself.
Especially the non-(1) one.
yeah like that one better, more of an original file in there, very interesting where you can get by using a simple few second sound file of stuff I would not even consider using as a base for music lol
I know we can't have an exact reproduction of an artist's voice, but I'm trying to make something close to Michael Jackson, the problem is that Suno seems to ignore almost all of the "(hee hee!)", "(hoo hoo!)" and "(aow!)" I give it.
I put "[Michael Jackson]" in the metatags and the voice is close enough, although not perfect, but I can't get him to do the signature MJ sounds (and honestly the tools to do AI covers really aren't that good, they don't detect the lyrics correctly when it's sung too fast)
has anyone managed to do something close with Suno ? Not wanting to be an accurate stuff, but close enough so I can call it "MJ style"
@dreamy hare
Support@suno.com — at the moment, the email support team is the official first-contact for these types of issues.
Thanks!
i would like to have a litle more control,
on what i try to create.
like i could command in the lyrics parts as [female, shouting, happy] or [male, growling, agressiv]
even for creating a Duett!
https://youtu.be/EZxbsi816sU?si=L_8ftzb6LMugMeBx Thoughts on my tags for this song? I'm trying to go for something similar to the beat used.
Clicky Cowbells, East Coast Hip Hop, Mysterious, Reverberating Guitars Riffs, Chilling, Post-Grunge, Hardcore Hip Hop
Provided to YouTube by Zebra Traffic
Shakedown (Interlude) · Cappo
Spaz the World
℗ Zebra Traffic
Released on: 2009-03-09
Composer: P. Spence
Composer: P. Adey
Composer: P. Littlefair
Lyricist: P. Spence
Lyricist: P. Adey
Lyricist: P. Littlefair
Auto-generated by YouTube.
Just try to copy Scatman John
Did you do something like "Hello (low-low-low)"?
Hi guys, I'm not sure if this is the right channel to ask this question but is there any API discussions here?
so suno doesn't have an official API i'm taking it?
illshouyu has the gist of it -- https://discord.com/channels/1069381916492562582/1253766525458841610 👈 here's the known-suggestion post asking Suno to make their API available.
thanks for the help 🙏 .
A Suno API would be ridiculously cool, to be fair.
I have so many liked tracks on my suno feed and just trying bulk download them without going one by one haha. Anyone with any tips?
big time
Ooooh, mass downloading is something that could be helpful -- I know you can select multiple songs at a time for grouped actions, but that just seems limited to "Add to Playlist" and "Delete". 😦
I've been going through my library downloading everything as WAV so I feel the pain involved with this (having to also wait a few seconds for each download)
Yeah it would be a solid feature, especially with accounts that generate thousands of songs.
You need access to the web to use the app
I agree, Hopefully they can add a feature to do a per page download
You can still download the songs directly onto your device to listen to when you don't have internet access, but you're storing them onto the device, not specifically in the app.
I'm very interested in API..
#💡┃suggestions-web is the best place to ask for an API, since the devs frequent that channel to look for ideas. I don't know if #1252298754787377225 is updated anymore, but https://discord.com/channels/1069381916492562582/1253766525458841610 has been posted a while ago.
to a degree, unless you liked the track and want all your liked content downloaded to process elsewhere
I'll just have to wait until they provide an official API with documentation. They seem to be on top of things. Hopefully sooner than later.
not sure what the difference is there.
well when you're a professional and time is money, I think that rule is obsolete
agreed haha
In part , the Michal Smith case may weigh further on AI music creation apps in maintaining a gatekeeper position, and at the lease, not put focus on developing a mass download system but keep focus on expanding the skills for song creation.
As is said, the few spoil it for the many.
Hey everyone! I'm trying to use the replace section feature and I tried over and over for changing one word, "revealing" and no matter what variations I try I can't get that one word to change! It's too muffled and doesn't articulate the "vealing" part.
Are there any fine tuning things I can do in the replace section to make this work? Love the whole song except that one word which is the title of the song lol. So kinda important.
I tried re - veal - ing and all sorts of stuff like that but no change. It changes the parts around it a bit but not that.
Hello everyone! 🙂 I'm not sure I'm in the right channel, I don't know where to post this, sorry. Are there screen reader users on PC here please? If so have you found a way to change the version to v3, and also to use the replace section feature please? The replace section button sends me to the home page, and the dropdown for selecting the model doesn't work. I'm using NVDA, I tried with Narrator as well. Thanks very much!
Where is the get whole song option now? I have part 3 of a song and it won't appear along with the other options.
You may need to wait until the song finishes processing?
I'm using NVDA right now, I've been trying to optimize 3.5 so haven't needed to change the model but will check this for you.
I can play the whole two minute extension so that means its done right?
Usually it's when the "extend" button appears beside the 👍 / 👎 buttons -- that's odd if "Get Whole Song" isn't appearing on an extension, if the issue persists after restarting your browser, you could share or DM the song and I could see if anything jumps out?
Okay, so use NVDA browse mode, NVDA+space, and shift tab/tab to where it says model dropdown section. hit enter on that, exit out of browse mode with NVDA+space again and go to the bottom of the page. The modal dialog should auto-focus but sadly it doesn't, so select the model at the bottom of the page. I don't know if this persists across sessions.
Haven't been online much lately and wake up to a whole lot of new exciting features.
- The "crop" feature is nice, but what would really be need is a reverse crop. I like the songs, but there are parts I want to omit.
- The "replace" feature really does not work. When I mark a section it reacts properly to the lyrics, but when I hit replace, all the generated sections include rehashing of other lyric sections, as well as changing later verses which is not at all what I want it to do.
I tried the "keep the same length" button, but that only makes it worse, now it generates 1-2 minute replacements when all I want to do is a retry on a verse that was not sung correctly.
Agree: Trim works better than crop. With a trim, you can "trim" the beginning or end, same as cropping but can also "take out" a chorus, an inside piece, a repeated second verse and so on. I guess there might be an issue with splicing two end pieces together and re-blend the pieces which would be nice-- whereas in a DAW we try to get an exact splice as much possible. That's double the effort to trim out and blend-splice. I'm assuming crop-and-keep was easier?
"Trim'n'blend. It even sounds cool.
Wow thank you so much, I'll try this right now, very kind!
Depends on your ultimate goal, deception is one thing creating quality is another. Suno is great but it usually takes me several prompt generations to achieve usable material from trial and error. That adds up. Sorting tracks is another task.
Suno has changed the game in many ways though. Massive props to the people behind the scenes
Hello, im not being able to subscribe anymore.. all my targets are being rejected, can anyone from the staff help me?
While it's not impossible that a staff member would see this and follow up, the best course of action is to email billing@suno.com directly 💪
3 weeks? seriously?? I "just" have it since 7 days now and sent an email to support@suno.com - and still no response from them...
The community support lead literally contacted you directly yesterday. I know you're upset about having needed to wait, but the other user stated they didn't reach out for help, just that they'd been experiencing that issue, and that it didn't resolve itself.
#📚┃suno-school message If you haven't tested another device or account yet, that could help be proactive about the matter
Yepp, that is what I said yesterday to my girlfriend after I tried on her macbook (with the result: system error...). A minority joined discord and a minority of that minority is reporting the issue. What is very annoying is, that SUNO-support does not respond at all - not directly to the people that sent an email nor in their website (I was contacted by a team member here, but only to ask me when I sent that email...), that they are working on that problem.
Okay, if your girlfriend is able to use Suno normally on her laptop, but you receive an error on your account, I think we've finally isolated the issue!
"Get Whole Song" Should appear under the "create" drop down menu. It should be second on the list. The list should be "Extend" then "Get Whole Song" then "Reuse Prompt" then "Cover song" then "Get stems" and finally "Song radio"
😉 I tried with my account. she does not have one. but yes. I would say the exact same: account issue.
Hmm. That letter case method for prompting works really well. Got a nice melody with a crap quality voice. I wonder how to use Udio's upload feature to make a better version that can give me the same melody with better sound quality and especially better voice.
would it be possible, as proposed by MOISE AI that I also use, to have the chords of the songs (and to be able to download them) and the overall tonality of the piece. in the same vain request could we have the possibility of choosing the tonality for creating a song
Would love to simply edit a song similar to changing the song details. Just change out a word, etc and hit save. simple.
guys
what else is wrong now with the platform?
it's stuck on loading forever on my side
yeah
it doesn't care neither the browser
chrome , firefox, nothing changes
let's try erasing the cache :|
no but even from mobile it's stuck on loading
wtf
it doesn't care
even changing ip
...
Web host was down for me too for like 8 min but it's back up for me now
I think i need help. I got a problem with extend. It keep skipping the verses. I erased everything, leaving only the verse I wanted. Well, what did he do? He ignored it and sang something I'd deleted.
I began an advanced prompting tutorial for creating a modular synthesizer in suno - if you are interested here is the workflow:
#1296601261944078396 message
How come we can replace sections on some songs and not others? Is it the length of the song that prevents us from using the feature?
4 minutes, yeah
Am I imagining things or have you been in the LANDR chat also? Your profile pic looks familiar. Is the Suno team working on increasing the length for the replacements? I have a song I’m in love with that is just over the 4 minute mark and there seems to be no way to get the full song even close to how it is now when chopping and re-adding sections.
It’s just one word that got muted between the bridge and outro that I didn’t notice till after the song was completed, yet going back to the original stems doesn’t work either when regenerating extensions.



Theoretically, you could;
- Use a DAW or https://www.bandlab.com/studio to clip out a section of the song to upload into Suno, 6-120s for paid users
- Use Replace Section feature to make the adjustments within the uploaded clip
- Download the modified clip & paste it back into the DAW/web studio
The cloud platform where musicians and fans create music, collaborate and engage with each other across the globe.
As long as you don't make any adjustments to the beginning or ending of the uploaded clip, in theory, it should stitch together seamlessly -- if that makes sense? Lemme know if you want me to throw together a quick demo video 💪
Anyone know the bitrate on the wav files Suno gives us?
Unless I missed something in that thread, its for MP3 only
any prompting help to get it to make a song in a specific key? using suno mainly for vocals
when i choose extra song mode, if i change the lyrics, it will be error no sound or only few seconds
is there a metatag to force the melody from part of the song?
I have a chorus that changes at the end of the song, so I put [Chorus] (I tried putting [Chorus 1] for the regular one and [Chorus 2] for the last), but it sings it with the melody from the Verses instead, I would like something to insist so it doesn't ignore the [Chorus] tag. Same for the [Bridge], it takes the melody of the Verses, when I'd like it to have a new melody since it's a bridge and not a new Verse.
Hello. I have one small question on you guys. I work as as firefighter and we are making videos from our responses and action. The only reason is to informate people about our work and activities on our FB and YT. For this purpose we use Artlist, but with more and more videos is almost impossible to find that much good songs. So I am doing some research if we can use free song from Suno for this purpose. But I can't find right answer. Do someone know?
I think you can generate free songs and use them with attribution to Suno, if it's not a commercial use.
Or, if 10$ a month is not too much, you can take the Pro Plan, generate songs and use them as you want.
For some reason 70 credits just disappeared, I had 120, now only 50 left.😡
did your subscription end?
No I am on free, but when I started I got friends to join and had more than 500 credits. Been using it sparingly.
You should probably ask support@suno.com then.
Thx, will send email
hey guys ,, any one know ay thing about suno no accepting payment right now i tried to buy more credits, i do this every week before renew , but just say error check card , iv tried every thing including loging in and out to useing diffrent cards , useing friends cards nothing is working
Please contact billing@suno.com
already did that.. waiting........ waiting .... trying to be patient....
How long has it been?
Ah, I see. Well, considering they are likely based in New York, it isn't even business hours yet
It is approx. 0718 EDT, usual business hours starts at 0900
ok cool thanks for that..
I say 24-72 business hours, is the more expected turn around for a response. But we hope the sooner than the later.
Did you check with the bank issuing your card to ask why it was denied, if denied, if suno payment "declined"? That can get you a step closer. If Suno says an error with card, especially after three different, then sure, talk to Suno.
yuh i check the cards fine used it on diffrent purches online to check , its not even try to take payment usally i get a notification if it tried
its not really a big deal.... but i need my suno fix .. lol half joking .
Do extensions now pull from the initial starting point's lyrics?
It's spitting out shit it didn't used to do before with extensions. As if other lyrics are now inserted into the extend's context or something.
Like stuff that did generate before, but I extend before that actuall started. I always felt and assumed that would leave it out of context. But now that I'm playing with ad-lib and spoken word, stuff just comes back, without putting it in my next extension prompt
Hi all! Sometimes I want communicate with some authors. I leave my contacts in my descriptions, only if somebody want communicate with me. But I think it would be nice if suno had the opportunity for authors to communicate with each other
From my experience, suno has always accumulated info with each extension, I suppose as it's not so much splicing an extension on and leaving the previous as the root or default to splice onto, but, in a way, pasting the whole song back on itself but now stretched to include an extended version. Thus we get the dreaded "2nd verse curse" where it may repeat verses. So, it reassesses and in making a new extension, fills in, adding previous verses, lyrics or words before, within, or at the end of the extension .
Always a total guess from my experiences and reading chat.
Also been having this issue for the last 4-ish days but no reply from support yet lol. Multiple cards declined
What's up with covers not working? I have a song with vocals. The lyrics are stupid, and a-rythmic. Unfortunately, its melody is also EPIC and I want to keep both the instrumentals and the vocal's melody, but replace the lyrics. So, I instructed ChatGPT to make a new song with an identical rythm pattern and also only having rhymes where the original had rhymes. So, I take this and run it through the cover function. And what does Suno do? It makes a carbon-copy of the song I'm trying to cover. With the OLD lyrics intact... Got any tips?
where do I add some suggestions for improvements?
#💡┃suggestions-web for general improvement ideas
#📱┃suggestions-ios for suggestions regarding the iOS app
#🚧┃suggestions-layout for the UI setup of the website
Ok... Sorry for not rtfm'ing before I came here. Will check notes next time.
Not a problem, man -- it's definitely a disappointment to use a cover & have it return the exact same song, sometimes with only the slightest of variations.
I wasn't meaning to throw you under the bus, just showing that you're by no means alone, but if enough people downvote the bad results, hopefully it'll blow us away when it comes out of Beta 😄
One cover? What about 30? 😅
But I get it now.
Thanks btw.
Is it normal for the official support from Suno to give no response for 11 days (except the generic answer)? #📚┃suno-school message Still can't use my account. Was about to finally subscribe in Oct, but well, I guess Nov it is...
[Covers Feedback]
I wish I could set the level of creativity when creating a new cover. Some of them look extremely similar to the original. At least when I'm trying to change a Trance Song to a Metal or Rock Song.
I am now a premium subscriber of Suno and I am still not allowed to make more than two songs at a time, how do I make ten?
I am a premium subscriber of Suno and I can't download wav files. Do I just send in email?
Have you sent a follow up email to their first response?
Also, not sure what context was provided, but I know that giving as much information to the support team as you can usually aids the process -- "help them help you" sorta thing.
https://sunomusic.typeform.com/to/HTrCNnQL They've actually set up a form specifically for feedback -- if you feel strongly about the covers feature for any reason, this is the best place to get those ideas directly in front of the most relevant eyes
Turn data collection into an experience with Typeform. Create beautiful online forms, surveys, quizzes, and so much more. Try it for FREE.
There might be a slight misunderstanding -- all tiers of Suno create 2 songs per "Create", but paid members can click the "Create" button 10 times in a row without needing to wait, letting them generate up to 20 songs at once -- free members can only click "Create" twice in a row before they need to wait, only allowing them 4 songs at once.
Yeah - did send a follow up mail after 7 days (so 4 days ago). I put them all details from the past within the mail as well as the current issue. I was just asking here since maybe there is a flood of mails (who knows?) and the support can't keep up.
Emailing support@suno.com can't hurt just to get the ball rolling, as the help article just suggests that you need to be a paid member -- https://help.suno.com/en/articles/2479873
I'm pretty sure @hollow hare has been suggesting a browser swap to resolve it -- if you're on Chrome, try Firefox, or the inverse otherwise (chromium & non-chromium browsers)
So I tried a few different browsers and that didn't resolve it.
I've sent in an email so I'll wait for support
Gotcha, I'm sure you've run the gamut of testing at this point, so I won't put you through it all again -- provided there's no error messages on https://www.suno.com/account, is it ALL songs you're prevented from getting .wavs for?
Have you tried Firefox? Most browsers are chromium-based and people that encountered that same issue told me that FF did not have that issue
No I've been making songs for about a 1 week able to download wav and now it's not allowing wav downloads as of yesterday.
I can still download wav files from the previous songs. I did have a bunch of network errors and I've cleared things on my browser, tried every other browser.
yesterday I kept getting network and upload errors. As of the update I can no longer get wav files.
everything else on pro works correctly
Oh hell -- an actual puzzle 👀
If you've shared all of that detailing with the support team, I'm hoping it should be a cinch for them to isolate what the issue is -- fingers crossed they get back to you soon 😄
There's never response from suno's team.. im not being able to subscribe bc my cards keeps declined.. i have sent an email and still no response, omg
How long has it been and did you email billing@suno.com specifically?
3 days ago
yeah, I know, and I've sent them a feedback yesterday. Too many questions there. Then I came here.
That's not exactly outside of an industry-standard turnaround...
im losing currently losing millions for this delay
Hmm, depending on the time you sent, that could be 1–2 business days. It can usually take approx up to 5 business days to get a response IIRC, but often sooner than that
How are you exactly losing millions?
I've run it through a gamut of options to try to circumvent the "repeating nesrly the same/the same covers" and so far nothing consistent, other than just typing "opera" in the styles section. But as opera (alone) is rarely my intended genre, then ... no options found.
Other techniques might NOT get a near repeat of the same song, but go off on wild tangents not even close to the original. So no go.
Gotcha -- #💡┃suggestions-web might be the next place place after the typeform script, since the devs frequent that channel for feedback and ideas 💪 🙏
So far, it's just click a dozen times and hope to find one that veers a bit.... still get some nice hallucinated songs though.
millions of milliseconds wasted everyday
Oh wow -- the option is just straight up missing.
Yup
It's showing as "uploaded" and also "part 3" so I wonder if the UI is confused by both of those labels
And there's nothing jumping out at you in https://www.suno.com/account, maybe asking you to update payment info?
nope paid up. it could be that label. I'm not sure why it started attaching that label to everything.
My memory might be fuzzy, but I've noticed "Full Song" and "Part 2" stacking for instances when that's the case, whereas I feel like it used to just say "Part 2" and only say "Full Song" on the tracks you explicitly stitched together-- could be wrong, but it seems like there was a minor UI update
Probably coincided with the release of Scenes, if I were programmed to make bets
I thought you were a robot 😒
Does anybody know if "someone" managed to create a list of let's say 80% reliable instruments SUNO knows of? Limiting to MIDI does not really do the job 🙂

I've seen these songs of lists make the viral rounds, but while there ARE instruments that aren't represented at all in the training data, I'm not convinced that getting a specific sound is always a matter of "using the correct terminology" -- Suno very much vibes with your choice of prompts, and we currently have no way of knowing how it parses the prompts you give it, or which it might favour, we're still figuring out what causes it to hallucinate and thus repeat or skip provided content (though overloading and "thoughtless" combination of input often seems to lead to this result)
For the sake of not holding anything back, people used to share https://www.suno.wiki/wp-content/uploads/2024/04/PROMPTS-FOR-AI-DANCE-MUSIC-🙏mista.lewys_.pdf
I have the very same impression. I tried several things for instance with Oud and when stating it, it does not consider it at all, but if I use "exotic" and/or oriental then I suddenly get the Oud without explicitly asking
Thanks for sharing this! I'll have a look
Keep in min that v3.5 will trend toward "generic" without effective use of prompting or an existing track to extend from -- try recording some Oud instrumentals to extend from, or try to solely focus on achieving Oud using v3 (which is more able to adapt to a broad and versatile range of sounds, as noted in the model dropdown), THEN exten with v3.5 and it'll tighten up the sound.
https://websim.ai/@theharber/song-prompter Just sharing again because I find it remarkably helpful in exploring new genres -- I made a tool that lets you enter your preferred genre & it breaks it down into keywords for you to assemble your Suno Style Prompt
Oh, I needed that info. Thanks! 🙂
You're very welcome! 😄
#📚┃suno-school message descriptions of the other suggestions thread as well, from just a bit earlier 🙏
This is gold! I know what I'm looking for now.
100bpm dub all day
If I then extend a song, must I have verse 1 pasted in again or can I start with verse two? The extension already remembers verse 1 and they just put verse 2 after?
Or they don't remember and I need to add the entire song in on the extension?
Extensions use the previous section as context to continue, so sometimes it can be awkward if you extend right fron the middle of someone singing, but if you extend after verse 1 has already been sung, you can just start with verse 2 🙂
Nice, thank you. 😁 👍
Now I had extended it and I see the other two minutes of the song that was supposed to be added to the first verses.
But now they are two separate songs with the same beat.
Any way to fuse part two and one together?
Yup! If you click ... > Create > Get Whole Song on part 2, it'll create the stitched version in your library.
Doesn't cost any credits to do so -- and while it seems to update immediately on the Create page, it seems like the Library page may need to be refreshed in order to show the Full Song
Does anyone know how to get Suno to minimize the reverb and noise once you've gotten past like your 6th extend?
I'm trying to stitch longer songs together but I find they degrade in audio quality after a while
Okay! Yes I've figured out that it's because of "Uploaded" tag and now it's applying it to anything I upload for some odd reason. The ones that don't have that tag I can download normal wav files.
Are you extending from extends (resulting in part 3, 4, 5 ,etc?) or are you using "Get Whole Song" in between?
first part yes
"Armenian guitar" maybe... let suno interpret what they means?
extending form extend
Usually extending off the extends, sometimes off the whole song. Does stitching it together between extends help?
Means? OK. I have too many ESL friends. I'm losing my ability to speak English.
Suno can only "hear" the clip it's directly extending from, so if you're extending from part 3, Suno can't hear what parts 1 & 2 sound like -- so while part 4 will bear resemblance to part 3, but doesn't understand what the rest of the song sounds like to make it "coherent"
THAT
Anyone know why I got this message?
Is HUGE advice
Guys..why my song was cut off after extended?
So if I'm trying to get a melody or genre switch, extend off the extension, and if I'm going for coherent, longer music, compile between each stitch
It's like photocopying a photocopy of a photocopy -- #1256301881966198805 message This is a workflow I wrote up for using this "fault" as a strength, I call it "terminator extensions". because you can use only a single second of an extension so that Suno 1) doesn't know what the earlier sections sound like, and 2) BARELY knows what the current song sounds like -- but you can still stitch everything together after the fact
Makes sense, I sometimes forget I'm workig with an AI and not a studio band that can just riff for me lol
Doesn't cost any credits to Get Whole Song, so I do it any time there's a section of a song I want to "lock in"
Good to know, I'm burning through credits like mad on these longer tracks
Can you elaborate?
That would actually be a really neat social experiment for a group of musicians;
- First group records a 30 second original piece
- Second group is only shown the last 5 seconds of that track and asked to extend from it for 30 seconds
- Repeat until you have a complete song, with only minor sections of understanding.
Arguably, if you have talented enough musicians, the song will cohesively flow together, but the sections will only be tangentially related by a general "feel", and that can be thrown off by any of the sections deciding to do something wild.
Suno does "Story-telling" Instrumental music REALLY well if you have the patience and ear fortitude to resist the wear on your brain, you can get some tracks that REALLY travel
Need to use Extend off extend though
I made 260 song, never happened before
My Favorite is just telling Suno stuff and getting it to fuse sounds together, it's... I got a banjo, flute and piano fused synthesizer once and it was hype
Sometimes Suno bugs out generating a sample; did this kill your workflow or did you just lose a single clip?
"Extend" doesn't always perfectly resolve the song, but you can extend again 💪
It doesn't make music no matter how I try
Are you... Out of credits?
anybody know what prompt should i use to get artcore
because "artcore" doesnt work directly
of course not
Because it's not an explicitly worded error message, it's just going to be speculative help from the community, so I'd get the balling rolling and reach out to support@suno.com because it might not be something that you can fix without asking for help.
Thank you for the mail address I did not find anywhere
so i need extended and put the missing lyrics?
On the subscription page it also links to billing@suno.com -- though I'm not super well versed on when "support" issues become the purview of the billing team.
I hope the matter gets resolved for you shortly 💪 I think the email team is only active on american business hours, so it there might be a delay over the weekend, but we can try to help explore troubleshooting options as a community -- even if it doesn't outright solve the matter, it could be helpful information to provide to the support team when they get back to you 🙂
Yes 💪
Because the AI model is still new, it is not a perfect system, and sometimes will skip verses or act strangely -- the most we can do at the moment to prevent this is to report these dissatisfying songs using ... > Report > Bugs > Doesn't Follow Style (OR) Lyrics (OR) Poor Audio
But until then, we can use multiple extensions to keep the previous sections we enjoy, and "overwrite" the parts that didn't turn out as we hoped
thanks
Hi folks, new to the stream here, I have been attempting to update my profile pic on suno and all that happens is it will hang up with a continual "saving" spin and never upload or update. Any suggestions? Thanks
Yeah -- @naive quail has a picture kicking around showing the "roadmap" of all of the extensions & Get Whole Song segments that lead to one of his recent songs. QUITE the journey, but well worth it 🙂
I completely understand not everyone is willing to take on a whole project or to "write a song" like traditional musicians toil and stress over, but I think that's another amazing facet of Suno -- it's super capable at producing amazing "one-shot generations", but if you REALLY care about how something turns out, Suno provides you with a brand-new set of tools to hyperfocus on what you want -- I'm convinced that uploads and extensions are all you need to chase anything you hear in your mind's ear.
Will give it a shot 🙂
Ahaha I think my first Boss Theme song is 33 minutes long, and it cost like 3000k credits
Worth every second and penny
I'm not aware of any site-wide issues -- have you tried from a private tab, or a different browser? If you're on Chrome, trying with a non-chromium browser like Firefox is usually a good place to start a comparative test
I am on chrome but just triedon firefox with same non saving nor uploading results
Extensions FTW! 
Haha it gave me an electric guitar that was used in the 60s or 70s 🙂
another example
no dice on the other browsers, good suggestion, but it seems buggy and has crashed actually, the size of the file is much less than the limitation,..?
is there a prompt for the website version that will basically put a time limit on? like sometimes I just want a 1:30 long song with only a couple verses and one refrain, but it will stretch it out, fill in a lot of time with repeated sections, long instrumentals. but If i'm writing like a podcast or streaming intro for someone, they don't want 4 minutes or even 3 minutes or even 2 and a half minutes
la la la,..just trying to change a profile pic, nothing to see here
@charred merlin also tried the private browser to no avail, everything good on yall's back end?
Not at the moment -- the most effective way to limit the length is to create an extension shortly before you want the song to end, then only using "end" in the style prompts -- no other lyrics or styles -- while it's not a 100% guaranteed command, when used at the right place in a song structure, Suno understands that you're looking to wrap the song up, and will do so however it fits naturally.
works for now, i'll give it a shot! It's easier when it only adds things after the song because then I can use audacity to add a fade out before the additions. Not so easy in the middle as AI music doesn't clip and blend as well
😅 I'm in the dark same as you, my Corvine Companion -- not Suno staff, just an evangelistic community support member.
Private tab is usually a good starting point, but shouldn't be the only check made -- if your browser doesn't disable extensions in private mode, it's not accessing Suno "with a clean slate".
Ideally, the point of using a private tab or a brand new browser is to access the site using as basic as a setup as possible -- no extensions, or download managers, no VPN, no cached data -- that way if it DOES work, we can slowly add things back to find what causes everything to "break".
When it's an explicitly worded error message, there's a bit more to investigate -- even "Network Error", as vaguely worded as it is, provides a BIT more information that helps the community understand what causes it (in this case, trying to upload a 16mb photo when the limit is 10mb) -- not explicitly worded, but still points us in the direction of it being network related. They HAVE explicitly worded network errors, but this pops up as a catchall otherwise.
When it just says "system error" -- that could literally be ANYTHING related to use of the service. That's why it's so difficult to troubleshoot or isolate.
hehe managed my Jester vs Clown REDUX, rewrote a bunch and added another minute and a ton of fast rap and chop
that took a bunch of working and techniques to get it flowing
but reminded me that there's no substitute for good lyrics
Yes, there was a network error message once the "upload" times out, I even tried having suno create my own pic, which should have been in their perameters, even though my file is 2 mb. soooooo yeah, good old fashioned network error, but i'm able to create songs and download and do my other work just fine, sooooo? Thanks for the help
When people had similar issue (Network error) uploading an song image, they mentioned that resizing (via Suno's UI) helped fixed it. Try seeing if that works
I am experiencing a technical issue with the excluded feature, which does not appear to be functioning correctly. Additionally, there seems to be a limited voice artists. I would like to know how to make changes to this situation. Furthermore, I am encountering difficulties with the love song style, as it does not yield the expected results. This situation is quite frustrating. Here's a sample https://suno.com/song/16745f5e-19a8-4365-8ab1-5f3e0838aa58
piano, flute, violine, love song, sentimental, memories, emotional, song. Listen and make your own with Suno.
You have to adapt to the model's interpretations of styles, the model can't adapt to each individual user's expectations. If "love song" doesn't work the way you think it should, you need to try other prompts that may fit your expectations better.
Can you elaborate on what you're expecting to hear vs what you're hearing? (using specific timestamps helps)
You've requested "piano, flute, love song, sentimental, memories, emotional" -- arguably it has presented a song that satisfies all of these requests, while the violin seems to be absent.
Maybe because they asked for a "violine"
Good afternoon, I generated a song on the free plan and have now upgraded to the paid plan to be able to commercialize it. However, after upgrading, I’d like to know if the song I created on the free plan can be commercialized. I tried recreating it on the Pro plan, but it’s not the same. Could you help me with this?
I'd posit that Suno can understand basic typos, but piano & flute is WAY more represented in training data compared to flute, piano, and violin -- I figure you might need to play with v3 if you wanted those instruments specifically, as v3.5 is going to drift toward the generic result of just flute and piano
You can email Suno & ask for the rights to the song, but you don't gain retroactive commercial rights to the songs made on a free plan 🙏
what about if you created a song with the free account, then upgraded to paid,..and took a song from the free account and then extended it and created a new full song with the paid account, would we then have commercial rights to the song? Also, got the upload to work, can't say why, but now it is working sooooo Abracadabra, thanks harb
The original song is still owned by Suno;
What you're proposing is similar to someone taking one of YOUR songs, extending from it, then just claiming they have 100% ownership of it, and your stake is invalidated. 👀
Really though, if there's a song you made on a free plan, just email support@suno.com, let them know what song you're requesting the rights to -- I've heard a ridiculous number of success stories, and that way you don't need to worry about your Youtube account getting shut down if you try to just monetize the song anyway
Good point, lol I'm kinda unclear about what does stop someone from taking one of my songs, extending it, remixing, whatever, and then clamining they have ownership,..actually? I'll email them broski, good call, much obliged
Kinda the same thing that stops people from just walking up to someone's property and making off with their stuff -- community support, people proactively protecting their work, being cognizant of what you share and where
As well as whatever overt protections there are, like the platforms where people would submit content making them declare they have the rights to do so -- so if you can prove you made something, and someone else uses it, hopefully you'd be able to show a papertrail of ownership & having handled things properly so that Youtube will respect your DMCA takedown notices 💪
It is a pain to deal with. The absolute best case scenario for you would be to make the song, successfully distribute it (plus copyright if allowed by your government), then make it public and/or share links.
Making it public or sharing links, even here, will end up having your song stolen.
By distributing it first, you shutdown the botters and others from distributing the song first and claiming it like it is theirs (a process that takes a long time to dispute with their distributor - if you figure out them in the first place)
The distributors would register your song with YouTube ContentID and other similar programs (though some requires you to pay for the service).
IIRC, Feline Cats had their trending/major hit "Cat" stolen and redistributed like 7 times before their publisher got it fully distributed, and quite a long time sending decease notices to those hosting the content on YT
When creating a song cover, there's an option to change the algorithm between V2/V3/V3.5 - does that make a difference to covers?
I believe the model-selector defaults to v3.5 when making covers?
The example I will give is this: When you publish an article to a scientific journal, you are giving that journal ownership of your article. They own the words and figures. If you want to publish one of your figures in another journal, you can't (without permission). But, if you make some small changes to the figure, like the color scheme or the font, then it is a different figure and you don't need permission to publish that in another journal.
Thanks homies, much obliged for all that jazz
hellooo, i have a question? how to use my real voice for generate a music?
Upload a sample of your voice
ok,
The longer the sample the better (up to 60 seconds for free users, up to 120 seconds for paid users)
The best thing is to record your voice for as long as possible.
but I have to extend my voice and put my text, and generate afterwards?
You can't change the model used for covers. @charred merlin mentioned that it defaults to v3.5 - in fact if you try to change it, it will switch back to v3.5 automatically. Covers seems to use a tweaked version of the v3.5 model. Similarly, extending audio uploads is only possible with v3.5. So I expect they are only implementing new functionality against the most recent model.
Yes, upload the file, extend from that file, and put your lyrics in the lyric box. The upload will hopefully make suno re-create your voice and use it in the extension
thank you 😄
does Exclude work with v3?
I just thought. There may be one or two songs I created on the free service and said, I need to get a subscription. Anyway to see which songs were created on the free service? More just curious than care.
Definitely would have been a v3.0 song.
Exclude should work with all models, because it's just an modification of your input request / the way the training data is weighted.
While image gen and audio gen aren't 1:1 comparable, as an example, I made a picture of a cat playing chess -- using the exact same seed and prompt between generations, I just added the words "White cat" to the negative prompting.
The benefit to this is that I didn't need to specify what kind of cat I wanted in my initial prompt -- the result didn't give me what I wanted, so negative prompting allowed me to make minute adjustments that kept the overall theme of the current generation, while making small variations that nudge the result closer to what I'd originally wanted.
It's not a superpower in itself, but it's remarkably powerful when used critically in your experiments and workflows
If you're asking for rock song and it keeps giving you piano -- you could try to make a super specific prompt that suggests ONLY guitar music, or you could just spam "piano, pianist, grand piano, keyboard" in the negatives and try to "hedge your bets" that way.
Probably safer to handle it that way if you're getting something CLOSE to what you want, since making massive changes in the main style prompt might push you further away from what you'd been working on -- all in the name of lessening piano prevalence in that genre's training data.
But now we have negative prompts 💪
Unfortunately there's no flag or something on the site that would help indicate which songs were made on free plans -- the easiest way would be checking when your account was active and making note what songs were made during that period.
It should do. I am able to select it with other models.
can this do loop tracks? im trying to no success
Right. Good idea. Maybe I can look back when I first charged for pro/premium service.
<@&1082792484700180531> you seem to have broken "Extend" it works as a cover. And it is unclear how a cover works - sometimes it just sings over.
I want to continue the song in the same style (or rather change the words in the song a little). I press Extend - it makes a new song
Is there a somewhat reliable way so the vocals/lyrics don't immediatly start at second 1 of the song?
Adding a few spacer tags like [pause] or aspirations like Mmmmm and ahhhhh to indicate space.
Using [intro] before any lyrics help to throw in an intro before singing.
Small lines/exclamations such as " let's go" , " 1 2 3 4" , "you ready?" Make up your own.
Also, if you subscribe to pro you can do covers. COVERS sometimes add an intro to a song without one. It's iffy, but I've used it for that when it gives it to me.
Use any type of symbols. ! . - % ^ # just to give it something to start in on.
You can also start with an Instrumental with no actual vocals added to the lyrics box, then extend from a point in time near the beginning of that instrumental track and add your vocals to the extension. That can give you a bit more control over the length of the instrumental intro.
Sorry, missed the reply yesterday, I haven’t quiet done all that, when I originally made it I did go back and extended and got something very close and tried patching it in GarageBand but the underlying instrumentals changed a bit and I’m not advanced enough to do the patch, but as you said maybe if I cut it and reupload the audio I could potentially get a new track that is identical for that little bit I need.
If the lyrics keep starting too soon I usually add [intro][instrumental] or [instrumental riff] etc then [verse] and it usually waits a bit before singing the lyrics. Starting with an instrumental also works then extending what you like… it just depends on if you want the melody to follow your lyrics more or be a more distinct change once it starts singing.
How to avoid repetition of some parts of the song? I have numbered verses, but it still quite often repeats them one after each other. Sometimes it repeats chorus with verse. It's making me a bit crazy, since I need to regenerate songs more often or extend from the middle what gives a different and unpredictable result 😦
Is there any way to cut a part of the song in the middle?
do you happen to have two [break] parts right after eachother, even if there is lyrics between them?
Hey sorry to bother, but i paid for Suno for two month and never got to use my 2500 credits for the second month
Now i connect and i see that i just have 50 credits ? How is that possible ?
check you bank or credit card statements, confirm the payment went through, and confirm it wasn't returned to you.
If the payment went through and you didn't get the funds back, contact Stripe with your payment details
if the payment was made less than 5 banking days, it might still be in 'limbo' and being returned
Sometimes the payment statement also say 'reserved', this is not actually payed, but this might still go through as successful payment. The reserved is usually a delay between bank and CC issuer.
Though I've not tested, in theory, the replace feature seems a means of highlighting a section and "replace" it with a single space bar click. I suppose of you did this specifically with lyrics , it might work, provided the lyrics are actually there. Often suno is singing something that is not in the lyrics, least I saw today in one of my songs.
Any ideas why some of the songs created do the second verse twice when this has not be requested?
Can we teach the AI theory? Circle of fifths, arpegio, instrument names? Standard vocal ranges if we say Soprano Alto Barriton and Bass even contra alto so on and so forth. Might make it a stronger tool for developing music without mimicking other things.
If you don't make a track public, but put it in a public playlist, will people be able to see the track in the playlist?
The public toggle only impacts 1) whether the song can be found using the Search feature, and 2) whether the song can be seen on your profile.
Just to make 100%, I'm gonna do a test
Things can change 😅
yup i can see and play it 🙌
Thanks 🤗
Okay cool -- I deleted it just to keep things tidy, but thanks for helping me check that -- hope your night is going well 🙂
How do i create a song for kids how to count, but heavy metal. I write the prompt but theres the phrase "Crash the wear" in the music. I wanna do something more "family frendly"
I don't think we can directly train specific information into Suno, but all of the feedback we give with thumbs up/down & ... > Report > Bugs > ... is used to improve future models -- so if you have songs that don't respect the style instructions, you can report those -- won't give immediate results, but it will help things in the long run if that's what you want to use Suno for
I'm always a fan of handwritten lyrics, but the "Write About" tool in the lyrics box could help figure out something family friendly that would lend itself well to metal 👉 #1270928841614823546 message
If "heavy metal" doesn't give you the right kind of sound, you could add some supporting prompts or thematic descriptors -- I threw together an AI tool to help turn your genre ideas into prompts and metatags 👉 https://websim.ai/@theharber/song-prompter
@void spade try mixing genres as a style choice. I used a prompt called Fugue Metal and got an interesting style that was really entertaining. Also tried prompts of Symphonic Metal, Djent, Techno Breakdown and got one of my favorite songs to date. Not sure it followed the prompts but it turned out awsome.
How to change the producer tag so it your prompts can sound different each time?
for example the "producer" name on my creations called MysteriousJam349
When Suno is given a prompt that doesn't match anything in the training data, it likely treats it like "prompt decoration", or "prompt noise" -- it doesn't generate any style in particular, but just adds some random variations to whatever song would have been generated based on the prompts it DID understand.
It might read your producer tag as lyrics, but unfortunately Suno doesn't know who "MysteriousJam349" is in order to generate music in that style. Here's a prompt experiment playlist showing that Suno doesn't recognize well-known artists either, so it's not a personal matter. #1256301881966198805 message > https://suno.com/playlist/0e2e154d-ef54-4125-a610-599e0f572562
Is there any ways/tricks to writing the prompt so you have more control over the outcome of the song?
there's a problem when we are seeking for our song with the filter. It doesn't give all the results. There's some song, i got to find them by my self.
I was only able to find the music because I had it in my laptop and typed in the full name. I was given the other tunes but not what I had to write specifically, even though they are the same titles at the beginning (all start with the first 3 words).
The server breaks, it can't handle even two songs generated. The editor throws errors while generating text (pasted text). Multiple. Replacing sections generates broken fragments, torn.!! Premier Plan.
Dear Suno Team,
I hope this message finds you well.
I am writing to bring your attention to an issue I've encountered with Suno's pronunciation of certain Arabic sounds. While using the platform, I noticed that Suno struggles to accurately pronounce specific Arabic phonetic sounds, particularly those represented by the following symbols:
ʒ (ج)
ʔ, q (ق)
t͡ʃ (چ or ك in some dialects)
θ (ث)
ð (ذ)
dˤ (ض)
sˤ (ص)
ðˤ (ظ)
These phonetic representations are essential for correct pronunciation in both Modern Standard Arabic and several dialects. However, Suno currently mispronounces these sounds, leading to incorrect or unclear speech output. For instance, the sound "t͡ʃ" (used in words with "چ" or "ك" in some dialects) is not pronounced correctly, often sounding like a hamza instead.
I would greatly appreciate it if you could address these pronunciation issues in future updates to enhance Suno's support for Arabic phonetic sounds. Accurate pronunciation is crucial for users who rely on Suno for communication, especially in Arabic and other languages that include these phonetic elements.
Thank you for your attention to this matter, and I look forward to seeing improvements in future releases.
Best regards,
Hi, Is there any possibility (mb in future) to show seed numbers and have random/seed number switch?
-+++-+- +s+s s+s++--a-a-a-a-a-
Who had also a miserable sound the last 2 days? Every generation was miserable sound quality. Probably most people from Europe ?
Can you tell Suno to do a producer tag for you? I tried [Producer tag: "my name"]
But it didn't work
I've had like 100 or so credits wasted recently with song style prompts being way off, or cover not working at all, etc. Sound quality isn't a problem here but it's not following structure or style well at all. Have had to report so many
im sorry but when is suno gonna update for thr free users???? its not fair the the paid uses are getting everything, stop doing this please yous haven't even released a good feature for the free uses, come on guys please udio is getting better then yous its making me stop suno, yous released covers, Instrumental and Vocal Stems, Exclude Styles, Crop Songs and Replace Section why can't we have any good new features for the free users???? we only get stingy homepage updates like why should i even bother with suno i should just move on to udio because the free plan is MUCH MORE BETTER With heaps more features for all users it feels like suno only cares about their paid uses and not everyone that just sucks in so many levels so please give suno a good update if not what's the point if you guys only care about the paid users??? aaand not everyone?
are you serious or is this a satiric post ?
In suno, is it at all possible to correct lyrics from a finished song so that the same beat remains
ifs my 2nd time saying it ofc im serious, i'm starting to think to use Udio from now on instead of Suno because Suno is lacking with free features
You're free to choose whichever service you prefer to use. To my knowledge though, Udio's free plan also does not offer: stems, inpainting (replace section) or exclude styles. They also do not allow audio uploads on the free plan.
Historically, new features for Suno have been for paying subscribers only, and at some point in future they may become available for free users (no guarantees that this will be the case for all features though.)
We understand you're passionate about the features offered on the free plan and appreciate your input. However, paid subscriptions are essential to fund improvements and new releases, benefiting the entire community. Sharing complaints repeatedly and posting competitor screenshots doesn't contribute to the discussion and goes against the server rules. Please help us keep the chat productive and on-topic, so we can focus on shipping the best features for everyone. @cerulean merlin
first of all you had no reason to delete my message, 2nd of all i said i meant no harm, 3rd of all im am just trying to do is say please suno keep up. i had EVERY right to say that 4th if you delete this i will be reporting you and screen-recording everything to prove your doing this for no reason, i have every right of freedom of speech, 5th if yous really understand, why delete my message? 6th nothing if offending me your just taking it all the wrong way suno is leaking and i'm giving them feedback how is that wrong? 7th if i said this ain a wrong channel i am daarle sorry and you should have only just told me wrong channel send your feedback in {whatever channel it is} 8th i just read the rules over and over and over and even copied and pasted the rules and asked ai if i did anything wrong with excialy what i said and i did nothing wrong, please understand i DO NOT mean any harm i was just giving feedback and what i think i have every right of freedom of speech
Your feedback is welcome, but please remember to follow the server guidelines. Messages were removed because they were disruptive, not because of your opinion.
Freedom of speech protects you from government censorship, but private platforms like this Discord have their own rules that everyone must follow. In Suno's Discord, our focus is on maintaining a respectful and productive environment for all.
Which is why moderation is necessary.
and what was so disruptive? that it had to be deleted??
Your message was disruptive because it contained repeated complaints and comparisons with a competitor, which shifted the conversation off-topic. It wasn’t deleted because of your opinion, but to keep the discussion focused.
it was not repeated complaints you are so off
it was feedback asking suno to please keep up like udio
I understand you don’t feel it was repeated, but the message focused heavily on comparing Suno to competitors, which took the conversation off-topic. That’s why it was removed.
ofc yous just think of money money moeny
Suno and Udio are different platforms with unique goals and features. Suno focuses on specific innovations in AI music generation, and while Udio may have its own approach, Suno's team prioritizes a vision that aligns with its strengths.
so me asking suno to have more features like udio is so disruptive it had to be deleted?
Yes, because the conversation should stay focused on Suno, not on asking for features from other platforms like Udio. Let’s keep the discussion about Suno’s development and feedback within the community guidelines.
well that doesn't make any sense at all you're just making a reason up i had every single right to say what i said i even said i meant no harm and still you even said nothing about that, thats proof yous dont care, this is what i mean yous dont care about the users yous care about money money money i have the message saved that u deleted and im gonna make an email report about this this is just doing my head in, you have no idea what you're doing and soon enough you will
I’ve already explained why your message was removed—it wasn’t about your intention or the fact that you meant no harm. The focus on Udio and accusations against Suno disrupted the chat and went against our community guidelines. Continuing with accusations and threats only further derails the discussion. This conversation is closed, and any further disruptive behavior will be addressed accordingly. It's time to move on and keep the community focused on constructive topics.
im not reading that because it seems you don't read mine for now on im not listening to you say whatever you need to say i wont read it i dont care your doing my head in please just stop and stop making it worse as i said you have no ida whaat your doing
You've been timed out for 1 day due to repeated violations of the community guidelines, including disruptive behavior and refusal to engage constructively. Please take this time to review the rules. Further violations after the timeout will result in more serious actions. Let’s maintain a respectful environment for everyone when you return.
helloo I'm trying to edit just a part of lyrics using Cover Song feature but they keep on producing songs with the lyrics unedited. am I doing it wrong??
I'm on Android mobile
Wouldn't it be better to use the "Replace section" feature for this, or is this unavailable for you
oh yes prolly that is not on mobile so I could try PC right??
nae, I can't remember if it's still currently only a "beta" feature or not. But if it's not available for you, I'd recommend using the extend feature instead, and set the time to a bit before the section where it's wrong
It is in use, but it makes a new melody in the background, I am also interested in how to correct the words so that the original sounds
Replace Section is PC only at the moment. I guess you're on a paid plan since you are using covers, so you should also have access to replace section
wait for reals? replace only works on pc, and covers aint open for all yet? time to pull out the phone
😮 neato
Pretty sure anything involving the waveform editor (i.e. Crop, Replace Section) only has PC interface atm
Crop's so useful
I was sure I had seen at least the crop feature available on my phone, but it appears I was completely wrong :x
oh wait I got it. I'm able to open Suno website in PC mode on my mobile Chrome👀👀 this could work I guess. though it looks so tiny that it's hard to edit on my mobile😂
I think I'll try Replace Section feature on PC. Thank you very much for your kind help guys🙏
yes editing just tiny bit of the lyrics..I wish that feature would be available too someday.
Bro there are not able to develop this with free users, but with people who pay. Just pay if you want more features.
I'm not sure it recognizes "producer tag" as an style of music to create, let alone something that it would consistently reproduce across songs -- but you could likely use Suno to sing your name or DJ handle, then clip that out & use it as an audio upload in the beginning of the songs you like -- working on a song that you might want to turn into a full track? download it, add your bespoke producer tag and reupload the track to continue working on 💪
That might have been me causing a mote of confusion -- someone asked about it recently, saying they primarily used Suno from their mobile -- I suggested that they might be able to trick their mobile browser into it by using landscape mode and requesting desktop view, but the UI/UX is largely designed for a desktop experience.
Awesome, thanks for the idea!
You're very welcome! After typing it out, it sounds a lot easier than initially expected -- https://www.bandlab.com/studio to clip out the section you want to use as a tag, and to overlay the tag over your upcoming songs (https://www.bandlab.com/splitter if you need to strip any instruments to just use the vocals)
I feel like the trickiest part would just be getting that first generation where it says your producer handle in a cool enough way that you'd want it used over & over again. Best of luck, and make sure to ping me if you end up finding something that you're excited about 🙂
Is there a way to limit song's length, like, to 15 or 20 sec. exactly?
My Fold 5 is considered a tablet by apps. But I think that puts me at the bottom of the barrel, under everything else.
I probably should begin all my comments with, "In my sole opinion"--- so, In my sole opinion:
Suno creates a song from what you provide, within the outer limits. So you can't direct a song to fit within a certain amounts of time. Trial and error and luck might get it, but it is likely not repeatable.
Even where a song is limited to 2 minutes or 4 minutes, suno ends the song when it has allocated for all prompts , tags and space it feels fit. Sometimes it's 01:58, or 3:12, and occasionally exactly 02:00 or a really close 03:59.
So if suno had a new button that said "00:15 length" , I don't think it could adhere to that parameter as of yet.
You can try prompting for "advert", "advertising jingle", "jingle" , "moto" , "slogan" and other forms which are associated with short quips. When I have time, I'll try as well.
#📚┃suno-school message
Basically, after 15s, create an extension point and only use the style prompt "end", or something similar -- no other lyrics or style prompts.
#1256301881966198805 message I took a 2m rap and turned into into a 30s clip because it was funnier that way using that exact method (I initially shared the regenerated version, not the original where you can see the transition with extensions)
Thanks, I found this before asking here but it dosn't seem clear to me how to do it, exactly.
I guess shall need to use music editor to cut
Nope, all within Suno.com 🙂
If you click "extend" on any of your songs (or ... > Create > Extend), you can either add to the end of your song, or choose an earlier point in the song to continue from.
And silly me to forget now we have crop, if you can write it so that it appears to end around the time you like, by putting a bridge or instrumental or something so the previous section has an ending feel, you can now crop it. Yes, thanks @charred merlin , I forgot about that. Perhaps even "replace" the area just after your mark and get a good empty second or two, then crop that. Alot of workarounds, but a direct means, not yet..... in my sole , un-expert opinion.
I'm going to tidy it up later and note what the steps are, but this shows my workflow for ending a song more-or-less where I want.
That's a good idea, regarding the jingle/advert prompting -- I think I'm going to dust off the Prompt Experiment Chemistry Set, and of course I'll give you credit for proposing the idea & the initial prompts 
Well. It's possible, just probably not repeatable. Lol out]https://suno.com/song/fa206715-4480-41c9-83ab-5c9be04b0d21
How do we use the scenes addition? I can not find it in the create menu (I am subscribed)
It's IOS only
should it be there?
I knew it wouldnt work on android browser but its not even on the windows browser?
It's currently only available on the iOS app.
Hey. Good for me. Turned my phone (fold 5) sideways and got the new features. Yay! So my "tablet" went up in the list. 😁
well thats a kick in the lower parts to us non ios subscribers 😦
#📢┃announcements message They've left a feedback form for that particular feature on the announcements post if you wanted to take a proactive approach to the matter & let them know 💪
That said, they're really pushing to get Android up and running -- the engineering lead is making personal broadcasts on his social media asking for talented android devs to join the team, so it's only a matter of time 😄
I would guess they know its not good to provide features to only one group of subscribers and will get scenesadded to atleast the browser fast... will give it a few days before complaining in any official stance lol.
just annoying as an image i created for another song came out so well i wanted to see the song it would generate.
Gimme a sec 😉
this is the song i generated the image for:
https://www.youtube.com/watch?v=Y4sSKTB4-tQ
Was trying something new. This is the Holy Engineers of the Ignavus singing a hymm about the reactor of the Ignavus.
I create AI music, AI art, AI text. The goal of all I generate is to make people smile.... thats it. Its not a "hey look at me, aint I awesome", its not a "here this song is amazing", its purely "I hope this makes you grin"
I h...
You get a field that asks Describe the scene when you upload the picture
Holy engineers singing a hymm to a failing reactor
First set using that description gave country, so I added "gregorian chant" at the end which led to terrifying industrial, so I'm trying with "rock opera"
try: Gregorian chant style, featuring monophonic, slow, and solemn tones with deep, resonant vocal harmonies,
I might DM this to you just to keep the tech chat from getting cluttered 🙏
if you can 🙂 that was what i put in the original song.... though i love the idea of the Holy Engineers of the Ignavus singing country style to aplease a reactor that is failing lol
ah yep forgot we were in tech chat. I am gettng ready for work and waiting on image gens for another song so bouncing about a lot
Hello! I would like to upload my voice singing a cover and have Suno create the instrumental track while completely preserving the melody. Is that possible? I'm experiencing issues where it either invents the melody entirely or keeps half of it while changing the other half. It's important for my voice to be included in the track as well. Is there a way to achieve this?
Hi - I'm struggling to get a "heavy metal galloping guitar" (very common in this style of music). Things I've tried:
- [Galloping Guitar] in the lyrics window.
- Galloping guitar in the Style of Music window.
- Putting the words, "Dum-di-di Dum-di-di Dum-di-di" in the lyrics window.
None of this seems to work...? Any suggestions?
What exactly do you mean by a "galloping guitar"
Hi, has your problem been solved ? I'm still not able to get the .wav from the Uploaded Audio Extension.
why is there no android app? we pay the same money as everyone else. Should we not be given equal access.
Would you like to make the Android app?
?warn @keen vortex Pleae do not post the same thing over and over again. See rule #3 @keen vortex
sorry!
its not my position to do the work of suno. They get paid to do that. if they want to pay me then i might.
Great, they are looking for developers, go apply
This is the kind of the thing that is possible to achieve as an end result, but it isn't a specific feature, so it takes some patience and trial and error. It may also involve at least a small amount of manual audio editing to fully achieve. It's not something I have specifically worked out a best process for, but if I happen to remember or notice a good workflow, I will reply again later.
Thanks for your answer Jonathan!💯
So, are there any plans to do something about the doubled / fuzzy / reverb'ish sounding voice quality? It's been quite bad since March.
I’m not experiencing that issue with my vocals, tbh — if there’s any particularly bad instances you could report them with … > Report > Bug > Poor Audio Quality so the teams have actionable data on which songs were dissatisfactory. 💪
i've done so multiple times over the past half year, as have many others. if you don't even know what i'm talking about, i guess this is pretty futile.
if the people behind suno cannot tell how bad it is, then they won't be able to fix it.
all i can say is: i'd love to have v3alpha back.
the sound quality of that was superior to both v3 and v3.5
I'm trying to do a choral version of a song, but it just keeps producing the same song (with a soloist omnipresent) every time. Anyone have any tips? This is my prompt: Haunting Full Choir, A Cappella, Melodic Counterpoints, Counterpoint Vocals, Cinematic, Haunting, and I have each prompt within the lyrics marked as full choir as well.
"An error occurred generating lyrics.
Please try a different lyrics prompt."
The prompt:
"eating pumpkin pie for dinner, then I transform into a giant bloodthirsty pumpkin, with sharp fangs, and roam the streets looking for unsuspecting victims."
Oh I'm surprised you're not getting the "has name of an artist" error for using the word pumpkin. 😅
I’m really pissed because I am prompting male voice and I’m repeatedly putting in to the prompt that it should be a male voice a deep voice and now I have like 10 different tries and it’s always a female voice. That’s really annoying.
I'm not one of "the people behind suno" -- I'm a community volunteer who is asking you to take action instead of complaining.
I'm trying to convey that for every report of a fuzzy/reverbed song, there's countless others that seem to satisfy the users -- this leads me to believe that either you believe that everyone else has zero complaints, or aren't musically-apt enough to hear what you hear. I'm also trying to suggest that perhaps your output might be influenced by what you're using as input.
So, without providing any examples of notable reverb, if you're going to suggest that any attempts to talk through the matter or providing actionable data for the team are futile, I'll leave you to the matter.
Best of luck obtaining non-fuzzy vocals 💪
Are you comfortable sharing an example of what you're getting, either here or in DMs? I'd be happy to look into it, as I'm used to acapella / a cappella giving group vocals, so I'm wondering if it might be something related to the combining of prompts, but it would be more directly helpful to have the lyrics as well, since they can influence the style as well 🙏
there are three examples of what i call "fuzzy" voices.
one more for good measure.
Honestly I went back and listened to some of my v3alpha songs recently and those also seem fuzzy to me these days. Covers seems to use a tweaked version of 3.5 which can be a bit better. I'm pretty sure if they offered v3alpha again people would say it's not the same model.
i have lots of v3alpha songs on storage, and i still like them
It was available for what, 2 weeks or something?
And then V3 dropped and everything turned into a choir and rapidly distorted into noise
No, the grass was definitely greener back then.
It's not a matter of all v3 models being capable of producing a variety of results, and it's not a matter of discussing the frequency rate, it's about an absolute and sudden change for the worse.
Probably intentionally making the product worse so we spend more credits, right?</sarcasm>
(the most common complaints)
https://suno.com/song/4f6619f4-d5c9-4b54-bd1f-3daae1eea1d7 v3alpha. notice how the voice is solo most of the time. sure there's some doubling in the chorus, but it's not super muddy.
https://suno.com/song/d137fa7a-9f03-474c-8733-f3ee70d641ad another v3alpha song with a really nice voice.
i think the difference is obvious
Are you comfortable sharing the track here or in DM? Earlier in the week I was asking if anyone was running into a song that absolutely refuses to present with a different timbered vocalist, especially gendered
https://suno.com/song/7b06a479-ed67-46e0-928a-384ce59477a0 v3alpha acing a good deep male reggae voice. how i miss this!!! ❤️
so ... now you have 4 examples of the current models + 4 examples of v3alpha. i hope you can tell the difference.
Another difference is the new models are more prone to making certain specific genres (namely pop). v3alpha didn't have that tendency. But that another story...
I think this is the last time i'll bring this up. From now on i'll just wait and see.
guys i need help whenever I click create i get system error occured can u guys please help?
v3.5 is often criticised for that. Covers seems to produce more diverse results so I think whatever that uses has additional training data.
v3.5 has that problem a lot. but v3 too.
these days i prefer v2
but nothing comes close to v3alpha, i'm sorry to say. but none of the current models would have made me pay a year's worth of subscription.
i'm puzzled that v3alpha isn't simply offered in the list of selections
guys i need help whenever I click create i get system error occured can u guys please help?
#💡┃suggestions-web if you'd like to request it, perhaps.
https://help.suno.com/en/articles/2551617
What all have you tried to fix the issue? Unfortuntately "system error" covers quite a lot.
This knowledge base is a great first step, but you may need some additional assistance. Since you're already using it, help.suno.com is the best pla...
covers may not be working? I tried to do the same song with wildly different genres like EDM and bluegrass, and I kept getting the same result. Can anyone confirm.
@charred merlin done (again).
i don't think there's more i can do really.
i hop they'll do it though, because you guys will blown away!
it's epic
This was an idea being raised a few months ago. One of the devs was chatting during a weekend and myself and a few others suggested making it available, even if it was just as a subscriber-only thing. In the end we got v3.5 instead. There must be a reason why they didn't offer v3alpha again. It was known to do very janky things with lyrics for example. I expect a lot of people would try it and go "ew what's this, it's so broken, I want my credits back", whilst a handful of people would be happy about it existing.
But, I do hope that v4 brings significant improvements given that it's been under development for quite a long time now.
You can report the songs that don't meet your expectations, and wait for the next model to come out -- that's the most effective way of expressing how much you don't like the current vocal offerings.
@unique stone given the direction things have been going in, i'm not sure v4 will even beat v3alpha 😅
i hope so, of course
I'm looking forward to when v4 comes out and people just complain that it's the same thing as making a v3.5 and then covering it. And ask why didnt they wait to release it longer until it worked better
it seems to be that v3 and v3.5 have been overtrained, and i'm guessing they need to start over to fix that
which is why v4 probsably takes a looooooong time.
then again, the fact that i have to explain and demonstrate that v3alpha sounds better (which to me seems as obvious as can be), makes me wonder if it's gonna get better ever.
"which to me"
The most common complaints about v3 were at least addressed with v3.5 (my main gripe was the metallic ringing that developed, so I made sure to thumbs-down everything that headed in that direction during the early-access phase).
Whatever they've done with Covers seems to further address the quality degradation. I've got some really weird results with that as well (a couple of guitars that are singing lol)
...I'm at 5/6 that are almost identical. One with "acceptable" variations, but not the requested style, and still very much so the same song.
The son or the songtext?
@safe hazel that sentence came out pretty bad. 😅 i hope it still made sense
One of the songs, that way I can re-use the prompt exactly as it's set up, and see what variations might help influence it, or to see what prompts might be negatively impacting the desired results
DM
My point was that it's a subjective thing, the version you like sounds better to you, it doesn't mean it is better objectively.
i actually disgree. when i listen to non-AI music, most genres have vocals that sound like v3alpha - clean, up-front and with not a lot of reverb. only certain genres have these super doubled vocals with loads of reverb on. for example some kinds of metal, and most 80s productions.
70s motown had clean sounding vocals. Or typical singer/songwriter stuff.
How much reverb and voice doubling you do is a sound engineering decision made by producers and bands in cooperation.
Some genres should have heavily effected vocals. But not all.
You can diasagree about a subjective quality all you want, that still doesnt make it an objective one
My point is that it's not entirely subjective.
My point is, that it is
Digs up one of my old v3 alpha songs
This sounds pretty crusty to me: https://suno.com/song/ac9c863c-491a-4592-a8de-29054350642f
I have recently experienced the same problem, even negative prompt doesn't help.
Example 1: https://suno.com/song/d202defb-059e-4dd2-99e3-9ec855b2935b
Example 2: https://suno.com/song/e65094a9-4255-493a-87fe-4350edf59bf7
@unique stone of course there were bad generations back then too. And you also get good ones now with the new models. But the general tendency over many generations is where i notice the difference. I can keep clicking like forever these days, and i only get reverbed voices most of the time... and pop...
What's your sample size, what percentage is "bad" versus "good"
Interesting discussion about fuzzy vocals, I am a relatively new user and only used V3.5, and I honestly believed fuzzy vocals are just an unavoidable byproduct of generative AI.
If you're still subscribed: Did you experiment with Covers?
In particular, a fun little trick: Upload 30 seconds of silence and cover it with whatever style you want.
@rain sky there are other AIs that don't have that problem
@austere dragon -- I just made 20 covers of a song I'm working on & wanted to share the results for your comparison -- 4 "obvious successes", 4 "companion pieces" that would work well on an album with my intended result, but bears little-to-no resemblance to my desired song, and 12 results that were the exact same track returned, with only changes to speed or pitch.
I know it's still beta release, so I'm going to thumbs-down the 16 results, but it's definitely a noteworthy ratio.
For the record, I'm using a rock song, giving the style prompt "ska, trumpet, trombone, saxophone", and using the negative prompt "guitar, electric guitar, distortion". Usually it's an obvious success rate of 25~50>%, but not for this run.
How do you determine "bad" versus "good"? What about in the middle?
in fact, if the fuzzy voices are annoying someone, try the v2 model. sure it's weird, makes short songs, but the voices tend to have less reverb on them. i'm sure you'll notice after a few tries.
I have tried v2, they dont sound better to me at all
@rain sky try skipping to 02:16 in this track. is this the kind of clarity you were hoping for? https://suno.com/song/1f6ad2c3-907b-40a2-bb01-2497b7f581e3
Ah, yes, that's actually awesome
I replied in the suggestions, but it should probably should have gone in this channel #💡┃suggestions-web message
The reason I am looking for clear vocals is I am trying to generate doom gothic metal songs in the 90s styles, with clear, high pitched, opera-style female vocals. I was not able to get to that point, so I simply dismissed the attempts as "the models aren't ready yet".
Generally, I have noticed the following specifics while generating songs with V3.5:
- The vocals are, more often than not, hard to understand, they seem "drowned" in the instrumental part. It's a matter of relative volume amplitude.
- The vocals are fuzzy most of the time, even while using prompts such as "clear vocals" or "high quality vocals" - now, I might be using the wrong tokens, maybe it's just me.
- The model tends to ignore specific vocal prompts, even while using negative prompts. Trying to get male vocals seems an uphill battle sometimes. See my examples above.
i'm glad i'm not the only one who noticed.
If I hadn't stumbled upon this discussion, I would have never thought it would have been (or it HAD been) possible.
Check out the reply I posted in suggestions. Try not worrying about nailing the perfect vocals in the original generation with V3.5. Instead try a workflow where you nail the overall structure and melody of a song, not super focused on things like clear vocals, and consider that like a first draft that you then iterative on with features like Cover.
I tried; cover feature eats credits like the Cookie Monster, over half the songs generated that way are identical, or near-identical to the original. I am not there yet (with credit eating) because I still have free generations left, but it was pretty time consuming and a bit frustrating.
I'll keep trying though.
re: 2 -- the prompts you give Suno aren't explicit hard coded commands -- everything you provide to Suno is a suggestion that it's able to ignore or move past if it starts to get overloaded.
That's why I'm always reticent to provide help when someone asks what "the correct prompts are" -- personally, I believe it's less a matter of using the "correct prompts", and more a matter of how you stack the prompts you're using.
It's super tempting to litter the song with metatags, thinking it's a way of bypassing the 120 character limit in the style prompts, but that's implemented for a reason.
Oh, I almost never go anywhere near the limit
Here's a song I just generated, positive prompt is 88 characters, negative prompt is 20 characters, it still generates a, well, mostly female voice.
https://suno.com/song/47771a30-879c-4cf6-9296-53a51cd11689
You have to be careful with negative prompts because negative promting gives the AI a very big range for what she can do.
e.g. Don’t use deep vocals.
… means she can use every kind of vocals which are not deep.
So this is definitely not that focused on what you want . So I would always recommend you to give clear comments or instructions what to do.
instead of saying what not to do and you can also say instead of “don’t do that” you can say “you are forbidden to do that.”
@untold briar hmm, interesting. i did a song, and then used the "cover" function to regenerate the whole thing, not changing any settings. this generated two identical files. but they both sound slightly more clear than the original. 👍 so this is promising.
Of course, this doesn't address the general tendency of suno to steer towards pop, but the new negative prompt helps with that i guess.
Suno can definitely use all kinds of vocals that are NOT female, but if it still uses female vocals... then what gives?
Cover has a tendency to overly duplicate a song. In one sense, this is very annoying users who are looking to change a song the way a cover typically does. But as you noted, when Cover duplicates, it isn't 100% identical - it often sounds just a slight bit better than the original. So this problem is also a feature, and I wouldn't want Suno to tune Covers in such a way you can never use it like that.
actually, the two covers are not entirely the same
Right. And furthermore, a Cover of a Cover, can slightly improve the Cover. And so on. I'm not saying it's easy on the credit balance, but you can really get to places like this.
so i guess the best clarity will come out of leaving all settings the same when you do the "cover"? or are there strategic modifications that can be done to make the 2nd "pass" improve things even further?
i just did a cover where i changed the style, and this seems to blurry things up a bit again
There's no simple answer. I've been out of action for a couple weeks and I think Cover was slightly tuned in the meantime, so I'm not even sure what a good guess as a recommendation is at the moment. But overall in Suno it's difficult to generalize advice, and it's always better to at least try a range of inputs and options.
alright, i'll experiment. 🙂 but thanks for the heads-up. i'd still love to get v3alpha back, but i'll give this some tries.
To give one example, I've totally transformed a song style with a Cover, and then transformed it back into a style very close to the original song with a Cover of the transformed Cover, but with no reverb and everything clear. Sometimes with multiple midpoint covers in between. So even when things are different or moving in the wrong direction, that could still be a step to get where you want to go.
oh wild
What do you mean by "midpoint covers?"
Huh, you know what, it IS interesting.
I just tried it, and it's pretty impressive.
Original: https://suno.com/song/ddb52510-6dc8-4cf6-86e5-612e55da383a
Cover 1: https://suno.com/song/9cba4008-4af8-40ab-8bfa-9580ecfe3731
Cover 2: https://suno.com/song/e8b5fb02-59e1-47f9-b2f7-20313a221f6f
Cover 1 sounds better (clearer vocals, voice is closer to being creepy)
Cover 2 is, well, different. Vocals are fuzzier, harder to understand, especially when there is musical background, but the ending is "better" 🙂
like covers of extensions?
Hi there! I am new to Suno and creating songs. I am really happy with almost all of them but have a question on how to clean up the robotic sound of the voice.
well, that's pretty much on-topic!
@foggy lark for starters, try doing a "cover", don't change anything, and then simply click "generate".
Let's say I'm trying to turn my Pop song into a Polka song with a Cover. It doesn't seem to be working. Rather than try forever, I should try turning a Cover of the Pop song (that is either close to the original, or even just a non Polka cover) into Polka instead. Adding a middle step or steps.
Not sure if this has been asked but can i somehow turn my scene into a full song and make public
Basically "Cover didn't work on the original song a few times. Before I waste more credits, let's try Covering a Cover (which may sound identical) instead..." This often works.
I see, make transitionary covers, cool thank you!
That's an interesting idea. I guess by changing it to something and then changing it back, it loses sight of the original.
Genuinely loving this idea -- turning reproducible negative into a strength 💪
Doing a cover cover cover cover now... this feels like watching a music version of Inception
sorry, not really understanding. I put in my own lyrics and put in the style of muic "cover"? All my music is original lyrics
Hey can someone help me? I cropped a song at the end of the lyrics and I dont want the song to end this soon. So I want to slow down and prepare for a drop and a last, full energy, chorus. I get it to work like 1 in 10. I tried [slow down][low volume][preparation for drop][massive drop][low-frequenzy build][drum explosion]
anything I'm missing?
basically, once you've generated the song, and you want it to become a bit more clean, use the "Create" menu -> "Cover song". That will give you this little extra section on top:
Then simply press "Create".
its worth "refining" the songs with covering several times?
so far i've only had luck doing it once.
You have to be on a paid plan to be able to use the Cover feature
twice or more still hasn't given me anything good
Ok but one time is totally worth it?=
i'm not quite sure yet... maybe i just got lucky... i just started doing this half an hour ago 😅
Oh ok, keep me updated 😄
Oh, I am on the paid plan lol
so far the best result was doing it once - without changing anything
Then there is a feature you can use to make a cover of a song, called Cover
https://suno.com/song/7218513a-f7cb-43a4-a537-44c1c2b2657f this is a cover-cover-cover-cover, and it's not exactly what i was going for. it's not easy.
hmmm, I do not see anything about cover.
Go to the Create drop down menu, there should be an option to "Cover song"
Suno is the best platform I found for creating fully custom music. Just sounds a lil too AI and warbly at times. trying to find an application to clean it up as well.
i seriously doubt it can be cleaned up very well
ooh, this trick works with v3.0 too
Just wait until we get every instrument, sing, voice line etc. seperated, so we can just adjust the tone of the singer or add/remove some instruments etc.
as well as Seeds for the voices ❤️
ahh, it uses v3.5 to do the cover even if you've set it to v3.0
the quality still degrades towards the end of songs despite using this cover trick
I am trying myself at the moment, lets see
Oh ok, I found it! That acutually helped!
It starts out so well... https://suno.com/song/11f8c7e7-5145-4549-a111-57631bf9cf7d
Thanks for your help! That is why I love jumping into the platform discord lol you learn so much!
Pretty nice results in the first run. I guess I will have to dive deeper. Thanks @agile veldt
Might want to start adding in [make a cover of "your" song]. People still confused that "a" song means any song. 🙂
yeah, doing a cover of a song noone has ever heard is a bit confusing
maybe it's a "phantom cover"
Ok. Now I'm just curious what "Will it get fuzzy" is all about. Lol
then click my last song link and notice what happens at the end of the song 😄
They just tuning up for the next session...... or clearing the spit valve.
oooh!!! this time the cover trick really worked!
sound quality went right up
but... still a choir
https://suno.com/song/b36bf96b-db71-413a-9409-d9e92ed7d544
what happened to this song lol? I used the cover function and now it is in some other language or so? hahaha
I would love to know if this is a real language or just something suno has come up with
it sounds like when kids who haven't learned english yet are trying to sing an english song
Here I used the simple "cover" approach:
- Generate.
- Do cover using the same settings.
Pretty clean!
Yup! Not quite v3alpha levels yet, but certainly better than I'm used to.
the instruments get a lot better this way
i also EQed it a little bit
But it loses some characteristics when the grungy style is appealing
https://suno.com/song/6601a14b-326f-45ad-8589-53ad41dcb47f <- this is the normal version
https://suno.com/song/617d6f76-ed47-4522-b438-28ce4592e24c <- the first and best cover so far
much cleaner, but I liked some of the dark tones aswell
it has certainly changed a bit in style, but i feel like the cover is a better recording (if it were a recording)
Ya like the original was some kind of garageband-recording 😄
@polar spade I'm making a more comprehensive demonstration playlist, but I think you might have hit on something really good with "motto"
Anyone else getting this while updating the song picture?
I seem to be getting it for only one song
-
- make sure the picture you're uploading is a
.jpeg/.png/.webp/.bmpunder10mb
- make sure the picture you're uploading is a
-
- adjust the slider from the default position, as this seems to resolve some issues
If the issue persists, you may need to contact support@suno.com, but I'd be surprised if the steps listed above, or trying a chromium-opposite browser doesn't resolve the matter
It was under 10mb, but it was 6.6mb, I reduced the size and it then uploaded.
Thank you for popping back in to report the weirdness -- super helpful in figuring out other things to test out, even when everything SEEMS like it should work 
Ahhh.. suno likes motto , eh. Nice work, brah.
When you're trying to force short generations, and you get credits refunded due to short generation times 18/44 times in a row, something is going right 👀
Did you hear that short 15 second cat song I linked?
https://suno.com/song/fa206715-4480-41c9-83ab-5c9be04b0d21. I tried :
- singing 6 seconds to audio upload, but add 6 seconds of Silence before stop.
- extend it , maybe I'd get a short gen based on the premptive silence ( didn't. Alway about 1:30 extension) but would get music that wrapped up nicely in about 10 seconds before continuing.
- Did a cover to clean up the sound.
- cropped off the first 15 seconds.
So, it's a fake short song, but, it's at least possible to get lucky . If I ran a couple dozen, I don't know how often I would get a short result.
Now I know I can add "motto" as well.
But of course, I used opera, so to get orchestral hits like that around 12-16 seconds is easier.
While we're not able to choose the specific length, I found through testing that "Motto" works PHENOMENALLY as a style prompt to force a short generation -- in my testing, 84% of the generations resulted in clips less than 30s, and they were short enough that credits were refunded ~43% of the time.
First Prompt Experiment Playlist in almost 3 months, and it feels nice to be back with such a resounding success -- thanks again to @polar spade for the suggestions -- looking forward to playing with some of these results and making some neat variations on a producer tag 🙂
Hi, anyone also experienced this issue? it has been weeks and i cannot get hold w/ anyone
I want to pay for membership, but it happened stopping me doing anything
Hey @vagrant gate, when you say you can't get a hold of anyone, have you emailed support@suno.com?
That's the only official way to contact the support team -- everything else is just supplemental community help 🙏
Hi, Yeah, I have emailed them couples weeks ago, and got no response
I know they've been busy lately, but waiting weeks for a response definitely doesn't sound ordinary, and likely isn't the experience they want you to have.
There isn't a proper support presence on Discord, but if your support request somehow fell through the cracks like it sounds like it might have, I'm sure the support lead would be very amicable to getting in touch, even just to see if they can get your support ticket bumped up the queue, so to speak.
I think support is on american business hours, so they might not be on tomorrow, but I'll ping them so at least the matter is on their radar when they're able to look into it 💪
@steady kelp
You are my angel 👼 Thank you soso much
In regards to the error message itself, that's a fun one -- unfortunately because of how vaguely worded it is, we don't have a lot of information specifically about what causes it.
I've seen someone people resolve it by swapping browsers, some seem to try everything we can think of and nothing seems to come of it. Sorry to not have more actionable advice at the moment, but I'm hoping that you hear back from them soon, just the same as I hope the error messages get a little more explicitly worded for clarity's sake 😅 😛
thank you for your kind help. really appreciate.
Tried to clear browser, using another browser, computer etc. but still not working 🙈
That's some serious testing, kudos on your thoroughness (this reads as sarcastic, but I legit applaud testing on a separate device; that's SUCH an effective troubleshooting isolation step) -- even though none of the attempts resulted in a song, it's still valuable data.
When you finally get in touch with the team, having this kind of information will help them narrow down what the issue might be.
I'm off to bed though -- best of luck 
Should I also give you my credit card information and other personally identifiable info?
my SSN? and the digits on the back of my card? of course i wont give that
Will making scenes come to the web eventually? I have videos I want to make into songs but they're on the computer.
Hey Sunny! If our team didn’t get back within a few days, there’s likely a block somewhere. Shoot me a DM with your Suno handle so I can take a look 🙏🏻
Can you increase the upload limit to 3 minutes for premium members?
https://suno.com/song/a2d16b65-786b-4686-8c6a-4b5e2a4b4172 Man, I suck at lyrics. I liked the instrumental I've made earlier so I tried to add lyrics to it but I'm not good at it. It's like Anthony Kiedis with the Red Hot Chili Peppers crapping on a good instrumental rock band.
Any songwriting tips?
Anyone got an idea what happened here XD
AI completely ignores what I wrote, and goes off on a nonsense song about Jesus Christ and politics?
I have the feeling there is some kind of "temperature" or "cfg" parameter internally, that influences how strong a song differs from the prompt or the cover
and sometimes they sample a really high value and you get something really different (which might, in some situations, be good and creative, though)
I really wished they would make this parameter accessible
don't really know where to ask this but is there a way of stopping Suno from automatically signing in on one predetermined google account, want to check out songs i had on other google accounts but whenever i try to log in on said accounts i don't get a say in the matter and put on said predetermined account
And at 2:55 it changed from English-y to German-y.
These are hallucinations Suno gives. It can be short blips, fill-ins where there is silence, or as you've heard, a complete anti-song, but still related to 1 or more of your prompts. You'll notice the increase noise towards the end. Suno may also speed up, or on occasions, slow down.
Some, though the words are, as I call it , babble-scat, are nice sounding, completely rhyming gibberish songs but with great music and arrangement worthy of radio/streaming, which we find helpful to just re-run and write actual lyrics to. If you listen to the suno radio portion, you'll discover members will often just post their suno hallucination songs as they sound great. Since often they sound like a foreign language, it's hard to tell it's just gibberish.
Thanks Mike, 🙏🏻
Hello guys, I have a doubt, I created an account with my personal email to try suno, I liked it so now I am looking to get a subscription, but I want to use another email and create a new suno account with that one, mainly because I am going to start posting on souncloud tiktok and youtube content ID and youtube videos, so I want to avoid suno or any of this plattforms to think I dont have a suno subscription and ban me, so thats the only reason, what should I do? should I just upgrade my test acc I already made even when the email doesnt match with my youtube channel and soundcloud and soptify email or should I create a second account, I heard that is not allowed to have more than one account and it is not my intention to abuse credits or something like that, I am going to stop using the first suno account I created if i create a new one
Good to see that there is s.o. else waiting for response and facing the same issue I still have. For me also no response now for 2 weeks.
I think the wording in the ToS says "you're not able to have more than one account for the purposes of taking advantage of the free credit system", or something to that effect -- afaik, you're allowed to have 1 free account & as many paid accounts as you'd like 💪
I'm fairly certain it uses the browser's SSO, so you'd need to go to a google-based website, click on your profile picture & "add another account" -- then whenever you click the Google button to sign in on Suno.com, it takes you to the account selection page. 🙂
Alternatively, if you don't always want to be presented with those options, you could visit suno.com from a private browser 🧠👈
thanklyou for replying, I get it now
has anyone managed to figure out how to get Replace Section to act like cover when tweaking lyrics?
Oh let's see if he has learned to behave himself yet
@final hazel Must I put up with this harassment whenever I speak here?
Daizy I did not say anything harassing
That's just it. I'm attempting to, yet I clearly won't be left alone
You 2.. STOP TALKING TO EACH OTHER
There is a block button. I suggest using it.
100% agree -- I'm all for raising hell when it's deserved, like when users call others "degenerate" for their personal or cultural identity, but everything seems copacetic at the moment -- I have faith today is going to be a great day for everyone 
now, Replace Section... making it act like a Cover. How do?
the original Bridge had a decent little melody that kinda descended during the 2nd line, bottomed out near the middle of the third, and then came back up during the 4th, but now all I get is lines where the melody just feels flat... not capturing the original at all, despite my attempting anything I can think of in the Style prompt box to influence it that way
Can I take a song that I have already confirmed to become a Full Song and change the lyrics in it still? Keep the rythm of the song/melody but just have the lyrics change still?
You can try with the cover feature or the replace feature. But it isn't foolproof. It often times doesnt work the way people want
I only managed to get it to happen once with a Cover. I didn't get all my changes, but I got the major ones. The only drawback was my male vocalist got replaced with a female, giving the same performance. I'm trying to get the same thing with Replace Section to keep the male vocals, but I'm currently having trouble with the bridge.
Apparently, it does not look at the part it's replacing for any kind of hints for what to do, so since there's no other Bridge for it to look for, it's going to just make its own thing up :/
I can't seem to find the option cover-feature. Instead I find where I can change song details?
Before I try anything I suppose, can I duplicate a song so that the original is still saved?
The Cover feature is only available to paid users (same with Replace)
It is located under the Create drop down menu
There is a helpful image
Yup, found it thank you.
So can I somehow duplicate the song in case I want the original to be saved?
Nevermind it all.
it will make an additional song, it wont destroy the original
Your original won''t go anywhere
No irony here either: I like it that you stay calm and try to help everyone with that same issue I expereince since more than a week now too.
So if I do then press cover, do I then change lyrics and press create again?
You can, but it's no guarantee that your new lyrics will be accepted
And as you said, the error message is just too vague. I tried to narrow my probably specific error down by opening the brower dev console. Went through the start of SUNO, saved the log. pressed "Create" and saved the log. Sent it to support@suno.com. No response still...
My chorus had three sentences copied to be sung twice, removed it in what I believe to be the cover-editing area and nope... Still singing it twice. 💀
Just says that the cover is still generating. 🫤
Nevermind, think it finally works now.
did the lyrics change?
Yup
@untold briar you are a king
Or well, in the lyrics-panel it has changed.
oh
maybe I proclaimed Jonathan a king too soon 😅
The code is completely non understandable scrambled. And my knowledge in js is on beginners level anyway. But... from the network side there are four things that don't look correct.
- [hCaptcha] Callback 'checkVisitor' is not defined.
- Failed to load resource: the server responded with a status of 404 () locales/de/translation.json:1
- Uncaught Error: Minified React error #418; visit https://react.dev/errors/418
- POST https://studio-api.prod.suno.com/api/generate/v2/ 403 (Forbidden) 5636-2c036c11a7a97a5b.js:22
No. 4 is the one when pressing "Create"...
Naw, you're right -- At the very least he's a Duke or something, pretty sure.
any tips on maybe directing chords via metatags, perhaps?
Now I see what this crud had changed in my cover...
I removed so that Verse 1 had one single line sung once instead of twice...
And then changed so that I had went over to I have...
INSTEAD THE COVER COMPLETELY REMOVED VERSE THREE AND FOUR!?!?
No no, screw covers.
Ooof -- it might not fit that use case, or might not have worked this time, but those are big words dismissing a remarkably powerful tool.
😛 Obviously I'm being silly, and it doesn't really impact me, but I urge you to keep an open mind about playing around with Covers some day if there's a song you love & want to hear reimagined.
At the moment, the beta-release of Covers trends toward returning the original song with minor variations, but you can get WILD changes to your songs!
For an example, "aporia" was one of my first "serious" Suno songs, and it was the first song I thought to make a cover of -- it's still a wonderful feeling hearing a lovingly-arranged reimagining of a song I care so deeply for.
How do you use uploaded samples?
I tried 'cover' but it just plays generates a copy of what I uploaded
You need to Extend from your uploaded sample
i extend from end, or from, let's say 00:01?
if you extend from 00:01 it will only know 1 second of your upload, if you extend from the end, it will know all of it. The longer the amount you extend from, the more suno will adhere to the styles in the upload
Extending from 1 second wont give you a result that is very much like your sample
extending from 60 seconds of an upload will
but then the end of the uploaded sample will flow in the start of the extended part, no?
yes
what's the point of that?
to get a song that has some similarities to your uploaded sample. Like voice, or instruments etc
suno can make a song in your voice this way
ytes but then my song doesn't have a start, it'll be the uploaded sample
you can just not use the "Get whole song" and then you wont append the uploaded sample to the song. Or you can crop it out. Just try it out
think of the upload as you telling suno what styles and vocals you want
Anyone else getting weird errors with covers? I tried to do a cover of a song created a couple of weeks ago, and Suno shouted at me that it was violating their ToS, yet the original song itself is available and fine. Can even extend it and everything.
where is 'not use whole song'?
And if i crop it, it will still not have a clean start
the cover model has a different copyright detection. it should have been detected the first time. sorry that it just got detected now.
it will only have the uploaded part on the generation if you use the "Get whole song" option. Just try it
i don't see a "get whole song" option
There is no copyright issue though, it’s the lyrics to a public domain Christmas song?
Although it doesn’t let me create a new song with the same lyrics either now.
it's the same "get whole song" option that you see on all extension parts.
you're just making an extension from an upload
ohh, ok i get what you mean now, after the generation ;)=
usually i make non sensical dance songs with just "yeah" as lyrics. i am not that keen how it changed recently.
yes
Ha. If they crack down on public domain lyrics, I’m having a problem, I promised to make a load of public domain Christmas stuff for a kids group.
Shall be interesting to see what else is impossible now.
we are here during the birth of all of this here. i am pretty sure things will change... If it takes the usual way from wild west to civilised it will only get more restrictions over the time... sadly.
look back 10 years and how technology advanced, now add AI to the mix and look 10 y down the line 😉
Probably. They still should understand what public domain means.
Gonna try another set of lyrics and see what happens.
in theory public domain is the exact opposite of "you own the exclusive rights" ... because everybody owns them...
OMG it's terrible!
Like i thought, it extends the end of the uploaded sample and makes a mess
Meaning everyone can use them as they please. Just tried two other public domain Christmas songs, same issue. Off to U, I guess. Let’s see what they say to that.
the beginnings can be a bit of a mess sometimes, where it like transitions from the uploaded file to the generation. but you just trim that part off
it's just a database of known songs, Suno didn't make this list themselves
it's not just the beginning, none of the song is right
keeps switching between original lyrics of sample, and the lyrics I put
I don't think there's a mutual exclusivity, but Mr Tree has a point -- public domain IS shared by everyone, but Tree is pointing out that specifically for use on Suno's private platform, they require you to have exclusive rights over the content you upload
Stupid example, but everyone is allowed to wear a hat -- but if a business says "no hats", it doesn't matter what you can wear OUTSIDE, but INSIDE it's just a matter of playing by their rules.
the wording in the tos of suno exclude everything that you dont own "exclusive rights" to. Thats the only real "issue" here.
"I'll make my own business where people are ALLOWED to wear hats inside" is more of an appropriate response than "this business should be forced to accept that I'll wear this hat because EVERYONE wears them"
Go away with reason and logic 😄
You can use the public domain songs as you please, yes. But you cant use Suno as you please.
In the position of suno in relation to the traditional music making / rights scene i would probably exclude the same things...
Dunno, I cannot quite see what problem would arise for them, but their house, their rules, obviously.
Recanted...error
Arguably -- if you use public domain lyrics and don't get an error message preventing you from doing so, while it's still technically against the ToS, I doubt that Suno or the estate of Good King Wenceslas will come at your for having used the lyrics from a PD Christmas Carol.
That said, if I had to guess at why they'd consider it a problem -- it could be that if user A uses public domain lyrics, and is able to copyright their output of the song, it could pose a ugly situation if user B uses the exact same lyrics and inadvertently uses the same style prompt -- given the nature of generative AI, there's a larger chance of Suno outputting something that would flag as having been copyrighted by user A.
It would eliminate a lot of these potential issues by making efforts to to ensure that everyone uses content for which only they have exclusive access.
there is either the link, or the embed. i am not sure what picture you mean... maybe the "unfinished render" static image... but thats not forcable.
See the lack of play button. I can put the < > to remove the embed but I am not allowed to post the picture
for me it looks like this (pc discord app)
Out of curiosity -- are you on a tablet or mobile device?
Oh..,well alright then. My bad
The other day there were some users discussing how the embeds looked differently for them in a similar manner.
Mobile. I know there are differences. That is my bad
Thanks for the answer
The only reason I even care about the embed is because it shows the original version of the song without the new picture and title, or the cleaned up lyrics
No worries! There's tips and tricks we've come to learn about how to make discord embeds do what we want -- I think it was Mr Tree that taught me that you can add a / if you've updated the image for a song but discord is still caching the old one
Oooh! I just mentioned that, but try to make sure the video is regenned on suno.com and update the link the the / at the end, since Discord considers that a different link
i think thats the case of discord has cached the pre regen version. you could fix that by adding a / to the end of the link after the vid has regenerated to make discord update the cache
Prost!
Ok, I shall give that a go
How long does it stay cached? Can you remove the / later after it's cleared?
just leave it ... it doesnt matter or influence the link.
That's a legit mystery to me 😅
None of mine have ever changed.
However, I haven't tried the / trick
Like a php force cache clear
More that it's considered an entirely new link altogether
Ahh...so not a refresh, but a way around the cache...touche
oohh i just had a look at daizys video, and i think its bugged. you dont see any styles or lyrics in the video. i have one of them too, but i dont know why it happens
But there's only the one chance, as it would seem -- if you post the / version of the link without having made sure that the video is properly regenned, it will just cache the old version for the / link as well.
Whatever .mp4 video downloads SHOULD be what's reflected on the discord embed
Yeah, like for each song that you make, discord & suno allow you to reserve one extra cached video incase you make a mistake -- but it's all automated
I had noticed it before, other users had it without even admin privileges. Those I figured were just errors
now I get to have fun digging through my unwanted sections for sick cover art
the "mod stuff" is just here for discord. we can post funky things but that doesnt have any influence on our suno. we dont have any special stuff (except maybe some fest features) that would make these broken videos possible on demand.
If you notice sometimes the embedded video doesn't actually have a "play" button, and it's just 2 pictures?
That usually indicates the song was shared before it had finished processing, so there wasn't an .mp4 for Discord to grab.
Similarly, It's possible that SOME as-of-yet undetermined change to a song makes it appear differently, like the use of a feature, or having renamed, cropped, etc
Yea, it is the lack of embed. I'd have no issue posting the link in <...> and add the picture manually, but I understand why it is not allowed. I only know because I was seeing on my server if there were ways to make it so people had to go to your page to listen. When I came back here I noticed I couldn't manually add the picture. I assume as time goes on there can be some webhook or other tie to the website so plays and likes and whatnot can go through.
Are there devs on this server?
Oh, I'm also glad to see the bug that caused the start of songs to be cut off after replacing sections got fixed. :3
I couldn't see it anyway. I won't be able to do any more testing until I am on a PC
So you linked to the picture and then posted the <link> or something of that nature
exactly. i copied the image link from the songpage, posted that and added the song link in <>
Thank you very much. I will give that a go
They're the users with purple names, but don't often poke their heads in for chat -- but they've set up #💡┃suggestions-web #📱┃suggestions-ios and #🚧┃suggestions-layout for ideas about suno, the iOS app, and the layout of the platform, respectively 🙏
Hello, I want to report a failure in the self-moderation system. It turns out that there is a false prositive that detects as prohibited a phrase that does not violate the rules, the phrase is this: "A ya esque hay gente que no sabe hacer esas cosas tan "cutres" y mucho menos compartir contenido para comentar en el grupo ....
Ya sabes hacer tema de conversación... No importa de lo que se hable. "
The translation:
"There are people who don't know how to do those "tacky" things, much less share content to comment on in the group...
You already know how to make a topic of conversation... It doesn't matter what you talk about. The point is to exchange opinions and make a topic... So I have achieved it..."
I hope they fix that bug because that phrase with which I want to create a song does not violate the rules.
Since Suno uses a ChatGPT-like interface for lyrics, you can feed your lyrics into ChatGPT and ask what might be causing an issue.
I asked, and it suggested that "cutres" might be flagged accidentally if it's being interpreted as being used in a derogative way?
Either way, it may be a false-positive, as I went to test what word might be problematic, but I was able to generate the original text without any changes.
@wet minnow worked perfectly. Much appreciated.
Yo, that looks really cool 🎉 Looks well formatted on PC, I'm guessing the same for your setup on mobile 😄
Good visual attachments work so well for songs in my experience. It helps add that picture in your head that most people think about when jamming a song they like. So, you will see on my page that my top songs all have custom pictures that don't use the "girly girl" gimmick.
Hi, I encountered a similar problem while helping someone who didn’t understand why Suno was refusing to generate their request.
After extensive research into blacklisted words according to the prevailing standards, nothing seemed suspicious!
Then, by changing just one letter in the name of a software brand mentioned, although it was open source, everything fell into place!
What was the name of the brand?
Guys, how can I continue the song without changing the style?
G.I.M.P. became GuIMP ...
oh, well, i can see how that might be a problem
use the extend feature, leave the style box blank, and pray.
prayer doesn't work.
I'll try now) but maybe there is a tag?
There is no tag
you just leave the style box the same as the original, or you leave it blank. You will get some results that are the same, some that are different, you just gotta keep rolling the dice until you like what comes out
are you from the team?
Nope
Thanks for the answers.
maybe the team will say something else? although I think you're right.
You have a better chance of randomly rolling what you want than the team speaking here
(The devs dont give tech advice in this channel)
This is already the 8th attempt. And he sings something completely different. + Then they need to be glued together. And these are two different songs.
You can glue them together within suno, on all extensions there is a "get whole song" option in the Create drop down menu. That will glue the sections together
The time of the extension point matters as well. The AI typically analyzes the song up to 2 minutes (in 3.5, one minute for 3.0) prior to the extension time.
For example: if you Extend at 00:01, you can make a completely different style because the AI would.have essentially no memory (which I did a few times when I wanted an abrupt change without [Instant switch], [Break] or other similar tags... mainly because I lwarned about those from the community after I was done...).
An issus some people have is that they may lose the vocals used in a verses because of this, especially when they use an instrumental interlude
Does [Instant Switch] actually do anything?
I haven't tried it personally but it looked like it did wheb the person who told me about it showed their link. Let me dig for that post
1,24
it's always hard to tell since they may have used a bunch of extensions to change the styles and then altered the lyrics page so you cant see if they did
- The convo thread that started that discussion #💬┃general-chat message
- SirBitesALot demonstrating one of their examples which has [Instant switch] present (Mr.Tree also said earlier they used that tag too but does not know if it is just a placebo effect)
#💬┃general-chat message
Thanks I'll check it out ❤️
“Gnu Image Manipulation Program” — unfortunately the acronym is a homonym for… a whole other thing. Pulp Fiction comes to mind.
But a Very good demonstration of the fact that the auto moderation service doesn’t care about context, as much as it cares about specific words.
I’m sure that context does matter to some degree, however .
@charred merlin hey bro. how can I continue the song without changing the style?
Use the extend feature and don’t enter in any other style prompts, or as few as possible, or only lyrics
He still changes the song. He starts singing a little differently.
is downloading wav not working rn?
Since we're in the early days of generative AI, music gen isn't likely to be 1:1 for expectation vs reality every time -- choosing an effective place to extend from can largely impact the resulting clip, but unless intentional, distinguishable changes are made to the style prompting, I'd posit that it's usually incredibly difficult to tell where the extensions were -- that's why I enjoy seeing the original .mp4s that show where the cuts were made for each extension clip, they get hidden after someone regenerates the video.
Either way, if the vocals aren't consistent enough between extensions, and you've exhausted all efforts to get anything sound remotely close to the original vocals, it might be easier to have Suno generate an instrumental track for you to overlay vocals onto afterward?
Thank you very much. I will try.
All seems good on my end -- any idea if it's all songs are seemingly impacted on your end, or just a select number?
And just to check, do you see the "download .wav" option, or is it missing? I know Lumine tackled a similar issue recently
What do we know about using punctuation to continue an instrumental? Would there be ways to use it to form 'words' for the instrumental section to play?
A guitar rhythym that is classically named gallop based on the repetitive aspect (e.g. Dum-di-di-dum-di-di ..) on a single chord for a measure, etc.
A good example everyone may know is from "Europe - The Final Count Down". The rhythym guitar in this song is a "galloping guitar". It is popularized by bands like Iron Maiden (e.g. Run to the Hills, The Trooper), and other 80s Metal Bands.
The only way to reproduce a particular rhythym would be to use the Upload feature.
Ok - I'll try that. Thanks.
You could also try using [chugging guitar] as noted in suno.wiki
As for me, I've had luck always using "staccato". Use [staccato guitar] or just [staccato], or [stacatto] [introduction guitar] . It means repeating a single note, and done abruptly, but suno will use the term to mean chugging a single chord, as well.
Its popular in opera and broadway tunes. So, as an example, A Modern Major General (circa 1870s) is sung in staccato. The defining style of most all rap is staccato.
all songs today. not even showing me the download.wav option
https://suno.com/song/5b1398cb-d810-470e-998f-d67117ee6a3f. This would be consider a staccato rhythm. https://suno.com/song/fda35b1f-fd25-42b6-b8ba-6b872788b2bb .... or listen to a litte Plug'n'Battery with the staccato guitar at 30 sec at the end.
Also, ive not tried, but two words describing staccato are "punchy" and "rhythmic" which sound like good prompts to try as well.
does anyone know why I can no longer download .wav files? my pro subscription is active
Thank you lots for the helpfull suggestions, moving to chrome fixed this issue of mine
@azure merlin @alpine rivet I've noticed that the .wav download option isn't present for songs based on Audio Uploads -- is that the case for yourselves as well?
Shout out to @hollow hare for having figured this out of the other day
what a lovely feature
this must be a new update because it worked a few days ago
still lets me split the track into stems and download the wav
Nah, there is some other underlying reason causing it. People have mentioned this issue for at least a month but it is only under certain conditions which we did not narrow down to yet
I haven't seen any update that .wav isn't available for uploads, so I'm operating under the assumption this is a matter for #🪲┃bugs-web
not using chrome. maybe i'll try using that
🤷♂️ I don't have the ability to download .wav for tracks stemming from uploads, regardless of browser or if I'm using desktop or mobile browser.
I don't think it's anything you're doing wrong, tbh
Not really, maybe [Spoken Word] and [Announcer] can help. I am trying some more exotic entries and I have to say the way this started off was surprising:
(Didn't work that well on its alt track https://suno.com/song/4d31ca1d-5af5-4286-ad84-1e65ca11f281)
Actually do not know if it is a usable tag, I was testing it out. We still do not know what the AI knows.
For example, "MySpace" is an actual style that works (and gives weird results). "Girl group" was a style since v1 but I do not think boy band/group worked, but been a while since I tried that
If you are planning to have a talk at the start or end, better to use something like Live Studio, Live show, or some similar tag to that (I cannot recall what they were exactly but you should be able to find some on Suno)
Best to make part 1 a simple talk then extend into the music you want, rather than attempt a one-shot
Well, adding "Live" is more for having an audience so maybe not the best idea to use that (see https://suno.com/song/cfb59ed3-f7c8-4948-983c-0000ae7d4ecf)
Probably my best generation from this test series but still not what I was looking for - https://suno.com/song/f2d6438a-4667-471f-a85e-867788d88b15
One message removed from a suspended account.
One message removed from a suspended account.
How much is the fish? 'bout tree fiddy.
Has anyone else had an issue where they can't download WAVs of extended songs when extending from uploaded audio?
Earlier this week when I did this, the extended version of the uploaded file wasn't tagged as being an upload, and a WAV download could be generated. Now it only lets me download an MP3.
how to make the background music clearer?
@brazen herald @charred merlin As a test, and premium user, I downloaded a wav file of a part 2 extension of a restricted RECORDED voice audio upload. So it does seem permissable in those circumstances.
This was not, however an UPLOADED audio with voice. I'll go find one or test with one and wait to be privatized first.
What's going on with having Part 2, and FULL song on the same line. Am I having brain fog? I don't recall this as normal. Scrolling through, my older extensions only reference the Part #.
Any brand names, sounding like brand names, possible artists or celebrity names, trademarked slogans like "shake it up", "fearless" , "lover fest", Disney, Coke, Jack Sparrow, Qtip, Kleenex
@buoyant gull Also, is the song inherently gender bashing, high or strong sexual innuendo, referencing bullying or torture-like actions, not even any violent acts but acts of cruelty. I'm not listing these as restricted items but saying that lyrics , as I last understood, are also evaluated for implied content, not just specific words.
So if your song goes in that subjective direction, it could be an issue..... or of no consequence.
The list includes an open-ended wording of "if objectionable"
So these concepts might also be reviewed by AI at the time of upload in order to set a restriction or not, but honestly, Ive read songs lyrics here that are extremely torturous in nature, so I'm not sure :
unlawful, harmful, threatening, abusive, harassing, tortious, excessively violent, defamatory, vulgar, obscene, pornographic, libelous, invasive of another's privacy, hateful, discriminatory, or otherwise objectionable; or (vii) in the sole judgment of Suno, is objectionable or which restricts or inhibits any other person from using or enjoying the Service, or which may expose Suno or its users to any harm or liability of any type;
You can email support@suno.com if you wish.
Can I change the video style somehow? Like upload a custom image or set a background colour?
Ok. Tried to post in bug support, but Woodhouse doesn't like that.
One message removed from a suspended account.
Is someone here from Suno who can contact me asap?
Can I extract the lyrics and timestamps to create a video in a different workflow outside suno?
I'm at a loss as I've not researched that. That's a good question when contacting support@suno.com. Just start there. I've not read where someone has followed up by asking support. So that would be a great start and let us know
The initial interest is to see why it was flagged. So that would be a direct question to ask support. The response would then likely be to change that specific content in order to get approval -- but "approval" would simply be that the revised song does not get censored by AI at the next try.
Discord isn’t quite an official support channel at this point, so you might need to contact support@suno.com
Hate email wan to speak to ter person it self. And need support now not yesterday or tomorrow
What's your assessment of your song based on the "censor " information I listed? Perhaps make your own initial judgment , tweak the lyrics for 1 or 2 new, original, first run gens and see if anything changes with no restrictions.
The easiest way to do this is by mastering the tracks — a reliable & free AI auto mastering service is available at https://www.bandlab.com/mastering
I understand the preference, however, Suno handles their first-contact support requests through email, whether it’s support@suno.com or billing@suno.com.
If you need them to get in touch with you NOW, it might be best to send the email ASAP so no more time is wasted 🙏
I'm just a member in here, but will ask. When you say "email" is that messages in Discord? Or emails to your personal email address?
Not sure what you mean by “restricted” — just that “public” isn’t toggled on?
Yes. Correct. Well. That restriction message. Lol. Ok hold on. I'll get it.
That one: Visibility - Audio uploads with vocals (and related songs) cannot be made public at this time
you mean after generating the song , we do mastering. what if i want to use my voice through third party app like kits ai? shall i master the song first,then separate the stem , input and combine my voice.
am i correct
I gotcha — restricted from being made public, but otherwise it’s completely the same.
I’m fairly certain the issue with downloading .wavs is when the song stems from an audio upload.

You’re becoming an audio engineer 😄
That absolutely sounds like a potential productive workflow. I like to run songs through mastering before splitting stems.
Keep in mind that you can reupload 120s of content back into Suno, so you can make 2m of a song then get Suno to improvise on it, continuing the improved sound quality in most cases 💪
Suno doesn't have a means of marking uploaded vocal audios, darn it. So I wouldn't know where to find one in my 1000 page library. Lol. I'll have to remember to title those specifically "vocal-music audio upload" next time.
I know your Ben-Folds-5 uses a tablet version of Discord, so it might not display?
Oh. Hmm... I have to remember to turn my phone sideways.
lmao -- what did that Daniel guy say to get insta-banned?
Even though one way it thinks it's a phone, sideways it thinks it's a tablet.
um... supposely after adding my vocal and combine using third party app like audiocity , the sound quality shall be good. Do you still suggest improving the quality via suno?
Best of both words, when there's features that only work on one or the other
Demanding in a threatening manner?
the problem with suno is the intro is short . i do think of putting it into udio to get the proper intro.
By "intro" , is that with prompting for an introduction? I often forget to add [introduction] tag to my lyrics.
I suggested using a mastering tool to improve the sound quality, Suno isn't the stage where you improve the quality, but create more content -- using a mastered audio upload input, you might eliminate a middleman by providing a "tidied" initial input for Suno to extend from
It looks like he left after the message that got auto-deleted for containing profanity.
I did not mention that that screenshot is of an extension of a visibility restricted song due to an vocal recording audio. So maybe those don't get marked but non-vocals audios will get the "upload" mark. I'm just asking to avoid the hassle of research and review. Lol.
Oh. And did you see this before? Where Part and FULL Song are tagged together...
yeah but still the intro is unstable. like if you want a 20sec intro, it will give you 5-10 secs even you write [intro{
Self-ban with less steps
thanks bro
Yeah -- that's a new one indicating that it's been stitched together previously, I genuinely love having the extension part alongside whether it's stemmed off of a full song 
But if those tags hide/replace the fact that it's stemmed from an audio upload, that's unpleasant.
Funky. But OK. I get it. Helpful and yet...... as you say.
We need a blue dot or something to indicate our restricted songs in the list. I'm starting to want something like that.
Blue dot = uploaded
Yellow = uploaded w/audio or recorded
Pink dot = visible restricted
Red dot = content restricted
Yea. Yea... i know... suggestion forum.
Good ideas wherever they are -- in your defense, it's not like the suggestions are being dumped here out of nowhere, they're following a related conversation.
Definitely worth chucking in there for the purpose of getting extra eyes on it, but I hope I didn't swat enough hands away to make tech-chat a "cool idea free zone" 🤣
i'm liking the idea of the replace section, but the ai just says 'yeah nah, im doing my own thing, sorry'
is the site buggy atm for anyone else?
its not even giving me the replace section confirmation page...
anyhoo, thats my only whinge for the day with 9 minutes to go 😛
I know it costs more credits doing it in multiple steps, but almost everyone one of my "serious songs" starts off as an instrumental -- that way I can choose where I want to insert the lyrics through the use of an extension.
This might just be my own confirmation bias, but I find that the vocals are better when extending from an existing track using v3.5, opposed to being created at the same time as the instrumentals.
Even if it's just a specifically timed segment, I find that the more specific of a song you have "in your mind's ear" that you're trying to recreate, the less likely you are to achieve it solely through the use of complex prompting alone.
When using replace selection. Is it replacing in the song or making a new generation with that selection replaced?
Making a new generation -- I don't believe any of the methods are outright destructive of the original content 🙏
since I'm challenging myself to think of destructive features -- edit song details doesn't keep a record of any previous lyrics or images, but that's all she wrote, as far as I know
Hey I found suno can write jokes:
Why did Suno cross the road?
Because the beat dropped on the other side.
Suno can be remarkably sweet at times
Hi guys! 👋
I don't know if this is the right place, although, it's quite technical anyway^^
I was wondering if anyone had ever made instrumentals
Without going through the Suno prompt?
– I mean, by directly putting the prompts, tags and markers,
In the lyrics section?
– I've been having fun with this for some time now, and I find it quite relevant!
If other people have tried it and succeeded, I'm interested in hearing about their experiences.
Yes it's one of the ways we can be creative on our own without prompting suno to give us lyrics, or clicking instrumental.
Click "Custom" and fill in with a variety of prompts, or even symbols.
Leave the style prompt area blank or fill in if you prefer.
Click create and see what we get.
Yes, that's what I'm talking about!
But I go further, write the AI directives outright.
That's how I tested and released some good music...
I like leaving everything blank and just type [Male vocals]. I'll do this in both Custom and Not-Custom modes. And sometimes eith [Male vocals] in the styles box as well, or leave blank.
But I do find seeding the lyrics box with an utterance such as "Mmm", " Ahhhh", "yea", "woop", "Ooo", "woa" really helps to get a random babble-scat (gibberish) vocally hallucinated song.
Are you asking whether it's possible to specify the key, tempo, specific directions for each instrument, etc?
Cool. What's would the AI directives be? And where? I mean, what are directives, prompts in the styles box or the metatags in the lyrics box?
Oh yeah, I haven't tried that kind of method yet!
But it seems pretty cool...
ok how did my suno radio auto-play start playing songs with skibidibi in the lyrics...... and Arabic at that
Yes! That's what I do, I can put here an example, just as a link, to support...
Ahh -- I've added some prompt experiment playlists on related matters, and I'm not personally convinced it's possible to exert this degree of control through prompting alone -- however by uploading a clip with the desired bpm and key (even if you take an existing suno song and make those changes externally, then reupload), that's one way of achieving the effect.
For us who love the hallucinations, I hope they remain in v3.5 once v4.0 comes out. 😄
I appreciate the thought behind the structure, but I haven't personally been shown a method that works relibably.
I notice that in the lyric section you've noted celtic flute, fiddle, bagpipes, synths, bass, drums -- and while it could be argued that there IS a celtic feel in some aspects, it like at 0:42, it seems to have drifted toward "generic pop", as I've noticed that v3.5 tends to without given effective use of the style prompt section, or being used for extensions.
Despite the requests to modify the BPM, from a glance it appears to maintain the same tempo throughout the song.
I have Scottish roots, so I would have loved to hear bagpipes as you'd requested 
I prefer the uilleann pipes myself
So, yes, indeed, it is far from perfect...
I have only just started to really test this method.
There are multiple adjustments, simplifications probably...
But !!! 😂
What I find quite great in return is the possibility,
to draw from it a working basis for a more complex piece,
and in the process stick lyrics to it!
Which I did too^^
Objectively it's a good quality song, but I wonder if the amount of instructions might have impacted the ability for the lyrics to be presented properly, as the only lyrics that seem to be included are ~1:04, I imagine it was intended to be "let the rhythm take you higher", but it wasn't properly delivered.
As a general rule, I've found that metatags don't work as explicit commands to Suno, and tend to be most effective when using 1~3 words, but this is all just personal observations 
I try to hit the Highland Games we have here in July. One locally and 'tother up in the Appalachians. Always loved Celtic music.
love me some celtic fingerstyle guitar
Let see keeping it AI related... what is that instrument that you wipe the inside of a coffee can with an oil rag? Gives that woopy woopy sound. Something like a ciaen?
I'm looking for a prompt that does that
cuíca?
This title is much better, with a lot of work on it, without being technically flawless, it can do it^^
https://suno.com/song/1b66cba6-23de-4a69-89ad-5cb96255ade5