bug: I've made a remix of someone else's song, as shown here. However it does not show the remix'es original song anymore anywhere in the song's page, neither for the version that I wanna publish, nor for any of the attempts I've made. It's just not displaying anywhere. It did when I created the remixes, somehow it got lost (and it's not a case of reuploading, the link is still displayed in the icon as shown here, it's just not displayed anywhere anymore when the song is played).
#🪲┃bugs-web
1 messages · Page 51 of 1
It's like the switch is off internally, but displayed on.
(also, the original song is still up & working, nothing wrong there either)
link to the song: https://suno.com/song/9ffa2bdf-a774-465e-8332-f81677ec96ee
again, THAT song's details page shows the "remix origin" as above, so something is wrong. I've flagged the song for review but there's nowhere to type in what's wrong so I don't even know if that's the procedure lol.
https://suno.com/persona/58365fc1-888f-4675-8aa2-abd6364b34ae
I use the persona function of another account to make a song, but the persona doesn't apply properly
If you press Create with persona, it does not go to the create page
Parece que la versión 5.0 no trabaja demasiado bien. Siempre, en las canciones clásico-operísticas, te lleva a lugares comunes de los que ningún prompt te pueden sacar. Por ejemplo, quieras o no te mete baterias pop en piezas clásicas! Yo trabajo con composiciones mias a las que hago un cover. Desgraciadamente, con la versión 5 siempre hace parecidas intro y muy parecidos cuerpos. Se ha perdido creatividad respecto a versiones anteriores, De momento uso de nuevo la 4.5 aunque pierda las voces avanzadas de la 5.0. Creo que es un tema urgente.
Hi everyone 👋
I'm a Premier subscriber having a frustrating account issue: I have two separate accounts due to different login methods (one SSO-linked to and email under username, and a phone-linked one "with all my songs). Billing is on the phone-linked account as well as all the music.
I emailed support@suno.com over 2 weeks ago (with follow-ups) and haven't heard back yet. Anyone know if there is way to escalate this here or if there is a solution that I can implement, or which channel/mod to ping for help with merging accounts?
Thanks in advance – really appreciate any pointers! 🙏 (Happy to DM privately if needed.)
Hi 👋 When are you guys going to improve sound quality of song generation at prime time? Why sound quality is always bad when there is an evening in Europe? Because you can't handle load when US and Europe are both awake? Are you don't earn enough money to buy more servers? It's starting to get boring, because this is never changing situation across many months. Noisy high frequencies, collapsed stereo field, hissing, drifting tape effect and of course, infamous underwater sound.
Legacy Editor have many bugs today and don't working properly. Why are you releasing untested code directly into production system? Are you serious? You should respect your customers who pay money to you. At least, if you broke something - just tell us, so that we save our time and credits.
Please, fix styles field in Legacy Editor (Extend function). It is limited to 200 chars even for v5.
Using Suno Pro subscription and since yesterday evening I get weird results when generating songs. The vocals are a mix of lyrics and style text and I don't understand why this happens, so I assume it's a bug.
This tend to happen between the following parameter ranges for me:
Weird 6-88% and ideally staying between 6 and just shy of 70, unless wanting the model to get a bit spicy and creative by deviating from common patterns.
Style 0-97%, and in some cases 98% will consistently result in output that doesn't include Style prompt poetry/singing.
Audio can be anything as it only indicates (though somewhat inconsistently if it's a cover of your own cover, or somebody else's song, or if a persona is utilized) how strong the influence from source material and/or persona is to be. When a persona is used, closer to 100% will end up in a song that incorporates elements from the song used to create the persona, but also the influence of the original track. If no persona is used, it seems to strictly relate to the song you are covering. It's all rather confusing, but through experimentation - it at least feels like there's logic to it.
There's also a thing where too high weird and style in combination that isn't anywhere near what is described above. But then my next move is to reduce weird by about 20%, queue a new generation and repeat until I'm close to 20% which is where things starts to become dull and sound like everybody elses songs, then use this as a way to identify the range which result in what I'm after.
Weird set to values below 20, is to me only useful when I want to reign in the craziness of an otherwise good and unique sounding song, as it'll carry the progression of the song over - but also make the play-style and chords sound more natural, or even possible to perform by a human restricted by limited reach and amount of fingers. :)
In short, if you want Suno to even pretend that you have written anything into the Style box, it needs to be in the 90's range, otherwise it's only considered a suggestion.
Also, try migrating Exclude prompt into the Style prompt by adding a - in front of the word, as I seem to get what I don't want when using Excludes.
So; yes.
Hmkay, so I should use values from 0-5 or 89-100% for weirdness and values from 88-100% for style, even though I then get red numbers? I thought the red values should be avoided
But I used values around 15-85% for all three options before and hadn't had such issues. Therefore, I cannot really understand that there is a connection
I'm just getting this message, although I am a Pro subscriber
Hey everyone!
Just a heads up that if you're bumping into issues, the team is aware and looking into it.
Thank you for your patience as we get this resolved.
Thanks David your a good guy
Will a new version be released? I'd be up for it.
So we should not try to generate songs atm?
5.5 lol wb addition
oh no, Just started having fun 😄
same lol
Don't worry, @sand depot will take care of it 😉
I created a few songs, it tells me that they could not be created and that I should get the credits back and I did not receive anything, on the contrary it took more credits and no melody appeared
Did you refresh the page? I got some credits back after refreshing
like 30 time
it tock me like 240 credits
I have an account specific issue if anyone is able to look into it. I believe I've had a feature flag desync and my premium subscription has not had access to all of the applicable tools for the past 5 weeks.
dunno why im seeing that as a pro subscriber.
Could you DM me what's going on and I'll take a look
I sent a dm a few minutes ago
Okay great thanks you. Will reach back shortly and we'll get this resolved
they demand more money now 😂
Thank you, I appreciate it. ❤️
pro no longer enough 🦊 👍🏼
eh they will fix it im sure 🙂
Happy to help!
more detail?:)))
new pro plan, only 1000$ a month create with the newest 5,1 suno 🦄
David, if my credits that were consumed without the songs being delivered don't appear, what do we do? It's not the first time that when I call, it consumes a lot of my credits and doesn't deliver what is requested through bugs, etc.
i am premier subscriber but also not able to create
i`m too premier :)))that doesn't change anything
so its not available for anyone...lol
Did they return your credits?
i dont know - but i hope
dont :))) hope here its fairy tale
:))) you are very lucky that the song turned out so well for you
oh actually it works for me...whoop whoop
:-)))
yes it works perfectly, it produces songs much faster and it seems to give them real voices of established singers, finally it started creating songs
:)))))
they give me suno 65.5
its a new world here
the songs write themselves
There's a new bug. I'm on the web version right now and all the songs I've previously extracted stems for no longer show up when I click "get stems". It lets me extract but acts as if it's the first time.
Yes, the extracted stems are still there if I uncheck the filter.
But the problems are
- I can't tell which songs I've already generated stems for anymore.
- The "see extracted stems" button is gone.
- Because this button is gone, I can no longer download both files at once. Along with the tempo locked versions being completely inaccessible.
I just tested by generating stems for the same track, then refreshing the page. The see extracted stems is there for that song now and works as expected. But all my previous stem generations do not show up under previously extracted tracks.
Actually I had a better result with the vocal going back to 4.5 then I remastered to V5. But in V5 no contralto vocal just a default female, in a million gens. or seems like a million lol
that could be nice if in Studio/editor mode we can copy and paste parts , without the current time limit of the song. Right now we have to replace an existing section, without being able to add new sections and lengthen the song's duration.
If you have a Studio feature request feel free to throw it in here: https://discord.com/channels/1069381916492562582/1420795299701588129
Every time I prompt Suno with version 4.5+ or V5, upload my lyrics, every time I click create they lyrics are mumbled or it says or sings words that aren't even in the lyrics, I keep reported the tracks for not following lyrics or poor structure and a few of them I've got credit back but most go left alone with nothing. I don't care about the credits unless it's a ton of them but damn...every song??
Glad you got the sound you sought. I have some gens that must've taken 10k or more credits, but we are a picky lot of artists aren't we. 😂 I had not thought about swapping versions. Apologies for my oversight.
I get this if I have words that don't belong inside the lyrics section that are un-bracketed. [ ] Which may be the case.
other times when this happens, I find the sliders can influence this. Too high and Suno occasionally adds in things that don't belong of muddies the first couplet or verse.
Yes, every song... happens to me more than I'd like to admit from my incessant perfectionism for getting the sound I want. lol
You could also try to clear everything from the prompts and load another song and generate one or two iterations of that, then return to your original song. It is like giving Suno a fresh palette to work from.
I'll Def try. The sliders I stay out the weirdness at 0% to 5% usually at first, Styles I typically go 100% when not doing a cover track and I prompt for Vocal key, BPM, tuning, with descriptors overall track. Song sections in lyrics are bracketed and any instruction in storage bracket below song section brackets each in own set of brackets. It's like it mumbles the lyrics at start and then clears up or starts of saying lines completely out of text or repeats a line or "skips"like a CD does in the chorus or somewhere half way into the track.
Never had a problem with this before until here recently and always prompted in same format
This honestly should count as a bug. Apparently SUNO only has one style for symphonic metal songs. Legit 10 diferent songs, all identical sounding. Same intro with the female voice humming, then there's a strong metal instrumental, after which there's a soft lyrical moment, then again strong metal. It repeats itself so much I am starting to think that the devs don't actually like symphonic metal or metal for that matter because they are clearly not teaching SUNO diferent methods for these. It even ignores specific prompts of where to do the instrumental, where lyrics and in which order. Like bro, you guys only like pop stuff or what? lol
🐛 Bug Report Detected
It looks like you might be experiencing a bug! Please help us improve by submitting a detailed report.
Pro tip: If you have exact steps to replicate the issue, that will help us fix it sooner! 🚀
Submit Feedback
Need Help?
Use the button below to get assistance from a moderator who can help reproduce and verify the issue.
Thank you for helping us make Suno better! 🎵
Hi, I have a Premier account and Studio playback is not working at all.
This issue has been ongoing since yesterday.
It happened before, but earlier it usually resolved within 5–10 minutes.
Now it persists, which is very frustrating, as I specifically have a subscription for using Suno Studio and currently can’t use it at all.
Environment:
- macOS
- Safari
- Web Studio
Thanks for the help.
Others have noted issue with Safari. Try Chrome or Braven (my choice). Never had those issues.
Thanks for the suggestion! Any idea if this bug will be fixed in a future update?
Yo creo, entre otras cosas, canciones de estilo clásico y le subo yo una referéncia con letra y melodia. Hasta hace poco todo bien. Pero con la última versión ha cambiado todo. En primer lugar, por más que instruyo la aplicación con un prompt donde especifico estilo clásico orquestral, sinfónico, que no pop style y no pop drums los resultados son siempre: nada que ver con el tempo ni el estilo de la melodia insertada, música plana y sin casi arreglos y más importante, con bateria! La versión 4.5 funciona mucho mejor! En segundo lugar, y eso no está ligado a una versión, o la canción termina abruptamente antes del final o aparece ruido que se puede prolongar varios minutos arruinando lo que podia ser una buena versión.
No idea, can't find suno roadmap, release notes, etc.
This! For me.
Da oggi ho problemi a creare scrivendo un testo completo, qualcuno sa perché?
Last couple of days I've been noticing a distinct volume drop off when using a dance genre. The drop starts to become noticeable at the 1/2 way mark in the song. Any suggestions for things to add to the prompt to stop this from happening?
Bug or by design? I can see likes on other's playlists but not my own... I also don't see likes in the Trending Playlists section. I think that is by design. But why can't I see them on my OWN playlist?...
🐛 Bug Report Detected
It looks like you might be experiencing a bug! Please help us improve by submitting a detailed report.
Pro tip: If you have exact steps to replicate the issue, that will help us fix it sooner! 🚀
Submit Feedback
Need Help?
Use the button below to get assistance from a moderator who can help reproduce and verify the issue.
Thank you for helping us make Suno better! 🎵
I've been getting that now and again with my gens. Like can be quite a dramatic shift of volumes in one track. Really annoying.
Ye, I'm plagued by an not entirely insignificant increase in volume after each finished part and if the song ends up being long enough; to the point of distortion.
In other news:
The context window is too small for generating Covers of a song which isn't close to the max length limitation, but result in gens longer than ~5:15 due to Suno slowing down the pacing or jazzing with the result being an extended intro, bridge, outro, new fills, etc.
This happens even without using a Persona or Inspo.
Hence, it cuts early. There's no doubt.
The "7 minute" limit is essentially wasting tokens - and worse; real world and scarce resources, when frustrated users repeatedly hit Generate until something that ends within this limitation is produced.
It's well past time time to think fresh about this aspect of Suno, as to ensure it functions in an as sustainable way as possible.
In hindsight; when contemplating the water and electricity consumption caused by me creating the below mentioned playlist alone, and while ignoring the materials required for components along with the potential environmental and societal (think employment) impact long term if it succeeds, and isn't just a short term cash grab based on speculation - does make me feel a tiny bit unwell.... :x
Just wanted to let you know that I'm not dodging, but that I've been working on a playlist to demonstrate what I mean in favor of stumbling over my own writing. :)
If you reverse your interpretation (You're not the first to be confused by my writing btw, just fyi xD) of Weird, you're more or less there. But also keep in mind that Persona and Inspo is akin to introducing a second song (and up to 4 in the case of Inspo) to the equation and to an extent changes everything. I think I've had an a-ha moment while creating this playlist, but need a bit more time to nail that part.
Alas, this would likely become somewhat evident when experimenting after having established a notion of what to listen for as a result of changing each individual slider while keeping the exact same Style and Lyrics prompts while generating a Cover.
Hopefully I'll be able to get this link posted within the next 12 hours.
Hi there! I can't acces to my account! Anybody had this trouble?
Extend just doesn't work for some reason. It keeps making 2-second clips. Granted, it refunds my credits, but it's still wasting my time and blocking my progress. Also, if I reduce the "keep" section on the left, the "extend" section will exactly match the exposed portion on the right, which means the song isn't extended at all, it is merely replacing the ending. I want the song longer.
Incidentally, I also have a problem with Suno v5 kicking out a 90-second song in the first place. It should have been at least 3 1/2 minutes, which is what previous versions (all the way back to v3.5) were doing consistently.
I agree but its not just to dance songs. Im on grunge music and its doing this halfway too. I think this is probably the same from all genres. It also seems that its going back to like when its was 3.5v. I hope suno pays attention to this as we are saying
I don't know if this is a bug or just stupid, but persona names can't contain the word logic, possibly because a rapper uses that name. It isn't even the entirety of the name I am trying to give the persona. The name just isn't allowed to contain a basic word, at all.
<Insert wholesome gag about not grasping context here> *stares at Woodhouse*
Generated me extensions of songs of 16 and 18s and didn't refund me credits I wrote to sup no response
Can you not use an alternate word that means the same thing?
suno cover doesnt even stick to the audio :/
My end result was even worse, the cover I just made was just psychedelic chaos. Better to save your credits and try again later. Last weekend the songs Suno made were terrible in terms of technical quality, hopefully this weekend won't be as bad.
Error fetching feed_v3 clips
{"detail":"Not found."}
[BUG] Web login fails when using “Sign in with Apple”
Platform/Browser:
Mobile: iOS (Sign in with Apple)
Web: Chrome on Windows
Description:
On mobile, logging in with “Sign in with Apple” works and the subscription is active.
On the web, access is blocked or the wrong account is opened, likely because Apple assigns different email formats (e.g., Apple relay vs. real email).
The issue seems to happen mainly when Apple login is used outside the Apple ecosystem (Mac/Safari), especially on Windows. Apple services work most reliably on Mac, iPhone, and in the Safari environment.
On Windows, the following often happens:
Apple logins rely on browser redirects.
Cookies or redirects are handled differently by Windows/Chrome.
Apple relay emails are not recognized correctly.
The system treats the same user as a “new user.”
If the same login were done on macOS + Safari, the chances would be higher that:
the Apple login is recognized correctly,
the relay email is properly linked,
no second “shadow account” is created.
Steps to Reproduce:
Log in on mobile using “Sign in with Apple” → everything works.
Try to log in on the web (Windows/Chrome) using Apple or another method.
Access is denied or a different account is opened.
The Apple relay email is likely handled differently than the real email.
Suspected Cause:
Depending on settings, Apple provides either a real email or a relay address, which can create multiple internal accounts—especially when combined with Windows browsers.
I would really appreciate help fixing this so I can access my account again.
#BugThread
Not sure if this is a bug or a glitch but playlists are no longer showing the number of likes (and there's no way to like someones playlist either)
(actually if you click into someones list you can still like it, just not on their main page)
Why does this keep appearing every time I try to create a song?
I am using an external program to do a final master of my track. Now, more often than not, when I upload my final remastered wav to merge it with the "old" version in studio, it won't let me upload at the first try because it claims to have detected "existing work". When I try again, it accepts.
I tried to create a song that started in v2 and changed versions with each verse until it ended in v5, but I can't put the parts together because when I click "Get Full Song" this message appears
Possibly, but that is hardly the issue. Common words being blocked because some artist uses them is an issue. "Offbeat" is the name used by a British rapper. Should a common music term be blocked as well? An artist using a single, ordinary, word as their name being allowed to block the use of that word in, not even as, a persona name is far too restrictive.
OK -- I think the issue was my lyrics were too long. Ironically the solution was to shorten the lyrics in the parts that made them too long, then extend again (no-op extend with [outro] tag), then "get full song" worked lol.
It would be nice if Suno had better errors in general. Instead of just "an error occurred", it could say: "the combined length of your lyrics was too long"
Or when generating stems: "Your song is too long to generate stems", instead of just showing them stuck in generating forever
I had the word "fuchsia" blocked in a persona name once because it contained "sia" in the name, though given all the litigation against them I can understand why they want to be restrictive.
A simple solution would be to allow the names but bar them from being made public.
Even if it's not public doesn't URL linking still work?
Also they may have technical reasons for doing this (like the model using the persona name internally for generation)
This particular bug where artist names are picked up inside other words for persona names/descriptions was fixed a while back btw.
Ah cool, good to know (so I don't need to avoid "fuchsia" in persona names anymore).
Outside of Suno I happen to work on a project called "fuchsia" so I make a lot of songs referencing "fuchsia" (and weirdly it never had a problem with that word in lyrics)
I think you had the same issue I did, your lyrics when combined are likely too long
that's something that could use a better error, although the issue happens once the gen has started so providing a clear error message to the user may be difficult at that point
I did send feedback from my account about that error
No, the lyrics are very short lol
thanks
No problem, also you should send feedback too as it may be a slightly different issue. If you're wondering how I fixed it, I removed the lyrics from the previous clips, then did another extend, then did "get full song" on the other extend
Also I looked at the "network" tab on Chrome to see if I could get more details, but that also just had a generic error status, although I was able to get the ID of the song it was trying to generate which I'm assuming will be useful for the Suno team to look at.
Also another "weird" thing in my case is I used a 3.5 song for "inspiration" which may have screwed up some metadata when it was doing the gen
Ok I have tried different versions and everything, simple prompts, complex prompts, lyrics are simple and every single song I try to make doesn't follow lyrics at all and adds its own. I'm getting tired of wasting credits trying to get it to just sing the lyrics correctly. This is getting old. I understand this is AI and Im not new to AI having a degree in it but simple prompting and simple worded lyrics are not complicated or hard for any AI to return. It would be like me voice chatting with ChatGPT and it coming to me saying words backwards or in another language for no reason at all. Just did 3 takes of the same song, same prompt, not changing anything and everyone messed up lyrics at very start or before intro was over with. Need this reviewed because this isn't something that has happened all along, I've been here for quite awhile and never had this issue until the past few weeks, like I said, I understand AI is going to get things wrong sometimes but every damn time is to much.
Just tried it again and same thing. What is going on. I have weirdness slider all the way down, and it will say this one line that isn't even in the lyrics and just repeat it throughout the entire track once it says it the first time. It may get the first verse right or part of it right but after that, it's over. I shouldn't have to waste credits on something that can't give lyrics back as instructed and simply put. Song sections are in brackets with no bracketed instructions within lyrics. All instruction is in song prompt itself. As mentioned I understand AI may mess up, but not simple lyrics. Getting style wrong or not what I wanted, understood, maybe I didn't prompt correctly, so my fault, but putting my lyrics into the lyric section and receiving lyrics vocalized in track that aren't even in what I gave the AI to say is uncalled for and unnecessary. Shouldn't be dealing with this especially in V4.5+ and V5
I subscribed to Suno Pro during Black Friday. Now I can log into the account, but it shows my account as a free account.
SUGGESTIONS: please for us to upgrade our personnas to new v5 let us pick which song to pick the voice from. I created under 4.0 but the voice doesn't transfer to well because some songs sampled didn't have the good voice I wanted from the original created persona but got better and thats the voice I wanted.
Canvas is so badly coded....
Hello Suno Support Team,
I’m following up again, as I haven’t received any response to my previous messages yet.
I’m currently unable to log into my Suno account using any method (phone number, Google login, etc.).
I normally log in with my Swiss phone number -------.
Mario --- - ---
I am a paid Suno user (not a free account) and I’m concerned about losing access to my existing songs and projects.
I’ve attached a screenshot showing my page.
Could you please manually check my account and let me know how I can regain access, or what steps I should take next?
Thank you very much for your support.
Best regards
Hi Mario. Sorry to hear about your issues, but keep in mind that the actual support crew aren't on this discord, so please don't display any private details like your full name or phone number on this server. If you've already submitted a support ticket you can ask for a follow up in #archived-ticket-help
is my dev console the problem here? Or AM i missing something ?
or is it really canva
I really am only trying to help if dev option is the issue il turn it off
Just let me know something plaese
Hi, I'm constantly getting robotic vocals and audio artifacts (clicks, static) in my generations. I have spent over 3000 credits on a single song trying to get a clean result, but the issue persists. It's really frustrating to waste so many credits for unusable audio. Is there a fix for this?
Thank you Poltergasm... where else but here can I report a problem...?
Hello i see in ticket only in 6 hourse you can ask for help, i dont what happen i have PRO, showing update payment, +-1500 credits gone somewhere.... when i buyed i got 2650 and now 1195.... @sand depot
or who can answer from TEAM
i cant send there because didnt see 6 hours cooldown.
Looks like video generation keeps timing out/failing at the moment
Hi there, I don't understand why clicking on "account" in the sidebar redirects to the subscription plans instead of the account settings
Run a hidden character check on your lyrics and style prompt. Use unicode for everything you put in the boxes. Something as simple as an apostrophe from Word can mess everything up.
Does anyone know that the website is not doing well with Safari/Edge?
That notification section keeps blocking the web view, and it cannot be closed.
Only Chrome can browse websites normally.
No idea where else to look for help with this, but for some reason SUNO is letting me create songs, but not listen to any of them? Even old songs I created a while ago won't play, which is less of a bother, but I was in the middle of writing lyrics and wanted to check a general idea of how it would sound so far, I've kind of worked around it by having to download the SUNO app on my phone, where it plays, but it is a bit of a pain to have to listen to it on my phone instead of getting options on my browser as it should be...
Hi there everybody. Anyone else lost the button Re-Create with the latest update? Mine just shows Cover and can't choose anything else. And when i use it to cover a specific track in Studio it covers the whole song
I've been unable to generate more than once at a time, going on 4 days of having this issue with none of my solutions fixing it. It started when I attempted to upgrade my plan to premier and the payment failed, but still gave me the credits and even shows I'm on the current plan. Any help would be appreciated 🙂
And still after trying Version 4.0 and 4.5+ I am still receiving song after song starting off with lyrics that aren't anywhere in my song. I write my lyrics and paste them into the Prompt, with and without song section labels i.e. - [Intro], [Verse 1], etc.. Also with and without instructions per section for example [Intro- Acoustic guitar, finger picked, strummed (whatever)] or [Intro] then below that in seperate brackets [instructions for instrumentation or style] and I've just pasted the lyrics without a space before the lyrics start and only single space between song sections with no labels or instructions at all and still receiving a bunch of nonsense and in the other versions, not just V5. I can't even get a song completed without wasting a ton of credits and time. I have even prompted directly
DO NOT ADD OR CHANGE ANY LYRICS. SING ONLY THE PROVIDED WORDS.
Use lyrics exactly as written. Do not add words, repeat lines, or loop phrases. Begin with a full verse and complete it before any chorus or repetition. No hook repetition until the chorus section.
create mode is repeatedly ignoring my lyrics, and using the lyrics from the next section instead, https://youtu.be/TC-TaGgE_xI
I am an annual subscription member, and I subscribed to the PRO annual membership on November 30, 2025.
However, since December 24, 2025, I have been unable to use all the features of the PRO membership.
I sent two email a long time ago, but support has not responded. If possible, please help forward my request.
Hi mrjam! Sorry to hear about this. Your query would be best left in #archived-ticket-help, as team members often check there and can escalate your issue to the relevant department. Hope you get it sorted!
thank u 🩷
@burnt jackal someone needs to moderate your web chat, you have bullying and harrasment currently in progress
Thanks for letting me know. Which web chat are you referring to sorry?
Thanks a lot for reporting this
yeah there are abunch of kids in there, its not a good look..
Hello,
I wanted to report some bugs that are making Suno almost unusable for me right now.
The main issues I’m having:
- Extend function: It very often ignores the new lyrics I provide in the extend prompt and repeats the verse or chorus from the previous part instead.
- Lyrics delivery: The vocals frequently turn into gibberish (mumbling or phonetic nonsense), even on simple English words. It’s hard to understand what is being sung.
- General stability: These problems happen much more frequently than before. A few weeks ago, extend and basic lyric generation worked reliably. Now it feels like I have to regenerate 5–10 times to get something acceptable, and sometimes nothing works at all.
What frustrates me the most is that these are not advanced features, they are basic functions (extend, correct pronunciation of lyrics). If even these core parts don’t work reliably anymore, what’s left of the tool?
I really don’t understand why the quality seems to be going backwards instead of improving with updates. I’m starting to regret my subscription because it’s very frustrating to spend credits and time on something that used to work well and now barely functions.
Could you please look into it? It would be great if basic stability could be improved again.
Thanks!
Only Chrome browses websites normally? Not true (maybe it's true on Mac...) I use Opera (which does have a Chrome core); pretty sure it's on the Mac. Been using Opera since 2000. Safari (by Apple) and Edge (by Micro$oft) are not the best browsers, for sure.
Seeing a lot of complaints about garbled words and the like...for me, that virtually only happens when using those damn sliders. Even slightly adjusting them tends to not give the intended result (ie, being obedient to the audio source -- to name but one flaw). The default settings look ridiculous (ie, Audio Influence 25% ?!? or Weirdness 50% - seriously?!?) So I just leave them alone. If you aren't getting the results you want without them, you likely will get worse results with them. Stick with prompts, for what those may be worth.
Boys, see how V4.5 reads the "J"s , seems to have a issue there. Im a poet from Portugal and the Js like Viajante are sound like Viahant
Or Michael Hackson?
Hello Support Team,
My dashboard is stuck on infinite loading.
Steps already tested:
-Hard refresh & cache reset
-Incognito mode
-Multiple browsers
-Different networks
The issue persists across all tests, which strongly suggests a backend / API-side issue rather than a local environment problem.
I’ve attached a screenshot showing the timeout error for reference.
Is anyone else having issues with:
Suno not following lyrics? Lately Suno seems to ignore lyrics and does its own thing!
Qulity of sound seems lower. Quality of music/vocals seems to be degraded. Nothing changed on my side prompts or persona wise, but quality and adherence to lyrics seems to have gone off the wall in past few days!
Anyone else with similar issues?
Are your sliders set to default or did you change them?
I was fighting with that most of yesterday
"cover" feature is broken at the minute... it completely disregards the original audio.
has anyone ever heard back from suno support? my first ticket i opened almost a month ago, and no one ever cared to respond. i have one open on this as well, but wonder if its better off to look at their competitors instead. anyone have any better experience with udio or the similar?
i feel sorta dumb paying for suno's top tier subscription, and everything going to /dev/null
Suno app is so bugged
atm suno seems to be eating credits for very bugged songs, broken rhythms, I've even gotten a remix as long as 7 seconds lol (I suppose it didn't count credits for that one though, but the rest was as much trash)
BUG: I think I found the cause for the disappearing "genre overwrite" on some songs. Basically, in the song's details, if you delete it, the box just disappears, and never comes back if you save. But here's the key: it only seems to apply to songs that have been edited (in the built-in editor). I can't replicate it on normally generated songs, nor on on imported ones.
Why does the create persona for a song not even show up as an option?
Creating personas requires a Pro or Premier plan. You also can't create a persona on uploads, and I don't think they work on remasters either
I have the premier plan, the option is just not here. It is frustrating when trying to make an album and we cannot get consistent vocals.
How are you trying to access the make persona option?
even if it's not available, you should still see it grayed out like this
No you're definitely not being dumb! As I said, it should still show the option, but just gray it out if it's not accessible. I've never seen it completely missing like that before
I am definitely premier plan.
One thing I did not try was a different browser. Installing chrome now. Let’s see if that is it.
Same thing.
It shouldn't matter, but is there any difference if you try to access it from the library rather than playlist?
is it happening with every song or just that particular one?
Ok. Now we are getting somewhere.
If I used a persona to create a song I cannot create a persona from that song.
OH. you're absolutely right
I can create a persona using a sing created using a random voice. Sad part is Suno is not doing a very good job keeping the voice consistent.
Yeah it's missing from my tracks made with personas too, not just grayed out
if it's a legacy persona (created on a model before v5) you can convert it to v5 which may help
Not sure where to post this but I am finding the extend feature to be completely unusable now. It used to work well but now every extend comes out sounding all weird - the vocals and instrumentation different from the source track and lots of bits where the music or vocals go weird. I have tried it on a number of different songs, and tweaked different settings to try and make it work but it just doesn't. It's just disappointing to not have a way to change the ending of my songs anymore, I just have to hope the first generation works. It also wastes a lot of credits trying to make it work.
The best I can do is to crop the song and then make a new cover of it and hope it makes a good ending, but then that usually makes changes to the base track I wanted to keep
The issue with consistency started with updating the persona to V5.
Have you tried remastering the track? The conversion feature is still quite new, so I'm sure they're still dabbling with it. It does have a few hiccups (glitches, variance in the voice, etc) I noticed that were resolved after remastering. Obviously your mileage may vary
Hi, I encountered this error on Suno. I uploaded an audio file and received this error, even though the file I uploaded yesterday was successful. All other files also received the same error:
"Can't fetch the uploaded audio.
Please upload a different audio file."
Suno is flagging a lyrics copyright issue on song, which is a public domain devotional I’m using only AI-generated music and vocals, no copyrighted recordings. Any idea how to get this public-domain lyrics claim removed?
i dont even use extend anymore, i try to just get it one go instead because it's such a waste of time and credits
Cannot create a persona from an uploaded track. The option is greyed out. But then when I go to Personas specifcially I can make it there. Stupid.
Also cannot stress enough how unintelligent and creatively stifling the moderation is. It simply does NOT work. I can upload any song in the world to Suno and it WILL go through without moderation. But sometimes I randomly upload a completely 100% original track and it gets moderated.
Just get rid of it. It doesn't even work. It's just music it's just words, no one cares. Advocate freedom. Let liability rest on the shoulders of the users. At least for paying members. It's such a broken and dumb feature that doesn't even slightly work.
Just me where the iOS app doesn’t work?
Log in with the same phone number, in ios, log in with the ID originally paid for, but keep logging in with the weird ID only on web browser. Does anyone know the fixer view?
hellllppp bros
its macbook
This situation has been killing me for a month...😩
You can't do it directly from an upload. You have to use it in a song first.
The persona? You can though. By going to the persona section. That’s why I think it’s goofy that’s it’s greyed out in the work space and library.
There's lots of issues with the code. Trust me.
But my understanding is when you upload you had to use it to make sure that it went through their filters.
Not gonna mention all the anonymous in the code. That's kind of suspicious.
I'm a Suno Pro subscriber, and I've been double-charged twice! The first time was a complete Suno site error, and it's been almost 40 days since I complained by email, and I haven't heard anything back. Then, a week ago, I was double-charged again for a renewal fee that was supposed to be a discount. They charged me both the regular price and the discounted price. They haven't taken any action. How can I get a refund? How can I get these incompetent customer support representatives to do their job?
I have an issue with one song that I originally created back in version 4 and now want to remix in a different style. The generation consistently skips two specific verses and does not follow the instructions at all. I keep generating new versions and discarding them immediately because none of them meets the requirements.
Web version currently has a bug: I keep encountering an error saying that I didn’t wait for the generation to finish, even though I’m sure the generation has already completed.
Logging out and then logging back in fixes the issue, but it’s quite frustrating.
I think I know why this happens now. When generating a song with multiple tracks, it seems to be counted as 12 separate jobs, so the system thinks my job queue is full.
I am having an issue subscribing to the Premier Plan, my details are correct and I authorise the payment, but for some reason on the webpage they keeps saying rejection of the payment.
don't know if this is a bug or not...if i created a cover of someone else's song, and then cover that again,..I lose the traceability of how the song is created. Shouldn't the user song details show the actual song that is covering, not the other user's cover seed. I know the non-user should be seeing the cover seed, but the user covering should be able to follow the chain of intermediate covers from the song details in my opinion....I feel like it used to work that way but doesn't anymore...I don't know if i did enough covers of others to recall if something changed accurately.
Trying to upload original songs and it's saying Uploaded audio matches existing work of art. Please upload a different audio file. But these are original works that have never been uploaded here. Anyone ran into this?
welp its not there https://suno.com/song/e18d9fa6-1d13-4042-bf37-e21ff672c0fc but it was genned last
Since today or maybe since yesterday, the subtil remaster often has a double voice that is time offset, and there are also other strange issues with the beats.
All songs 7:59 long , every try it ends when its time for the Outro, with 4+ min of beeping at the end.
hey, i want to create an instrumental for my song, but every time i choose cover and choose instrumental only i got the song with vocals !!, any help here ?
I uploaded an instrumental and it was given lyrics from other songs.
remove the lyrics from the edit tap, it will work fine, i faced the same issue
i wasnt aaking about the beat itself its perfect lmao
hi there everybody. Anyone else experiencing this? Whenever i select a track, i only get the option to Cover it. Shouldn't there be a replace option as well? And when i press cover with my style of drums for example in this track, the generation is making a cover of the whole song
i already cleared all cache and data from the browser in case it was something from the cache
Lately, I'm finding it difficult to go to the TRASH section. Occasionally a garbage-can icon appears in the page. But, when I click it, the TRASH section is COMPLETELY EMPTY. Aren't deleted songs first put into that folder, so if someone accidentally deleted a song, or someone changes their mind it can be restored?
If this is a bug, please either fix it, or give me a way to access deleted songs over the past few months.
However, if this is a new "undocumented feature", please advise when it was started, and when we were notified. Also, if moving songs to trash instantly deletes then, why would you even need a TRASH section?
are these bots a suno feature? 🤔 pretty sure ive had about 300 follows in less than a week.. i kno a lot of ppl wouldnt usually complain about this kinda thing but i view it as cheating the stats and i am against it, if its suno's doing, stop it, if its not, please fix it.
Selecting the track or clicking on the track region doesn't cut-it. You need to click at beginning of region and drag across the length of the region and 'replace' will be available.
Clicking play on the new Mashup popup (to select a song) closes the popup instead of previewing the song prior to committing
Can you tell me why it won't let me use my voice?
- Persona Style text box does not update when adding or changing it.
- Song Edit Mode - Can't extend the song, hitting the button does nothing.
- Song Edit Mode - Replacing sections or regenerating them do not use the selected Persona or styles for the song.
- Song Edit Mode - Lots of general jank and features not working as you would expect.
It feels like the Suno site was made with an LLM exclusively with all the issues like that I run into.
Thanks, I hate it. https://suno.com/s/LULg8EBEB83LeixV
Just stops halfway with random background noise.
@full nacelle Sorry to hear that. Please note, however, that the new feature is a beta version. Errors may occur. We hope you enjoy using it anyway 🙂
Anyone else running into the issue where attempts to cover a song with slightly different lyrics causes Suno to only make instrumentals? Keeps happening to me over and over again and it's frustrating.
do a search of of the discord or the prompt development thread for...cover with different lyrics (or something along those lines)....that may help you narrow the issue or find others with similar problem
Try with default or at least less extreme slider settings. Your song was created with 0% weirdness, and 100% style/audio influence. Generally this is likely to have unwanted side-effects.
I swear they need to rename things or something. The number of people who try to turn off the creativity of their gens astounds me...
Im having issues using persona audio influence.. it continues to cut my songs off at the final chorus or verse anywhere between 5seconds- 40 seconds left of the song to complete how do we fix this issue
Extend from X time.
The cover + persona + whatever sliders is eating away at the context available for the AI to work with. A one-shot may not be possible at all due to this. If you're consistently ending about the same point this is likely the issue.
You can also try removing almost all [metatags] and rely on the cover itself to provide that information but it's not guaranteed to give back much time.
Yeah It continues to cut off at the exact same time example 4:36 on every song if I remove the cover it shuts off at 4:03 on every song
What are your sliders at?
Style 75 audio 90 plus.. because i want the same voice and thats the only way to get it
Hmm... yeah, that's eating the context available for song length. I repeat you're going to have to extend from one of the best available and see if you can make it end comfortably.
And if I change to legacy persona it changed completely like its not even covering the original sound
Yeah im going to try and see if I can split it at a point I like in 2 different versions extend in different versions break into stems and try to splice in Reaper
@solar grove ok so i tried to turn slider completely to 0% across the board and its still stopping at 4:36
Any use of the sliders, any cover, any persona will result in a reduction of total context available.
When Using Mashup the generations stops at the exact same time, every time.
https://i.imgur.com/Ty7BhIg.png
@solar grove is there a way to turn slider off?
Just use it as default. But I really think you're just going to have to find a good spot before the cutoff and extend from that time manually.
Ok thank you
I’m experiencing insane hallucinations with lyrics in remaster mode, as well as in create. If it were a one-off, I’d ignore it…but it is happening on nearly everything I’m generating today. All of a sudden the language has changed and pronunciation is altered. Any advice?
@open garden report the music as bug in Suno
Hello team, would love to know why my sliders are grey on web chrome, can't use it... Thank you.
Which model do you have selected?
I restarted chrome it works
the links to the workspaces are broken
I've been experiencing my own lyrics being ignored and replaced with other lyrics in weird languages in custom mode for the last 2 hours or so. Tried a couple of song generations, but it just got worse.
Report the music as a bug in suno not following lyrics
Bro, did you just delete all my songs?
Have you tired logging in in an incognito tab?
all my workspaces are broken too @static iron
i can copy and paste the link to it in another tab and play the songs, but it looks like the workspace browser's links are broken
and workspace links don't work in incognito
who let the intern make a commit without ai double checking them?
@static iron delete /v2 from the url and it will work:
if you compare a working link to the broken ones, thats the difference
it should look like this
Hey there! Thanks for flagging this!
Could you walk me through what's happening here? Are you selecting something that directing you there?
Looking forward to figuring this out
any workspace i click on loads the page not found above, tried on two machines:
i had a workspace open in another tab, it didn't have v2 in it, and worked as normal. all new ones i opened had v2 in the url
Thanks again for reporting this. Could you DM me your suno handle (@example_handle) or a song link from your account whern you get a chance?
I tried and went back to front page
https://suno.com/create?wid=1c73438e-6f75-4180-aa53-11b7b2cdbebc
im going to miss my AI songs
I am having that too. Every time I cover a song with the Voice persona the song comes out 5-7 minutes (original song 3:10) and 2 minutes of beeping or gibberish repeating the last chrous or whatever. Is this known with V5 and Voice Persona? Until a few days ago it worked totally fine.
@static iron @twin shell
There's a fix on the way for the workspace issue.
Please check and let us know if it is still an issue?
that resolved it. thank you 🙂
@humble cobalt Suno Bug Reporting and Monitoring
Using a conversational forum (like Discord) for bug reporting, well... it sucks. It is not only difficult to see if someone else has reported the issue already but there is nowhere to see the priority, classification, fix progress, etc.
There are certainly well known issue management systems ($) Jira, etc. I am sure there are probably some open source options as well.
Heck, a google sheet with an 'new bug' dialog would do....
Subject: Corrupted stems, export crashes & lost credits (Pro account)
Hello Suno Support,
I am a Pro subscriber and I am experiencing a serious issue with the stems export feature.
For a many songs I generated, the master file is completely clean, but all individual stems contain noise, clicks, and static artifacts. This happens consistently, regardless of the device, browser, or audio software I use.
Additionally, when I try to re‑export the stems, the page freezes and logs me out, causing me to lose credits without receiving any usable files. I have already lost approximately 1,000 credits due to repeated export crashes.
I do NOT want to regenerate the songs because I want to keep this exact version.
I kindly request:
- Manual re-rendering of the stems for this specific track (without altering the song).
- Restoration/refund of the lost credits caused by the export failures.
- Confirmation that this issue is being investigated, as it affects Pro users.
I can provide the song link and audio examples of the corrupted stems if needed.
Thank you in advance — I appreciate your help.
HI there
Since we had mashup updates today, there is an issue with workspace issue and cover has a big issues.
Lots of stem ticks, noises and inconsistent generation.
I was so looking to working on my music tonight but very dispointed 🙁 help please.
Also, pronounce slurring, skip happening as well.
Is there any reason we can no longer see the number of likes on playlists? @sand depot could you please let us know
Another point, in the TOS it states that if you cover someones track (or they cover yours) then you become joint owners, surely if this is the case then BOTH creators should be able to download the track, @sand depot are there any plans to address this?
Hi Neil 555, thanks for reaching out about this.
When you get a chance, would you mind sending me a DM with a screen recording of what you're bumping into?
Cover DL re: this is quite a bit out of my scope but something I can ask about for sure.
Looking forward to figuring this out.
What on Earth is wrong with SUNO? Nine out of ten times it fumbles the lyrics in the first verse. Lots of pops and clicks. It's broken.
It's completely messed up.
Greetings. I've been trying today to download "Video". As far as I know, when I've done it previously (I haven't for a while), it creates an mp4 with the song and scrolling lyrics. I'm not getting any lyrics. Any help? (I have a PRO account).
(I tried at different times to see if it was server load problem)
Good morning. I have done a few more test and it seems Persona cover is not working properly. If I dont use persona during cover, it seems works fine. But as soon as I add persona and make cover, stem ticks, skip beat, slurring, unecessary extention etc.
Actually testing them again, Cover is overall not working properly.
IOS updates came through last night and I used cover (no persona option via IOS and I couldn't test) it was fine hence I said it was working in my first comment. So maybe fix is on the way for Web?
what am I doing for Suno to, once again, not display the link to the original song, while there is clearly still a link made in the song's details?
Is it when you do a "cover of a cover"? Because that used to work fine.
It should be displaying this
Suno v5 Pro Beta giving 3+ minutes of dead air at the end of generated tracks.
after I crop out the 4 minutes I can't land back in the editor to finish the edit with a fade out. I get this error.
All my songs are now 7:59 long. Startet with some a couple of days ago, but yesterday it was all songs - same thing today. And it's not only the endings of the songs, there's random clitches in the middle of the songs.
It's not a bug. Your remix already submitted when you made it public and it will show the original origin too when you play that song and look at the right side box where lyrics are displayed.
no it's not, that's my point
that's the second time it happens for me
the link isn't visible to anyone but myself in the song's details page
it's displayed properly for my other covers
just not this one and another recent cover
it's not that Suno "forgot" the link to the original, and forgot that it's a cover, because it's in the song details page
In my remix entry, it is visible and correct. I don't know what is wrong with your remix entry 😔
Is the new update our yet? Wasnt it supposed to clean up the sound? I got so much hissing, whitenoise ect in the songs.
You remixed other people's songs or what! Check once.
Use the proper music track from the artist to make it remix.
how else would it be displaying this if I wasn't
this is for this song, it's a bug
I don't know how it happened. Maybe someone will look out for you and resolve your issue. Good luck 🤞
(this sums it up better)
I've been trying to edit a song to replace some lyrics. In the legacy editor, I can hear the replacement clips, but after a while the editor gets really weird, adding extra lyrics and not showing the window with the replacement clips. I have to click on the section to open the window.
And when I try to use Studio, I don't even see a section for the lyrics like in the legacy editor, but either only hear the first second or two of the replacement clip, or not even at all. I'm not sure what I'm doing wrong.
Been trying to make an extension of a song, after 16 attempts, on 13 of them the audio completely craps out after just a few seconds. It returns to normal after a while but makes the extensions completely useless. Starting to burn through credits here trying to finish it
Am I supposed to get captchas for every song I generate?
I've had about a half dozen gens cut off in mid lyric or riff at somewhere between 4:30 and 4:45, I've tried adjusting the [Outro] tag but it just keep cutting off the end of my songs
I've been trying today to download "Video". As far as I know, when I've done it previously (I haven't for a while), it creates an mp4 with the song and scrolling lyrics. I'm not getting any lyrics.
WORKING NOW.
Could you please fix the persona/personality of the AI when using this because some people have absolutely no problem when using this but then I do and I have to generate at least 6+ songs to get the lyrics correct because either the AI will mess up lyrics, which people say that “Suno doesn’t mess up lyrics.” Which is bs…… and or it will cut half of my sign in half and generate it to be like one minute and 30 seconds. Or nine minutes. It’s very frustrating…. and would also be nice if you could expand more on its capabilities I guess of what it knows versus what it doesn’t.
Previously, uploading an audio file via the cover interface worked fine, but today I'm getting the error: "UPLOAD_STATUS_ERROR: Uploaded audio matches existing work of art".
I constantly receive notifications (3 or 5 or 7), but only 2 instead of 3, or 4 instead of 5, or 6 instead of 7 are displayed. If there is only one notification, I don't see anything in the list at all.
My mashup songs sound weird... Sounds like the singers mic has too much gain and got squeezed through a compressor afterwards. Same for some instruments. Sounds like something between distortion and white noise...
Hi Suno Team,
I noticed a decline in Japanese vocal pronunciation quality in the generated songs since January 20. While the arrangement density has improved and sounds great, the clarity and naturalness of the Japanese vocal delivery appear weaker.
Here are two songs with the same Japanese lyrics:
January 20: https://suno.com/s/ubcS7edEPFDKXAhc
January 23: https://suno.com/s/ZfKZ42RqB4Z1h9uw
I hope this feedback helps improve the Japanese vocal synthesis.
Thanks for your support.
has v5 been updated? i'm trying to prompt basic stuff and it's like educating pork
I covered the contest song, then covered my cover as per my normal workflow (I use personas to tune the sound on my music) and the end result doesn't show up as a remix of the contest song... see below
https://suno.com/s/1ANynuzNZsV5g7Xf
Is anyone available to help me locate the mashup feature I've been trying for about 2 days now and am starting to think it's not activated on my account
in Studio, I'm trying to create a new bass track - what I'm getting is a full song: guitar, drums, vocals, everything
The last two days, most of my covers and remasters end with a very annoying hum at the end of the song. And that wasn't in the song that is covered. And it wouldn't be such an issue if it was fully at the end of the song, but it already starts in the outro, so it's difficult to get rid of it. It a bit like what you hear if you short circuited analoog equipment.
This should be bugs-generation
Please tell the AI to skip the breathing sounds.
It's funny, not funny, so not funny is becomes funny.
In your create menu pick a song, and go to the three dots at the end and click on there, see where you do your edits etc and go to the top dropdown menu"Remix/Edit" and from there look in that menu and it should in between Extend and Adjust speed
There is a bug with playlists it seems.
-
remove song from playlist
-
update playlist description
-
removed songs will come back, it also seems like it messing up links \ actions for songs.
So I found a slightly elusive bug on desktop. If you make a Persona abed use it for a few Songs then change the name of the persona, it only updates the persona name under the source song and not any of it's children. Not a massive bug but it can be confusing.
Hey, when i was canceling suno subscription it gave me a 50% off on my next purchase but i didn't get that 50% off😭😭, i want a yearly plan please help me out.
Hi Suno team, I need help with a subscription issue that has completely blocked my account.
Around December 19, I attempted to upgrade my plan from Pro to Premium, but the payment did not go through due to insufficient funds. Even though the payment failed, my Suno account now appears as Premium, but without access to Premium features.
I cannot create songs, export stems, or generate anything — I can only listen to existing songs.
This seems to be a billing/state error. I currently have an active Pro plan that is paid and valid until October 2026, and I just need my account reverted back to Pro so I can continue working.
Payment method is credit/debit card, and the issue happens on both web and mobile app. I’ve contacted support by email but haven’t received a response yet. Thanks.
Does it happen to anyone else that all the songs generated now last 7 minutes 59 seconds even if half of it is final noise or silence? He's been doing it to me since yesterday in practically all the songs.
hello! anyone else here getting a bug where lyrics are generated for instrumental when using the mashup button? the lyrics are about "no words, just silence" lol
i know it's in beta
Between 40 cover and Style and Lyrics gens, it only happened once and it was good from start to end.
Anyone know why all my generations does not tempo match i tell suno to make it for example 170 bpm, when i drag it into daw and try to tempo sync, it says "Sample is 170.535" and i cant seem to get it to sync... PLEASE HELP
every generations is like that
Personas haven't worked with uploads (or covers based on uploads) for quite some time (not a bug)
Can't adjust speed of some of my covers. Adjust Speed is disabled. But I can adjust some of my covers a few days. Someone help me. Is this glitch/bug?
what if you save as new song, then edit the new song? 🤔 does it still count it as a cover? (i don't know because i never use cover)
Uploaded audio can't be speed adjusted. But my problem is I cannot adjust the speed of my new covers anymore.
it could be that it's because it's not the original 🤔 (and not a broken feature, like deliberate)
i think the work around for it would be to get stems, adjust tempo, then write new lyrics on it
i hope we get a better model soon, one that has all the best aspects of the current and previous models, because v5 feels a bit hit or miss. like example, i have absolutely loaded my lyrics section, it's like 4700/5000 characters and i'm getting tracks that are 1:30
Hey guys I have a problem. So my subscription was paused because I didn’t have sufficient balance on my account (suno pro 10$) as soon as I’ve seen that I added funds to my account and went to my stripe to finalize payment. I got an invoice and everything but my account is still on “free plan”. Can somebody please help me?
I am sure others have the horrible heavy bass and distortion problem on the V5 model and extending the song to over 7 mins with the genned song finishing before that but with the mentioned problems with audio. As it isn't on every gen (but is probably more than half of genned songs) then it is clear the model or something else has some problems that do need fixing. Although the V4+ also has problems including similar audio problems and the extending with either silence or noise after the genned song, I do use it if the V5 model is causing issues.
Desktop: Getting a Failure to Fetch API call when trying to save a persona (studio-api.prod.suno.com)
Note: This is happening at the level of creating a song from scratch and attempting persona creation, no cover or upload.
I saw someone's else prompt
‑intro vocal adlibs, ‑vocal humming
Interesting nuance.
Hope that helps 🙂
Helle here !
I've got a problem: one comment of me and all the nested comments are on two songs. The main problem is that comments were published on an unpublished song, so… "kind of half-private". Relational DB problem ? 😅
And sorry, it's late here, I'm going to bed, so I can't give more information before probably 8 hours. XD
So I see Suno is cutting my songs short again and ignoring full verses.
All cover songs completely cut out long before finished, and whole verses are ignored. I'm not paying $30 a month for this. Sorry.
Cannot play audio in suno studios. Been going on for over a month now. I have done everything i can think of from clear browser history, cache etc. log out and back in. All of it.
I can hear my audio when im not in the suno studios app in my account within a browser. But as soon as i add a track and hit play in studios the play bar doesn’t move and there is nothing i can do to get it to play. Wasn’t always this way. Just started happening about 1-1.5 months ago.
I'm literally singing brand new lyrics into SUNO studio and getting this message "Uploaded audio contains copyrighted lyrics.
Please upload a different audio file."?
Regarding this problem, anyone here also cannot adjust the speed of their covers?
Just following up — this issue is still blocking my workflow.
Hey there, I'm really sorry to hear about your situation. Have you raised a ticket by emailing support? #support-guide
Lately have been getting songs created with an additional 2-3 minutes of dead space/low extended tones at the end of the song. I have reported each one as they come up, but it's a little disconcerting at how many are now happening (about 35-40% of my generations in the last couple days!)
I can’t open my notifications
It keeps giving me this message and I can only access half my songs on homepage without having to close the app and then reopen. Also, trying to comment on others songs isn’t possible alot of the time, it seems to time out. Same when I am trying to post a hook.
I assume those last two issues mean you need to add more servers? I am on iOS and have the latest update.
Why did my CDN links all change to removeai.ai ?
is your app/device up to date? That can be the problem sometimes
1: Consistently cuts off song creation 1-8 lines of lyrics shy of the end. 2: Covers consistently more mellow vocals/instruments to vastly different when using the exact same style structure. I've wasted 300 credits trying to fix 1 song because the software is too dumb to fix simple mistakes. Really, really struggles with rock/metal songs.
Also, website feedback can NEVER submit.
hi, anyone having issues generating lyrics on Suno ?
Hm
i sent reports on ios bug, but maybe it belongs here since its in discord itself when i try to put a playlist in showcase playlist, it auto deletes and my playlist is public with no curse words or offensive art. This is the 4th time now. Please emote on my message so i know you looked at it so i wont keep sending reports cuz idk if u see this. I know there’s a lot of people. Thanks
it was due to the prompt :/
another thing, i have updated a persona (which was making great songs) with the last V5 promising better quality. but the voice has changed so hard, is it possible to reverse this ?
Hi,
I’m having an issue with keyboard shortcuts in Suno Studio.
On Windows, the Mute and Solo shortcuts (Ctrl + Shift + M and Ctrl + Shift + S) are not working in web browsers like Chrome and Firefox. Instead, the browser’s own functions are triggered, or nothing happens.
I’ve tried this in multiple browsers, but the problem happens every time.
Is there a way to fix this, or are there alternative shortcuts I can use?
I’d appreciate your help as soon as possible. Thank you.
I was automatically logged out and I can't log in on any browser, neither on my phone nor on my PC, because the login button isn't working.
Just a heads up to everyone that's having trouble logging in. The team is looking into this now. Will update as soon as this is resolved.
Thank you for your patience and understanding.
Should be good to go!
I ran into a bug several days ago that generates songs at double length and breaking into feedback or noise, for 7:59. its a known issue and support said theyre monitoring it and working on it. I dont want to waste my credits until resolved (it wasnt as of yesterday) but my credits run out as of the 1st and i also dont want to lose them. Any knowledge on how this is going with a fix? Also i ran into an issue with personas that ive developed that have a flatter delivery and they previously worked during the chorus where they wouldn't uplift to an absurd amount and now they seem to ignore prompting and the persona model. not sure if theyre releated or two different issues.
same problem. Sometimes i get 100% of generations with this bug.
Original live recorded audio being falsely flagged and discarded, breaking service utility
Hope this is the right channel to post in. I'm pulling my hair out. My process involves recording improvised scat/beatbox, and I normally use the the built-in Record feature in the web app or ocassionally on Android (less often because the app is hot garbage, sorry). Starting yesterday, more than half of my live improvised recordings are getting flagged with false positives ("Uploaded audio matches existing work or art."). When I record in the interface directly, as I typically do (less interruptive to flow than recording externally and uploading), if it's flagged, the recording is just...gone. As it's improvised, I won't get that take back. If I try to re-create it, odds are fair it will just get flagged and trashed unceremoniously again, amplifying my frustration and souring me on the tune in most cases.
I did a search for "false positives" on this Discord but only saw things about prompt words triggering false positives, but I'm all ears if this is an open topic somewhere and I missed it.
I can just record in Audacity, but that's a massive cut to the value of this product for me, and presumably won't be as simple as "Record, then upload," as the same audio will surely trigger just the same, leaving me with the work of trying to pitch- or time-shift it to overcome the trigger...and potentially just more frustration if that doesn't work. And if it does, well I'll have compromised the audio to get it to take.
I did open a support ticket via email yesterday, but from what I've read it can take weeks to get anywhere. In any case, I really don't want to have to cancel and press for a refund, because it's not like I have an alternative product to take Suno's place. But it would help a lot to know this is something the team is aware of and working on.
Why I got getting this error every different audio uploadings?
I'm also getting this while trying to import tracks from EASUS Ai which uses the SUNO API. I think its a fingerprinting issue on SUNO backend or at least according to ChatGPT
Beware: there is NO official Suno API
🚨 Suno does not have a public API. Any claims otherwise are unauthorized.
there may not be an official API, but i have read there is a workaround that some companies have used. Im familiar enough with the SUNO modesl that the results are very too similar, differences being EASEUS has less control and tighter character count. They also have the same versioning of the models. 4.5 4.5 pro and 5.
Beware: there is NO official Suno API
🚨 Suno does not have a public API. Any claims otherwise are unauthorized.
Yeah I don't know about any API; I'm a Suno Premier subscriber and just use the service directly...or at least I did until it started accusing me of stealing and burning my work.
Windows Desktop Web – Suno Studio Bug
When extending a fade beyond the currently selected section of a music clip, the fade handle becomes invisible. The handle is still interactive, but it cannot be seen and can only be located by hovering the cursor over the area until it appears or becomes clickable.
Ideally, when i click on the title of that track, it should become visible in my opinion, but right now it is invisible.
I still want to be able to drag that fade as far as i want to because it's a useful function for editing music.
Sample of a false positive. If Suno wants to make people believe it's not just a slop machine, they should care that they've effectively broken functionality for people using it in conjunction with real audio inputs. This seems to have started with the same update that brought the "mashup" feature which, frankly, has no apparent utility to me directly. I do have the impression it brought a big improvement to stem detection, which I don't use directly in terms of generating stems from songs for external use, but I believe it has been better at identifying different voices/instruments for assignment in my vocalized inputs.
I can't describe what is attached here because the words necessary won't pass the Discord filters.
firefox 14.7.0: for any style page, if i try to scroll down requiring the page to load more songs, i get the "this page is slowing down firefox" message, prompting me to stop the page. This only started happening w/in the last 3 or 4 days if I recall......There is also general laggyness that shows up sometimes making working on a song impossible like 10 seconds or longer where it buffers my actions then bam spits them all out...dangerous when its doing clicking instead of typing...and its the only website doing it among the open ones..this second issue (general laggyness specific to this website) has been ongoing for a month or so, and can be somewhat alleviated by deleting cookies and data till it eventually shows up again. Tried to submit via feedback on website, but the submit button was blocked/greyed out even though all required fields were populated
I am trying to make a payment for the year, and for the last 2 weeks, I have not been able to. I have contacted my bank, and they said they are authorising the payment, no issue, but my payment has been declined every time on the website. Having searched Google, I can see that so many people are having the same issue. Can anyone advise? i have emailed billing 2 weeks ago and had no reply or acknowledgement.
I have this weird bug where sometimes, when I generate a song, it'll give me an 8 minute song, however, only the first 3 and a half minutes is the song, the rest is sometimes idle static sound.
[SHELL] - Suno Studio - Recording vocal, then using cover, then recording vocal, causes match/ copyright conflicts
Repro:
- Suno Studio
- create track, set record
- Sing lyrics to the track
- use
COVER - record lyrics over track again
Expected:
Recording is completed, with lyrics displayed in side panel
Actual:
Copyright flag, or image
Repro level:
unclear, but high freq
Severe outage now ? cannot generate songs anymore for 15 minutes
I had that too for a bit, refreshing worked for me
Ah that'll do it. High demand.
Hey everyone!
If you're bumping into any issues generating, please know the team is currently looking into this. Will update once we get this resolved!
Thank you for your patience and understanding!
Quality dropped to the level of version 3.
The 7:59 bug is really annoying me. Any updates on when a fix will come out?
I just reported a bug for same 7:59 issue.
Please put back the option to create personas from recorded audio. I need Suno to sing with my voice! At least please enable it as a one-off for me! I understand the reasoning, but I should be able to request a manual review or something, it will be very clear that a recording of me saying and singing random words into the mic isn't copyrighted.
I have the same issue.
Suno just keep cutting my songs in half, totally ignoring 70-50% of the lycris
Same thing happened to me too
Reporting bugs on quite a few songs as Suno keeps generating with 1/2-second skips/gaps.
So I’ve been sitting here for quite a bit and I even reloaded the page and I’m just trying to do a fade out and it’s not working.
That happened to me as well. And I hated it. I actually had to shorten my song. Or what you could do as you could make two different shorts of that song and then mash them up and then put the full lyrics in. Maybe it might unstick itself.
The Suno overall seems to be extremely buggy/broken last few days.
The devs already said they are looking into it tho, so I would just wait it out, until it's fixed
Alr. I also did try editing it in the new editor and it’s still fade isn’t working. I kind of hope that they fix it soon because I’d like my end to be faded out because it sounds like garbage. I mean, I could just crop it, but I don’t know how well that would sound.
And my song also has many spots there the AI messed up simple words and it sounds bad. @humble cobalt
how to reverse a persona update, i would like to get back to my first version which was better than the "upgraded" one
I agree
Yeah, it bad.
Ty for actually replying to our worries. I’m also having problem where my song caught halfway through it and I’m having a glitch where my song sometimes will generate only 24 minutes in 53 seconds. And also fade out or fade in doesn’t work in the editor.
Also on top of that, replacing the lyrics and song editor is also not working. It’s just completely ignoring the new lyrics I put. @sand depot
The play back feature is broke on web page some song play some won’t and some won’t load I’ve reset I’ve cleared cookies different devices even still won’t make a difference
you mean you don't want to have keep choosing legacy mode when loading it defaults to v5?
I had a message 'exr to my persona. Update to v5
I had to choose a 30sec part
But now the voice doesn't sing like before
When will clipping the end of songs be fixed since extending doesnt work? All extends end in 4 minutes not following either style or lyrics.
Someone on Reddit reported they cannot export to MIDI anymore, that the option is missing. I tried it myself with 2 songs. When clicking the Get Stems/MIDI option, it brings up the option for All Detected Stems or Vocals + Instrumental and I choose All Detected. I cannot change the option from Tempo-Locked WAV, so I can't even choose MP3 or normal WAV let alone MIDI.
So i tried using a prompt template i've used in other tracks that are typically 3:30 - 4:00 in length and yesterday it started giving me 6-7:59 gens. Looks like the old structuring bug is back..
SO 7:59 BUG is still active. So is the "uploaded Audio matches existing work of art" false flags. I haven't been able to use SUNO for days now. Any estimate time frame for a resolution on these? Or just the team is working on things?
🐛 Bug Report Detected
It looks like you might be experiencing a bug! Please help us improve by submitting a detailed report.
Pro tip: If you have exact steps to replicate the issue, that will help us fix it sooner! 🚀
Submit Feedback
Need Help?
Use the button below to get assistance from a moderator who can help reproduce and verify the issue.
Thank you for helping us make Suno better! 🎵
When I choose to do vocals in Suno Studio, for some reason it does it with the entire instrumental, although this has never happened before
Estou enfrentando problemas na cobrança de uma assinatura anual, sendo que a minha mensal já foi paga e essa cobrança anual não me deixa em paz!! Não sai e não liberam meus créditos já pagos do mês!!! Alguém da comunidade consegue me ajudar, já entrei em contato com o pessoal do billing@suno.com e com o pessoal do suporte e até agora não recebi nenhum retorno deles, só uma mensagem automática dizendo que estão trabalhando para atender toda demanda de mensagens!!! Por favor alguém!!!
Please remove your screenshot for your own safety (don't publicly post your address/email address). The accounts team is not on discord so we can't do anything with that information anyway. Please mention your enquiry in #archived-ticket-help without disclosing any personal information. Thanks!
Obrigado, não sabia, novo em comunidade!!! 😂
No problem, and welcome!
Is the trending section acting up for anyone else? I tried to change the language and it just went blank on me, and now I can't access it at all. If i try to change language or time period, it changes immediately back to Global - Now with No Results Found, and if I click the link to visit the actual Trending page it's just blank 
Seems to be an issue for my account, as it's the same on the mobile app too
Yes I updated all. It’s not a huge problem as I just use the app for fun but it is plenty annoying. Thanks for the response!
Update: I had a friend recreate what I did, and if someone tries to change the Language in Trending to 'Dutch', Trending becomes locked out entirely
This needs to be fixed
No
It’s weird how you’re having problems with the languages I don’t know if you’re like lyrics or not, because I’ve done other songs with other languages, not exactly Dutch or anything, but I have and I never had that problem.
@echo marsh this was what was said yesterday, but I wish that they would replied to us more.
No it's no an issue with the lyrics, its the site. I'm saying that when me (and my friend) changed the language to Dutch, it locked us out of Trending section entirely.
Oh, damn….
Im dutch
Can you access the Dutch Trending list?
Yes
When I tried to view it today, it locked me out of Trending page entirely. I have screenshots. It's now stuck on Global - Now and cannot change it
extends have volume and voice fluctuations
As soon as I selected Dutch, it went blank (2nd image).
I've refreshed, closed browser, tried mobile app/browser, all of that.
Other accounts working fine on the same devices, it's just mine and my friends after we tried to view the Dutch trending page
Now no languages work (can't even select them, it changes back to global instantly) and no results.
I never saw this bug until now👀
Yeah, I have no clue what happened T.T never had an issue like this before myself either.
I contacted support via ticket and they are looking into it
Yea me neither
I have a few songs where the persona only shows up in the workspace selected while in create mode, outside of that view the persona isn't showing up the song landing page or the profile recent song list....so the information about how its created is kind of lost unless i add those details in manually....seems like a clear cut case of a bug to me. one of the songs in question can be found here: https://suno.com/song/5a26ef72-f859-4dad-b7cc-7153f6e78801.
It's the time of the month again? Trying to get one single usable gen for a song.
-Sound degradation goes from bad to worse. It's not shimmer, it's droning or clanky or dull/muted (as if Kermit is giving it a shot at singing).
-Instrumentation is all over the place. Sometimes sounding as if they're playing on deck of the Titanic while it's sinking.
-Pronunciation mistakes galore. Or they just leave words/lines out.
-Remaster doesn't work. Again. Not subtle, not normal and don't get me started on high.
-Something new: the AI starts to sing the style prompt. Great. And it's in the style box not in the lyrics box. I checked twice. Thought I had made a mistake. Nope. The prompt was used as an intro.
-I have been using Cover until my fingers bled. Not one gen is usable. Pronunciation could be fixed in the Editor but not with that many mistakes and sound problems.
Credits are going up in flames. 🔥 It's really annoying as I want to finish songs for competitions which got a deadline.
I replied as the 😆 because you mentioned something about Kermit.
oh is that private persona's only show up in list view but not in the landing page of the song...this seems to be what I am finding now that I am looking for it...so maybe this is a feature not a bug..i can see the persona when its list view of the profile songs page, it's just missing form the song's landing page when persona is private....
Hi, I have Suno Pro, using v5. Why is Persona greyed out?
something seriously wrong with song generation in Suno. noticed it for past 24 hours. both v4.5+ and v5 are broken. please fix it ASAP
Theres something wrong with the Remix Contest link. I click on the link from https://suno.com/collab/donut, and it doesnt take you to the song. It can only be clicked on in the suno front page. Tried it on chrome, its broken.
If you are using firefox, try using chrome. As much as i hate google chrome but it seems to work there. But if you are using chrome, then idk.
Tried to submit a link to the song for the remix contest, and your suno and discord broke it showing shit emoji. Had to remove it even though its your own link your suno gave it, yet, it broke. Now i have to wait 6 hours. wtf
The inspo seems to be broken since the mashup update. I would usually select a playlist and inspo will randomly pick 4 tracks. Now the random selection doesn't work anymore and it would just pick the same first 4 tracks of the playlist.Can you please fix it?
@fallen lark
the 3rd point in that list is the bane of my existence on here. i write a 4 verse song, and it gens a 1 verse 7 min song
me and my team also facing this problem please solwe this asap
I also facing same issues with remix since new feature mashup come.
I didn't get a single song perfectly after suno introduce this feature please fix it as soon as possible.
seems like audio quality is degrading for me after the 2minute marks on my jazz/orchestral renders, sometimes earlier...a pretty clear and distinctive volume change and static/hiss....has anyone else experienced this for the jazz genre or other specific genres(note: I have been doing covers, not from scratch renders)?
anyone else facing this ?
Looks like it might have been resolved. Tried some songs.
the “Create Again” button doesn’t utilize the reference track anymore. it just cooks a random song with the lyrics and no reference. Day 6 with this problem and no response or even acknowledgement.
I have noted a bug with Persona creation.
I created a stem of a public track because I wanted to only make a persona for the vocals, but when people try to use my persona, I am told that they can't use the persona because the basis track is not public, but the persona is again based on the stem of a public track. But I don't believe it's even possible to even make a stem public. Would appreciate some focus on this asap, I have a bunch of people that would've love to use the persona, but I don't know how to resolve it myself. If I based a persona on the original track, then I will not just get the vocals, but also the instrumentals, which lowers the useability of the persona.
Here's a link to the Persona if it helps.
https://suno.com/persona/df2f165a-1f5e-46b1-8bf6-a88497d0d2c2
Based on one person trying to use the persona
"Just tested with your persona... yeah this looks like something else. It's not even giving me an error, just not starting to generate anything."
suno cover art need to be fixed plz
i think i narrowed it down to the v5 version of a persona, the legacy version didn't seem to introduce the volume drop hiss and static the v5 version dose
Super curious about this, suno suddenly started making a bunch of 2:35 songs consistently. I'm using v5 and I haven't changed any prompts because I wanted to see if it made longer songs and so far nothing
Hey there. I have put out a few emails to support .. but no luck ..
i paid for a year of pro, but my invoice says i paid the year price of pro but only recieved 1 month of premier.
can i get some help with this please (Images Provided) @sand depot please can you look into this .. I fear it may have been lost due to the most recent outage … I just wanna make tunage !!! Lmao Good luck
Hey everyone!
Just a heads up that if you're bumping into issues, the team is aware and looking into it.
Thank you for your patience as we get this resolved.
Great, now I’m having the bug like everyone else, I can’t sign onto Suno onto my phone. Cause I was gonna send this discord server a link to ask somebody about my song, and now I can’t sign into it at all.
Same issue - I just bought premium from the web version, instantly got signed out afterwards, and unable to sign back in. Thought it was an issue with my Google account at first. (But it’s just suno being down)
When I try to create something, Suno Is pushing me back to this Page, why?
suno is down
Ty
No Suno! Where is it??
Everything should be good to go. Please let me know if you bump into any problems 🙂
Can't get my library!
Yeah, I also found out in the main chat as well. Because I was able to sign on yesterday night in early this morning at like 2 o’clock but now I can’t. I hope it’s fixed soon.
I actually have a request. I know this isn’t the channel exactly to ask this but, i’m not exactly code savvy when it comes to Suno and how you coded it that is all your information no idea. But just a suggestion, not just for me, but literally for everybody else who uses this, whatever cheap code or whatever it is that you’re using that keeps breaking down every five minutes, could you switch to something a little more expensive whatever that is exactly? Cause it’s driving me absolutely insane that say I make a song and I really like it but it mispronounces one word, I go into song editor to try to fix it, and then, when I regenerate a certain part, it doesn’t work the way it should. I mean song editor right now is somehow working fine the way it should, but then other times small bits and pieces of Suno literally start crumbling like a cookie. And yes, I know that you and your team are working fantastically hard to fix this, but it would take a lot less strain off of you and your team and us who put our money into it if you use something more expensive that’s worth everyone’s while and you wouldn’t have to go back and fix it every five seconds. Honestly, I would rather pay a little more money than save $24 a year for premium or whatever than for this to breakdown because you’re losing subscribers because of how fast it breaks. And I’m just suggesting it’d be more worth everyone’s while if you up the price and made it a little more expensive. Like for yearly $60 or something, but Suno doesn’t break as much because you’re using more high-end, whatever it is that makes this function.
I did also put this in #💡┃suggestions-web
did it have to do with audio quality on renders or service down?
there is no static or his in my v5 persona vocal selection, so why is that the culprit in adding hiss, static degraded audio in my renders?
Hey Rexy2271
First and foremost, thank you for taking the time to share what's impacting your experience and how we can improve. Really appreciate this suggestion you brought up and look forward to hearing more feedback on how we can continue to do better and better.
Feel free to always reach my way via DMs if you'd like with some song link examples of things that aren't going as you'd like them to and I'd be happy to take a look and surface what's shared. Also, suno.com/feedback is a great resource to make sure these things get connected directly to the team.
Hope you have great day making awesome music!
-
why 1 specific song always lose around 50% of its volume when downloaded?
-
why cant it reliably sing/scream/rap "romance"? It's like 50/50 that is works it usually turns into nomance,rumans or numans. anything I could do to walk around that problem?
like [row·mans] type pronunciation or something
Hey Bored Ape!
With songs that are using sliders there's some super important things to consider:
Use sliders carefully! When using sliders, tweak them one at a time. Don’t jump above 50% immediately — adjust, test, and roll back or forward carefully.
If the results still don’t feel right, reset the slider back to default with a double-click and try generating again without sliders.
Going to link some additional guidance as well that may be useful: #1422711383014768681 message
Im saying once its downloaded its lost half its volume
i linked it to you and asked you in dm a while ago
Tysm, that you personally answered back. And I did send you a request to be able to send direct messages to you. It’s also really good that you’re actually taking the time effort or at least trying to make time to do so to talk to your users in the discord server because it actually means a lot to us even if most of us don’t say it. So I guess I’ll speak for the people who aren’t speaking up. It makes us feel that we’re actually seen and our problems are being heard and seen and even read. 🙏🏼
Very welcome! And thank you for this thoughtful response. I'll check for the DM request, looking forward to better understanding. Hope you have great day!
Tysm, an you too
Hi, I just wanted to share a quick observation as a long-time user.
Over the past 1–2 weeks, I’ve noticed a gradual drop in perceived quality on models 4 / 4.5 / 5, in vocals and instrumentals.
It feels similar to what sometimes happens before a major update or model transition, so I was wondering if there’s ongoing backend work or tuning.
As a lyricist, creative freedom really matters to me, and I hope original songwriting will always remain central to Suno.
High pitch screeching tones at the start of a track and about 2mins of constant screech tone at the end.
I’m truly disappointed. I’ve been a Suno user for almost 2 years, and for the first time I subscribed to the PREMIER version, paying for a full year in November 2025. Suddenly everything went downhill.
The “Persona” mode is basically unusable now: it adds noise at the beginning of every song, or a loooooong tone at the end that stretches the track to 7:59, or it just cuts the song randomly. I thought using Covers would avoid this, but it also cuts songs at random. Even in “normal” mode, it fails.
I’m burning through credits like crazy. I sent a support message a week ago and got no reply. Are we going to be compensated in any way? Will you fix these issues and return the wasted credits? Will memberships be extended for the time we already paid for?
@sand depot please give me a satisfactory answer.
Some examples:
Cuts at 0:25: https://suno.com/s/DwR2T1mkqULyXRMk
Mispronunced words: https://suno.com/s/0IptcjiKxdn2gmoy
And I've also started collecting 7:59 songs on a playlist: https://suno.com/playlist/e4a42c70-1690-4bb2-b2a9-931e8e7ca7a4
@rancid fossil Hello .. me again .. I'm awaiting any response to my ticket, post and bug report i sent (as suggested earlier) could be looked at so i know whats happening next.. can you dm me maybe so we can sort this please. 🙂 thanks in advance
I’m having the same problem. I’m burning though credits trying to stop the noise at start and end of generations when using a persona.
I hear you! I need to come up with a new song daily and it takes me forever to get a decent one.
Sweet Lord, it's a total nightmare trying to get ONE single decent song!
What style are you using?
I’m trying to create 90s vibe European electropop, IDM, electroclash De/vision or Camouflage, using my new v5 persona. Really struggling with garbage square wave noise at the start and end of the tracks.
Just on the end pieces. About 3seconds at front and about 2-3mins on the end. Turning “single arrangement “ tracks to about 7mins long
It’s a weird one!
Definitely! I'm glad you posted about it. Do you think you could link some examples?
Via DM?
Either by that or by posting here. Understand that I'll just be forwarding them to the mod's section and hopefully sending it the Team from there.
I have a theory as to why the generations are adding in about 2mins of square wave tone.
The base Persona used as the source is about 7mins long as a vocal file. Applying the persona means Suno seems to be trying to make up that time by filling in the new generation with “blank square noise”
Can you try clipping the vocal and creating a persona to test the theory out?
Il try do that and report back 👍
OK I’ve:
- Took the source persona track and cut it down to 48 secs.
- Stemmed out the vocal.
- Created a new persona from that stem.
- Used that to “cover” a prior song I’m working on.
It generated two tracks.
4.52 sec and 5.11secs.
So… early results are OK time wise but the second generation made that intro “noise” sound at the beginning..
I’ll stress test it to see if it does more and adds that sound at 2mins of noise at the end
I've tried many styles but now that I'm using v4.5 seems to be working, not better... working... at least it does the job, v5.0 must be poissoned
I seem to have got rid of the end 2mins of noise using a cut down persona but the intro noise squeal prevails
I switched languages on the trending page, next minute it's gone:
They run out of tokens?
Yep, if you switch to "finnish" language the whole thing disappears and you cannot change it afterwards either.
DON'T switch to it, it breaks your trending view.
Anyone else finding V5 Extend and cover to be broken for songs over 4 mins?
(possibly all songs)
Exactly the same. 2500 for each song, just trying to have one without mistakes in pronunciation
yeah im on 1.5k already, managed to correct lyrics, trying to cover now to correct imperfections from edits and degredation, its like v5 went back to how it was when 1st released.
been in and out of editor/studio/covers (repeat several times)
also stemmed it multiple time
The vocals no longer make any effort, no interpretive work. And Suno tries to keep up with the rhythm with unbearable kitsch percussion. It is almost as bad as V3.5 just before it ended to work
It has become impossible to make any song. Thousands of gallons of water on the planet used for nothing. I'm leaving.
Extend - Broken / Covers - Broken / Mashup - Broken. wtf. ??
and degredation back big-style after 4min.
I have had some wierd flags pop up recently saying that "DAMN" was an artist name and I couldn't use the reference in my lyrics also could not use "cross-fade transitions" in style prompt had to change due to Cross-Fade being an artist name... While I appreciat the flag actually stating what was flagged as that is helpful for fixing the problem, I just wanted to point it out and say there are some annoying ones that are flagging the system. Would think production cues would be prioritized over a band or artist name (just saying)
#🪲┃bugs-web Can anyone here tell me how to fix a bug that Suno has created in one of my songs? It starts the song normally, then there is a silence and it starts again from the beginning. I would like to be able to fix the bug, but when I edit to delete the buggy part, it removes all the original lyrics and I end up losing the whole song. Has anyone else had this problem? If so, could you help me? Thank you.
Hi @crimson zephyr , well.. sounds like an AI error that can happen 🙂 I would download the file and cut the beginning with an audio editor (like audacity)
I Got the same trouble as you, I've tried do to this on suno, but I loose all my lyrics, just because AI Suno has a bug... I don't know if you can do on Audacity, didn't try myself
well.. when i get it right the only bug you have is: track is starting.. then a pause and then the track starts normal ???
🐛 Bug Report Detected
It looks like you might be experiencing a bug! Please help us improve by submitting a detailed report.
Pro tip: If you have exact steps to replicate the issue, that will help us fix it sooner! 🚀
Submit Feedback
Need Help?
Use the button below to get assistance from a moderator who can help reproduce and verify the issue.
Thank you for helping us make Suno better! 🎵
thanks helpy.. but i am ok 😉
Looks like Extend functionality (v5) is broken recently. For example, when I put cursor in the middle of last chorus and hit Extend it never respect what was going right before this point - it doesn't maintain intensity, the voice is only remotely similar etc. Previously it was possible to make several generations and select something good. Now it is always doing something on it's own and it's absolutely not possible to do seamless extend anymore. I don't know when exactly this started, but seems like a week or slightly more..
@sand depot Sample lets you select a section of sound as the jumping off point for new ideas. Whether it’s a snippet from a Suno song, a melody or mumble in your voice memos, or something you heard and loved out in the world… chop it up and drop it into the Sample feature!
How to use Sample:
Right-click on any sound in your Suno library, and choose Sample this song from the Remix/Edit options. Then, trim your selection, add lyrics, type in a style prompt, and dial in the details with the Advanced Controls. -I also added in a Persona, but when I went to try to Create, no song was produced.
I have tried a couple of times in Studio to extract stems from a song but the output it gives is not usable. After extracting the stems then adding them to the project area and simply clicking play the song starts ok but sounds slightly rushed the 35 seconds in it slows down dramatically. I have tried setting the BPM manually but the result are the same. At this point trying to extract stems is only throwing away my credits.
#🪲┃bugs-web Hi guys, I had a pro subscription and more than 2,000 credits left, then I decided to switch back to a premium account. When I upgraded, the credits I had left disappeared. I ended up with a free account, then I was charged for the premium subscription, but basically my previous credits from my pro account are gone. Normally, I should have had the rest of the pro account + the premium in credits, i.e. +/-12,000 credits.
So, if I send an email to support, and a few days later I send another to provide more detailed information about the same issue, I get bumped to the back of the line? That is not a best practice - which is to have a ticket number reference so that the ticket would be updated, and the it's queue not be effected.
Please direct issues related to support emails to the #archived-ticket-help channel. Thank you.
Been doing that @solar grove. Still doesn't help me any though. And for what it's worth, your advice is being heard - and the work you are doing here is appreciated. 🙂
everything dosn't work, your website is rubbish, (your ai is great tho) but your website is rubbish, nothing works *will also add this to andriod chat
There might be a problem with the MP3 file that Suno has created (you mention that the song restarts from the beginning.)
A couple of things to try:
Use "Adjust Speed" from the Remix menu but leave it at 1.00x to just create a copy of the song.
If that doesn't help:
Try extending from the end of the song. Just put something like "immediate stop, end" in the style, and leave the lyrics blank.
It will likely create a short extension.
Then, for that new part, use "Get Full Song". This will create a new MP3 that shouldn't restart the song.
You may need to use the editor on this new clip to cut the end part off (if it's not silence.)
Finally, try downloading the WAV and see if that has the same problem. If it doesn't, reupload it.
All of these would create a new MP3 and hopefully get around the issue you are experiencing.
Submitting feedback on the web app doesn't work. " You must be logged in to submit a feedback"....But I m ligged in.
I'm noticing model 5 recently reverting to the days of 3.5 mispronouncing of words. At one point it seemed to notice it made a mistake and added the section back as another attempt and then says it correctly. So the model is having emergent misalignment with a guilty conscious and will likely develop insecurities and imposter syndrome. See this fake diagram for more useless information regarding stuff that isn't important or real, distracting you from what I'm really trying to say which is it's engaging in Sloppy English like NOFX's Liberal Animation.
use a hyphen with it or a synonym, annoyingly there are a bunch of words that are also artists, I couldn't use "bratty" as a descriptor for the same reason.
go into editor, crop the begginning off and repate the lyrics into the new song once saved.
Hey suno, fix your extensions! Fix your song generation!
Hey all,
I was watching a video from last week about making a persona from recorded vocals, but it's grayed out for any recorded clip. Is my premier account glitched or something or did they remove the feature again since the beginning of the week?
Accidentally posted this in #💡┃suggestions-web. Many apologies for the double-ping it caused. I am what I am. :)
Dear, @sand depot.
I post this in public as it deserves attention, but also as I seem to have been forgotten or drowned among other DMs.
Anyway - I think I stumbled across a workaround for early and abrupt endings. Though it does introduce a new issue, at least now we can manually fade the song and not have an ongoing ringing string or vocalist get cut early.
I added ```
Example 1: https://suno.com/s/6sZAhw0IfsRTtVbV
Example 2: https://suno.com/s/yXeW2E2xzLouQP1d
The second example is severely broken in many ways rendering it useless, but it goes to show when comparing the two - that parameters have _nothing_ to do with the issues gens end up having, as these issues happen with defaults as well.
I will post more examples as I stumble across them, but it also happens to a piano in the second example. :S
This clearly demonstrates exactly what's happening both with resonance feedback driving some weird kind of amplitude thing for each finished bar, and also that there is 0 need to destroy otherwise decent gens by cutting early simply due to there being no more lyrics and the source material being covered doesn't have the bars to let instruments and vocals play out naturally.
I've been thinking hard about explanations and workarounds for various issues this week, and after today's testing; I'm close to being ready to publish a proper bug report before I ultimately cancel my subscription as soon as I've burned the remaining credits on my account, as the product currently stand is not worth the money and further compounded by my frustration with biased and down right unfair treatment in regards to moderation as discussed in the last DM I sent you.
Best regards from the strange internet Norwegian,
Geir.
#freechronicx
I'm almost afraid to say the following, but I will do so anyway and it should become evident as to why when you read on.
Hopefully what follows end up interpreted as constructive feedback/criticism and not a vile personal attack, for it is absolutely not the intent, but I'm simply unable to phrase this much more kindly.
What's reported here; is such a common issue and present during playback on the website, and it happens with default parameters as well as exotic ones. And in the case of the latter; you more or less always point to as being a cause every time someone reports a repeatable issue.
In context; the suggested workaround and reasoning is almost concerning as the issue not at all presented as being exclusive to an exported MP3, nor at all aligning with the current experience of being a user that is actually fuzzing and pushing the service in order to come to grips with what's happening when things go wrong, while also verifying your proposed solutions which ultimately serve to shift the blame on to the user. Issues which by now; there should be no doubt about being rooted in underlying technical issues that this platform have been suffering from for months (ref. above mentioned pending bug report).
I'll end with stating that someone in your position should be well aware by now that the Extend feature is entirely broken, and that if one actually gets vaguely usable output from it which is remotely close to what you claim and more importantly; it says on the tin, would be the equivalent of being struck by lightning while seeking refuge in an underground bunker.
I guess I'm getting banned now, but honestly; I'm well past the point of caring. These suggestions needs to stop and focus be shifted from lackluster and unimportant UX changes which seem to break more functionality than they implement (certainly doesn't improve) - to fixing the parser which ultimately feed the text encoder which your service rely on, and it too is likely incorrectly implemented.
Final readability edit performed, and I'm done for now. Until next time, if at all. Have a good day regardless. :D
Following up on ☝️
Perhaps the best example of what I'm describing as a cause of the ever increasing volume issue (found at the very end, as with the rest of the examples provided so far):
https://suno.com/s/v0qgtvqgZ1e5cum4
An example of Suno somehow concluding that "Wait, no more lyrics and 4 bars with actual silence? 1 is likely enough. Let's do that.":
https://suno.com/s/yXeW2E2xzLouQP1d
🌪️
Another example after getting a decent result with a single -, but since it faded out the reverb I went with two and got this:
https://suno.com/s/u29pQZN2LXfuxgZG
Suno seems super into the whole reverse fade-in and feedback thing, which is likely something to target during re-training. At the very least how it should end songs properly, because this isn't how it should be. Not a single engineer in the dataset would have done this to their work, as it makes no sense.
The second song from the same gen exhibits similar traits as when using 4 dashes:
https://suno.com/s/KzdbJbfbpI7ZfvNF
Testing the effects of . after N dashes next. I mean, it's gotta be possible if there's nothing wrong with the back-end and it's all on users prompting, no? :D
I my account won't let me sign in. This happened to my phone around a month ago, and here is my trace id:
Trace ID: 140212950509331277706571853637255495793
If you have placed 2 clips in the "Mashup", and then try to "Sample" a track, then try to click Create for the sample nothing happens. You have to click Mashup and remove the selections there and go back to click Sample in order to create. These shouldn't be stacked, like a Persona for instance. Should act more like a seperate tab. (Until you're ready to mix match more features together.)
Point being, someone using Mashup, and then try to use Sample (desktop) may not know why Create isn't working. This is why, have to clear the mashup selections first.
Subject: Major regression in "Cover" function quality (Starts ~Jan 28/29)
Hi team and everyone, I suspect a silent model update or backend tweak happened during the overnight maintenance window on Tuesday, Jan 28 (EST), which has severely degraded the Cover function.
🛑 The Issues:
Harmonic Errors: High frequency of wrong chords and dissonance that wasn't there before.
Style Drift: The model now ignores the nuance of Style Prompts, producing results that sound nothing like the previous baseline.
🧪 Testing Evidence: I tested using a "controlled" method:
Same Reference Audio
Same Prompt & Persona
Result: It worked perfectly on Jan 29 AM (Beijing Time), but broke completely by Jan 29 PM. This aligns perfectly with the US overnight window.
👥 Community Confirmation: I cross-referenced this with a local community of professional AI music producers (on TikTok/Douyin). The consensus is unanimous: everyone noticed the degradation starting at the exact same timestamp.
Applying a new v5 Persona to a cover doesn’t work. I’m getting no application of the vocal onto the cover at all.
Works fine when creating something new, but applying the v5 Persona to an existing track doesn’t replace the vocal with the new voice at all.
Been trying to create a persona and im on the premier tier. except create persona is greyed out?? anyone else getting this?
Create and Remix not working at the moment? I have pro and credits
Not that I'm here to be the one providing support, but you told @crimson zephyr to try downloading the WAV of their damaged song and then re-upload it to Suno.
That's gonna be a bad time in all likelihood. For at least a few weeks, Suno has been matching re-uploads against the database of its own generated and user-uploaded library of songs. If the re-uploaded audio isn't significantly changed (e.g. pitch, tempo, compositional alignment) it will be rejected as matching a copyrighted song. I didn't report it as a bug before because I assumed it was intended, although it may be a factor in the bug I reported days ago here and in email without acknowledgement from anyone: There is an extremely high false-positive rate for recorded vocal improvs, possibly due to "matching" of similar timbre, room tone, and phrase stable in past uploads and/or high-similarity (95%+) generations based on those uploads.
Just deleted one new song via the regular way (... menu) - but in fact 42 went into the trash can und got unpublished. Had to restore and publish them manually again.
Since last two days sound quality of v5 became absolutely unacceptable. Overall sound is very noisy with accent on hissy high frequencies and bad stereo field. The sound now reminds me old v3.5 days.
This is especially noticeable in covers - they are very noisy and absolutely unusable. Besides, I often encounter pronunciation issues in simple words. Thing that never happened before.
It's very sad, that Suno treat their customers as alpha-testers. Do you guys even doing internal testing before releasing something in production? Currently the whole platform became absolutely useless...
Yeah, I also release this problem as well
Suno fix it as soon as possible.
Suno web is down for me. when logging in from chrome browser in european union, I get the birthday verification page which I complete (I am an adult). When sending the information with pressing on the „Continue“ button. I get the studio-api.prod.suno.com ERR_NAME_NOT_RESOLVED
try clearing cache
is suno down?
Yeah seems like it is
Cloudflare (sunos host) is experiencing some issues which may impact some people
It's up for me, but with the current V5 degredation, Extends, covers and mashups all screwed up, it might as well be,
It's up for me too. It just depends what node you're routed through. They're (Cloudflare) also performing maintenance which can cause latency and timeouts
Extends are horrible
Issue: Payment to @wooden pelican via @Stripe succeeded on Jan 26, 2026 (Invoice 3D1C9D8F-0002, Receipt 2391-5766). The $10 monthly Pro subscription was NEVER activated. Support tickets for account ........ are ignored. Day 4 without service. Please escalate or I will proceed with a formal chargeback. #Suno #Stripe #PaymentIssue
@burnt jackal @dry fox thanks you guys. Unfortunately, it is still not working but I will check cloudflare status. I am on mobile right now, therefore I can‘t check the network packages to see at what connection it exactly fails. I tried with isitdownforme and it‘s up from USA but not from Europe it seems.
I don't know if it falls under the "bugs" category, but v5 cover feature is broken. You upload a 5 min audio, you ask specifically in the style prompt to maintain duration, you even guide the model in the lyrics so it follows the song, and it just doesn't and generate stuff sometimes half the original duration.
i've tried mashing up covers and they are corrupted withou lyrics and less than a minute at times
I have an account issue. I changed my Google domain and when I login into suno with my google login, it has created a new account with my new email for Google. But this email is an update to my old one within the same account, so now I cannot login to my Suno account. It is somehow still using the old one. Which cannot be used.
I went within my Google settings and updated myself back to the old email, and now I can login to my Suno account. I'l;l try to manually change the email from there. Or take myself off Google login haha
Couldn't generate that.
Your tags contain artist name menor - we don't reference specific artists on Suno, please change your tags and try again.
Promt
"title: "Juegos Peligrosos (Treino)"
language: "es"
mode: "Custom"
genre: "Electronic / Brazilian / Workout"
mood: "intenso, físico, enfocado, humano"
bpm: 134
key: "Fa menor"
voice:
type: "male vocalist"
style: "short phrases, rhythmic, percussive"
delivery: "dry, close-mic, human breath"
processing:
- "EQ Eight (high-pass 90Hz)"
- "Glue Compressor (light, 2:1)"
- "Saturator (soft sine, low drive)"
- "Delay (1/8 dotted, low feedback)"
- "Reverb (short room, low mix)""
its better just to make a new song using "cover " features
Why do some of my cover images turn into a pink and red default image when I'm on my profile songs list or when they're pinned on my profile. I thought it might be since the images have people in them but one of them is literally just a tree. I noticed only the larger file sizes of my custom images didn't work (around 10mb) meanwhile the ones around 1-1.5 mb worked. Then I compressed one of the 10 mb images down to about 560 kb and it still didn't work, though only when I edited it to have the character in frame, when the character wasn't there it worked... though this is only on my desktop, any browser. When I look at my profile on my Samsung phone the images work.
I disliked a song and now when I try to go back for in my dislikes it’s not there, I even checked my trash….its gone
Some body help me. When I click create I'm getting a an error occured. A system error occoured.
Alright I figured it out if anyone else runs into the same problem: Even though I compressed my cover images to a smaller file size, the pixel dimensions of the images were still quite large. They were still 2304x4906 pixels in dimension, but once I resized the image to 600x1060 pixels it worked. The image looked about the same, so it's all good. I had used an AI image tool to create my images, but some of them I then edited afterward, and it appears that the edits increased the dimensions of the image to be too large.
Ye na, at some time during testing yesterday I noticed a sharp degredation in audio quality as well. Vocalist started mumbling, misspelling words it hadn't up to that point, same parameters that were fine prior had same issues. Tried manipulating lyrics, replacing parts which sort of worked, but then the issue moved further into the song, returned when lyrics were reverted, etc. At the same time, the guitars would start getting fixated on missing chord changes and sticking with the penultimate chord before a bar concludes, over and over. Something akin to 1 out of 50 gens came out sort of alright, but there's always something that isn't even present in the original track I was covering so far it can be called a mistake or error that it would make sense for Suno to carry over. Entirely new stuff. 🤪
I mean, can we pretty, pretty please, revert to like November? I'm talking the whole thing. Frontend, backend, the works. Everything. I see nothing that can be called progress since then. The trajectory this platform has taken since I created my account is bordering on shocking from a development and UX perspective. It makes me wonder if the whole thing is legit the product of vibe coding. I can't for the life of me understand. It's as if the people whom designed the Library and how it present its content to the user, have ever used an actual filesystem. It's lacking every basic feature that is actually useful (a simple directory structure would be sooo nice), but then I'm like hm. LLM's know what and how filesystem works better than what we have to live with on this platform, so I'm thinking - maybe you guys should let the models roam free for a while, because it literally can't get worse. xD *stares at @white ice* On second thought, maybe it can.
Besides of severe sound quality issues, I have a strong feeling that v5 model was silently changed under the hood as well, because it is now produces semantically different content with the same prompts that worked well before. Now, no matter what you prompt i hear a major influence of contemporary mainstream pop. Like style influence not working properly anymore. Or... maybe it's a silent start of WMG era?
I’ve noticed the same. I’m getting very homogenised glossy pop results no matter how many time I write dark, indie, experimental etc
Only way I vaguely get the things I plop into the Style prompt is with the Style parameter set to 97-100%, but at 100% it's super hit and miss in regards to the encoder correctly handing off the full prompt in a way which result in output that's not singing the prompt or otherwise having a mental breakdown. At 97% it is extremely unlikely ending up with output where the prompt is either spoken or sung, but even at this value; instructions wrapped in parentheses to indicate that they're very much desirable, is more often than not entirely absent (Ie. snare ghosting). At 98% it's more or less guaranteed that it'll include at least something close to what's described in the prompt, but more of a sprinkle than following instructions to the letter. 99% it starts becoming something which actually match the prompt, but this is also the point where the chance of getting Style poetry mixed in with the lyrics skyrockets. But with enough "rolls of the dice", at least there's a chance of getting something which resemble what's described.
Running defaults just result in the most boring dull generic ... stuff, that sounds just like what everybody else is generating and comes nowhere close to resembling what I'm meticulously describing in my prompts, while not putting the equivalent of shackles on Suno. So I can't for the life of me understand why it is such a common advice touted by staff, especially since the same issues are present with defaults (which I'll continue to challenge until they stop, because it is really that far off reality). It's as if they are no longer using their product and defaulting to responses based on "it used to be like that, so it probably still is", rather than what users are actually expected to keep throwing money at them for.
Good points being raised here. I've definitely noticed a shift into 'bland'
It's amazingly boring stuff. All of it has this super whiny tone to it, excessive amounts of gasping for air, and incorporating the inhalation as some kind of singing, and it doesn't matter if it's bluegrass, nu metal or whatever passes for current day "emo". If I'm not covering a song which is already free of these traits and running a high Audio Similarity parameter value, most of my prompt ends up being spent on describing what I don't want as opposed to what I want, as the Exclude field is in reversi-land and acting as a positive prompt. To put it this way, Vocal:{(no insert_undesireable_trait_here), male baritone, ...} with a high Style parameter, will end up in what is described - but with need to iterate on the prompt to remove undesirable features. Type the same undesirable trait or genre into the Exclude field, and it's more or less guaranteed that you're gonna get a lot of it. xD
Anyway, it dawns on me that perhaps I should finish that bug report rather than sit here and rile myself up. If I were to put it kindly, the UI demonstrates that whomever is making it, or those whom are making the calls - haven't got the slightest idea how and why NLP sort of works to the best of our understanding of these strange black box things we define as LLMs, as clearly demonstrated by we as customers being limited to 1000 characters which includes blank space, delimiters, newlines and likely other characters which are entirely meaningless so far context windows are concerned from the perspective of a text encoder. To the point where the parser which does the initial conversion of the 3 input fields into whatever is handed off to the text encoder, is actually translating punctuation which is used to define weights such as (desirable feature:1.2), desirable feature, (less but still desirable feature:0.8)into (desirable feature:1, 2), desirable feature, (less but still desirable feature:0, -8)... I mean, come on, one can't make this stuff up. It's amazing! :D:D
Just fyi, I've had this issue even when applying an image generated with Suno's image generator, regardless of downloading the image and uploading it the old fashioned way, or via the new Cover Art section in Library. The only way I've been able to get around this consistently is to delete the old image with the trash-can icon (Barely visible in the top right corner on top of current Cover Art image) while in the "Edit Song Info" dialogue box, adding a space to the "Description" field, then uploading, returning to the main dialogue box, remove the space (otherwise the Save button can't be clicked) and voila - the image sticks. xD
Now, this is why I'm kinda sounding a little bit worked up over the quality of code which is being pushed to production. I just can't comprehend how this passed QA, if any were performed at all.
All said; Suno as a model sounds amazing when it works and doesn't end up increasingly distorting the output for each finished bar, but oh my gosh. Oh my gosh. Oh, my, gosh. I hope somebody with an incling of being capable of comprehending what I'm writing sees this and actually do something about it sooner, rather than later. I'm not kidding when I say that this is the last month I'll be paying for until I see what can be defined as actual progress. I love this process to death, so it's kind of heartbreaking, but I've had my fill. Even though I have all the time in the world to spend trying to "make this not be me doing bad prompting", the novelty has worn off and then some. Now. I shall go do other things before I, like Suno, have a 100% Style prompt mental breakdown.
Feature request: API access so that we can migrate to a local front-end that is sane. Kthxbye.
Suno API? There is no Suno API.
I need help. I subscribed to Suno, but when I download a video, a watermark that says “Suno” appears. How can I remove it?
You mean like this?
Sim
It's by design.
If that's what he's talking about. I think it is but without confirmation...?
This is yet another misleading statement. Sorry.
My request, past it being more of a gag leaning on the context provided by my rant just above it and the sole reason as to why I wrote it in here and not #💡┃suggestions-web, is exactly because I'm aware of Sunos current stance on this topic.
The context provided above outlines why allowing public access could be a good idea, and the joke being that it would free us from relying on the UX that the current website provides, and thus in a sense eliminating this aspect of what is a huge part of the problem with Suno as it stands.
In fear of however this may be perceived, all I can say past this point is by returning to why I opened my response as I did by calling your statement misleading with the following counter:
Of course Suno has an API. It's just not "public" in the sense that it is "allowed" to be accessed through third party applications. Perhaps parroting these things should be reconsidered.
Take away what you may from this. I've got screenshots of what I need, and I'm ready to be jettisoned. *braces*
It happens sometimes. If you cannot figure out the cause, which could be anything, try changing song words until it no longer occurs which could be time consuming, or just try something completely different.
Yes, this is happening to me. What can I do to remove it?
You don't. As stated this is by design.
Videos like that are really a legacy thing IMHO from back when everything was generated on the Discord and output was in MP4 with a generated cover image. These days we have WAV and MP3 options, can add video album covers, add our own images for covers...
Not much need for video output any more.
One off issues like yours put in for https://discord.com/channels/1069381916492562582/1372994553233276970
Past that everyone has come to expect this feature which have been mostly working to continue being a thing. I mean, sure. Why not throw it out.
What do you need a video output for? Asking to see if there may be other options to give you what you need.
Check your filters.
Other than it can cause issues like this one where people think it's watermarked when it's just the way it outputs it.
Hmmm...that should be a feature request though.
So you're saying that putting your company logo on a thing which people potentially own the distribution rights for isn't the equivalent of a watermark? :D
I’m new to this. Which tool do you use to perform this function?
Not in the sense that it's owned by Suno which is the usual concern.
People tend to think it's a sign that the generation was on a Free account.
I usually output to WAV for long term and MP3 to play in my car.
I'm sorry, but "you" in this context is of no concern.
I was asked and answered what I personally do. Look above.
He literally asked you "how do you export video", is my interpretation - which may be incorrect.
Thank you very much for the responses.
There's so many reasons why having the video format is beneficial not only for users that doesn't wanna or have time to get into things like video editors, but also Suno as a means of advertisement - especially when the company logo is stuck on top of everything else. (Edit: Which honestly, is why I thought this feature existed to begin with and have found super useful for sending things to family members that struggle with concepts such as webbrowsers in general as it can be easily played directly within most messaging apps - and also they get to see the lyrics. <Insert "Suno mumbling excessively"-gag here>)
Been trying to create a persona and im on the premier tier. except create persona is greyed out?? anyone else getting this?
did you cover an upload or anything?
Also - i have a strange little bug here, when i zoom in/out a part of the waveform changes into a brick haha. Sound remains the same though
I did.
Using uploaded audio as remix/cover in SIMPLE mode, with a prompt for music style and lyrics about whatever. It keeps generating only instrumental tracks now. I made sure the 'instrumental' box was not selected but it will still not generate lyrics over the track in simple mode like it used too.
MAJOR BUG: If you generated a song in the past on some sort of premium tier, or if you use someone else's song, and IF that song had negative styles, you cannot Remix/Edit -> Use the Styles and Lyrics.
If you do, you will get an error saying you cannot use negative styles, and you cannot remove the negative styles either. That part of the menu simply doesn't show up!
i cannot generate music with suno
im having this bug where it shows error when generating music
im going to stop using suno
i get this for many days
i was also a paying user
Is the static noise and lengthy generation issue fixed?
Not for me. I was getting that high pitched whining noise on the start of generations again
Arbitrarily and unreasonably freezing user accounts—such business practices are a recipe for bankruptcy. A rogue AI system.
Hello Community. For some time Suno has stopped updating my "plays" and "likes". It is also not showing all the notifications that it is telling me are there. I have uninstalled/reinstalled, cleared cache etc. There is no change. Can anybody offer any advice to help resolve this issue, please?
I unpublished a song while it was pinned to my feature song list, leaving 9 songs. I was unable to pin a new song in the 10th slot until i located the unpubbed song and selected to unpin it. browser tab was open with a workspace that had not been reloaded (since it was actively being used to create) and so still showed up as published and pinned, after unpublishing (as to be expected since reloading the page would lose my current create settings). So maybe it can be reproduced by publishing a song and pinning it from a create workspace, opening a separate tab, unpublishing it from the featured song list (w/out unpinning) w/ 10 songs, and try to to pin a different song .
I try in every way but when I try to crate it's still show up "An error occured. A system error occured". Somebody help me for this. God help me please.
Any body here could help me please...
Can you try opening an incognito tab and seeing if it works there?
What browser are you using, what is your operating system, etc?
Windows? Mac? Linux? ChromeOS?
window
OK. What are your prompts?
i just let ai prompt for me by click at the magic wand tool
Try as a test, putting something in there yourself.
I fear that you will need to try emailing support@suno.com for further assistance. I do apologize.
you always have tio use a refreshed tab to pin a song after removing one, doesn't need to be the one in-use
Thanks Hepcat
Looks like we get 0 reply as to why Mashup/Covers are limited to 4min, why degredation after 4 min is back, extends and replace is broken.. Just silence as usual.
Not just that but the general structuring of a track is completely messed up now as well. You can add 5 lines of lyrics and a few but very specific prompts and most of the time it'll spit out 5-8min tracks where 80% is just instrumentals, on top of the degradation. Can even use the same prompts that would generate a 3-3:30min tracks in the past and it'll now generate 6-8min tracks.
Ngl, it's not looking good right now.. A lot of people's credits and actual money is being wasted while there's zero transparency.
Not for me either. I was testing the new Sample mode with some tracks and got several. i also got two that were all noise for 7:59. I emailed tech and i get the same generic autoresponder that they're working on it. Its frustrating. i got some decent results using sample when i wasnt getting errors. Id like to go back to what i was working on before the errors started. My sub rolls over tomorrow, and i have way too many to just lose because of this nonsense.
tab was refreshed....the issue was unpublishing a pinned song , still occupied after refresh....if you read the post, it mentions that the song was visible in an unrefreshed separate page...i just mentioned that in case it was part of the bug....to put it simply, i unpublished from pinned, refreshed, and it wouldn't let me add a another song to the featured, until i unpinned the unpublished in the non-reloaded page.....so its a tale of 2 pages....legitimate bug...with a common use case of someone keeping a un-refreshed working tab open generating renders.
Yep. Meanwhile how many credits are wasted by the overall user base because of this bug?🤔
omg omg omg omg ! autism squeeeeee...... when the vocals started it was me in a live room no mic..... weeeeeeeee......... autism sqeeeeee
but i broke sunos samplers brain...... the backing was pretty good. but it keep playing the wrong instruments... i kep hearing a drum kit. so i fixed one part of the song but i still broke suno brain without lyrics instruction
https://suno.com/s/NycpxZ02YMNC0rHj
(fixed one error in edits)
just so you know this was the prompt.
I should be putting in but i went with simple lol
Production: cleanly produced studio TURN LYRIC VOCALS INTO INSTRUMENTS:
CONGA DRUM[E], CONGA DRUM[F], CONGA DRUM[G], ACOUSTIC GUITAR, ACOUSTIC BASS GUITAR, 1954 ROCK&ROLL,
rich yet dry, ensuring every rhythmic detail and vibrant arrangement shines with pristine clarity.
Tempo is 140 bpm.
Description: A pristine, studio-quality track,TURN LYRIC VOCALS INTO INSTRUMENTS:
CONGA DRUM[E], CONGA DRUM[F], CONGA DRUM[G], ACOUSTIC GUITAR, ACOUSTIC BASS GUITAR, 1954 ROCK&ROLL
Songs are for certain genre are creating noise after completion. At most case it's between 6-8. Will there be a fix to this soon?
Calm your CAPTCHA down a bit. You really shouldn't have to pull yourself out of workflow just to solve a weird puzzle and render a track every time. This is ridiculous!!
If I'm to extrapolate on what I've read from staff, this is due to what can only be defined as a severely limited token limit, aka context window.
This is why the Extend feature exists and why songs which are generated purely from prompts are limited to 7 minutes in length.
When doing Covers, Mashups, using Personas, Inspo and so on, the reference material (essentially Songs) also count towards the token limit as Suno must "listen in and memorize" in order to replicate it.
This in turn means an already limited context window becomes even smaller, as there's less tokens available for what is to be generated.
The shorter the source material being used is in length, result in longer Covers (going by my own experience) being achievable through instructions and is likely the cause of issues such as Covers ending abruptly when the source material exceed a duration of ~3:30.
The basis for this theory is that when I cover a 2-3 minute long song, it's extremely rare for it to cut early. When covering a song which is 4-5 minutes long, it will more or less consistently cut at around the 3-4 minute mark depending. There have been rare exceptions where it actually makes it to the end, but more often than not with heavy degradation in sound quality, increase in tempo or rushing to fade out the song before the band is finished playing or vocalist is done holding that last and sweet, sweet note, unnaturally cut off lingering reverb, sustain or delay, etc.
Similarly, the Covers which are about 2-3 minutes in length, rarely suffer from degradation in audio quality (except for the ever increasing volume to the point of distortion that render the song unusable in any professional setting).
If I'm correct, they are burning real-world resources and our environment for no real value in so far the work performed is concerned, due to fairly misleading descriptions of features and promises in regards to quality.
An example of this is me as a user attempting to debunk most claims put forth as fixes and workarounds, by spending most of my credits on testing rather than effectively iterating towards something that resemble what I'm after. And this being a direct consequence of frustration caused by what is described above in regards to any given feature's descriptions and promise of quality. All while being assigned blame for not prompting correctly.
This is why I'm vocal, and will continue to be so, because this is not sustainable economically, nor environmentally - and likely will drive people whom are like me, but far more persistent; up the walls and subsequently straight into a bed at an asylum somewhere.
As per protocol, contextual refinement and reability improvements have been made in todays episode of "Meow, meow, please stop destroying the planet for no reason due to limitations which make no sense, unless this is actually the goal, driving people insane while wasting their credits in the process".
I think perhaps it just dawned on me why the new Personas feature introduce so much noise.
For those whom have actually used Studio and Stems, it's extremely common to end up with individual tracks which have an absurd amount of noise (often bleeding over from other tracks) which need to be cut (if at all possible).
If the new Personas use vocal stems, then the selected song for a Persona is likely not cut down sample-length clips consisting only of pristine vocals, but also the noise which is described here. This would also explain why Personas severely hamper the available context window...
As the feature is likely relying on the same trick as we known as "Audio Similarity" but isolated to vocals - then in context of my speculation above; this is not a surprising outcome to me. At least not anymore.
Please, team and staff; consider this an open letter and challenge to counter my speculation with some form of evidence rather than attempting to shift blame on to users. It's obvious that we're not the problem with your service. You rely on our perception and ongoing funding long term, not your investors.
Sorry for the cross-post, but I believe this song I doodled toghether will at least somewhat demonstrate that my attempts aren't entirely effortless while attempting to get Suno to co-operate, and that my comprehension of the ramifications of my efforts isn't lost on me - as such; making this song exceedingly relevant to the topic of "Suno v the environment (and by proxy; users)", have a listen: #🐣┃suno-showcase message
The irony this song represent borders on satire when considering everything - yet I still find myself very much in love the process which is Suno. That is all. Good night.
Somehow managed to copy the wrong link.... Final post updated. Sorry, I'm word-blind and it's a hopeless affair.
What's the effing deal? I'm only getting 1-2 minute songs here (when not getting 7 minutes of a single tone). Also tons of plop plop sound artifacts on personas. Give me my credits back and fix your sht!
Yes. I have this issue with uploaded or recorded audio, has anyone answered this?
It's disabled by Suno on purpose, it's not really a bug
I left Suno playing my tracks in the background for several hours, and the song description panel got stuck on the first track while other songs kept playing. As a result, all the listens were incorrectly attributed to that first song, which isn’t even published, nor titled (i.e. the song titles are "untitled") and they have the same styles.
As a general bug, if I select instrumental in the prompt box, and my prompt does not contain any vocal requests, Suno automatically adds vocals in the style description
Hey ive noticed when ever i try to do a mashup the song never has an ending, its always cuts short
Add instrumental to a song cuts it off too. Won't go past the 3 minute mark or so
If you put in the style box
Artist name Style
And use 2 artists names this way, it lets you generate but the outcome is nothing like those artists.
I specifically used a male and female in the same song and I did copy/paste. Had no idea those names was in there till after I did a few generations.
Also if it helps those 2 artists did several collabs before the lady died
i'm getting a lot of songs that will generate to 7:59 but when you play the song it will be like 3:40 long the rest is just noise. When I try to repair with replace or extend, what I get is so far from the original output that it is useless. The style changes, the persona, pace, overall audio quality. Sometimes it takes 10 or more tries just to get an ending that is a few seconds long, each costing 10 credits to generate.
[Bug / Critical Issue]
Account Plan Not Updating After Payment
Issue: I subscribed to the Premier Plan (Yearly) on Jan 20th, but the web interface still shows my account as 'Free' tier.
Impact: I cannot use any paid features despite the payment being completed.
Note: I've emailed billing@suno.com over 2 weeks ago with my receipt, but have received NO response. Please help check my account status.
The play counter is borked again too. xD
@sand depot ☝️ How hard is it to resolve this issue and why isn't it prioritized? This has been haunting the platform since before I joined.
Right I get that. Is it permanent? Because this is the entire reason I paid for pro
anyone else getting weird dips in audio quality in your song generations? like one moment the music sounds crisp and has great quality then it abruptly becomes muffled and poor quality, and, no, it isnt the music genre nor something i put in the prompts (ive tried to delete or change the prompts to find the source and it doesnt seem to be anything i input). it comes and goes, so like one song generation is fine, or maybe even a couple songs are fine, but then another generation is of a crappier quality. i think imma take a break from Suno to let it sit because maybe the AI is getting overworked or something
Yes I l’ve had this. The track loses all clarity and anything about about 400hz is cut. Usually it’s the first verse in my experience
yea, it's usually early on in the song. 1st verse as you said, but generally around the beginning then may continue through the rest of the song in that muffled, poor quality. like i said, idk if the AI is just being overworked right now and cant generate properly, or theres something wrong w/ the site/server or something thats making the AI hiccup
I've uploaded my own vocal to try make it a persona to then write lyrics and my own voice will sing
But I cannot select "Create a Persona" - it's greyed out
Anyone know how to fix?
It’s not a bug. It’s a copy protection thing that stops you uploading potentially commercial tracks to then create voice clones from.
right, its my voice though so how do i do it?
Unfortunately Suno cannot tell the difference between your own voice or Shakira!
yea thats fine, but how can i turn my voice into a persona
Bottom line… you can’t on Suno. But you can using third party software like Audimee or Kits.AI
Is anyone else experiencing very inconsistent vocal quality lately?
In my recent generations, the instrumental side often sounds great and very detailed, but the vocal engine feels unstable and unreliable. Vocals frequently come out unclear or washed out, with slight pitch vibrations, metallic artifacts, or a muffled, noisy character.
What’s confusing is the inconsistency: one generation might sound fine, then the next suddenly drops in quality, even with identical prompts and structure. I’ve tested different genres, removed vocal chops, simplified lyrics, and adjusted prompts, but the issue persists.
At the moment, it feels like the vocal engine is struggling, while the beat and production quality are actually very strong. Because of that, I’m taking a short break and waiting for things to stabilize again.
Just wanted to share in case others are noticing similar behavior.
Extending has become a serious problem. It no longer stays with the original song’s style. It says "remix" above the clip. How can I fix this?
For reference, I didn't have room for all of the lyrics I wrote for the song. Not enough room for half of the last verse and chorus, and the outro.
Well there is still a way to persona your own voice but let's not say it too loud
idk if this is a bug or where to report, did it deduct my 50 credits, if so, how to get it back?
"Get stems" still doesn't work for long songs, it just hangs forever. Any plans to fix this? I'd be happy paying more credits for this feature.
It does at least refund credits for the stuck gens, but it would be nice if I could just one-shot the stem gens for longer songs without having to split them up
Hey guys, quick question, are you also seeing a prompt on the Suno website (desktop login) asking you to confirm your age? Just wondering if it’s new or device‑specific.
Has everyone received their Suno credits for this month? I haven’t gotten mine yet.
Is there a reason why I can't seem to upload my own cover art. It just is this neon orange splotchy thing.
ok I was curious and clicked animate and now I see it. but it also cost me 50 credits
If it happens again, there's no need to shed 50 credits, just do a little hop, skip and dance through the myriad of QA-mysteries that makes up this platform: #🪲┃bugs-web message :3
i had remi make lyrics but when i went to make the song it said there was inappropiate words which doesnt make any sense. if remi made the lyrics they should be ok in suno. plus suno puts bad words in songs itself.
#🪲┃bugs-web ETA on fix for "Cover" function regression (Started Jan 29)?
Hi team, the Cover function has been severely degraded since Jan 29 (wrong chords, Persona weights ignored).
Isolation tests confirm: Persona works perfectly in Text-to-Audio, but is completely detached/ignored in Cover mode. Is the team aware of this pipeline break? When can we expect a patch?
#🪲┃bugs-web Multiple communities have confirmed that Cover mode is ignoring Personas since Jan 29. Resulting in generic humming and broken harmonies.
Hi on browser web lyrics not work, its a bug?
It’s completely ignoring my persona when trying to apply as a cover to a song I have generated. Totally useless on covers. Seems It’s only working on new original tracks
I cannot access my phone-number login account on PC.
On iPhone, the same phone-number account works normally.
But on PC, signing in with the same phone number creates a new empty account.
This seems like a broken session or account linkage issue on Suno's side.
Please repair my phone-number login account so that it connects to the correct data on PC.
hey MAKI, when you sign in using your phone number please ensure you're using the exact same format as you signed up with, otherwise this can cause it to create a new account instead. Some common mistakes are omitting or adding the 0 at the front. You can read more about it here: https://discord.com/channels/1069381916492562582/1462888142829912125
Thank you very much for all your support.
After removing the leading zero from my phone number, I was able to sign in successfully on the iPhone app.
The PC version took some time to recover, but it eventually started working again.
Now everything is back to normal, and I can return to my creative work.
I truly appreciate your help.
I often have some glitch, where after I create a song if I click anywhere so that it loads a new webpage, the song vanishes... It happens both when the song is generating and also after it has been created, anyone else has this happening?
So frustrating, makes me loose some real gems, and just to confirm it's not when the generation gets restricted... Restrictions are a whole topic of bugs for themselves, I literally can't use some of very common words in my language because of seemingly some artists existing with same names...
I published this and have been trying to share it with someone, like in discord, but it just stays buffering and never loads even after days. The page the song is on even doesn't show the cover art either and when trying to go to the link, it doesn't work. I don't know what to even do to fix this lels... This is the link/song.
https://suno.com/s/b1AomMcEobGT54By
Listen and make your own on Suno.
Hi all! When creating songs, the ending always seems to get cut off. Whether there's a reverb of an instrument or lyric or whatever, it always sounds like it's cut off like 2 seconds too soon. Anyone else have this happen? I think it started with v5, but I'm not 100% sure on that, but I'm pretty sure it didn't happen with previous versions.
Also, is there a way to add like 2 seconds of silence at the end of a song? It would really help from me having to do it manually using Audacity or whatever. It would just save time (probably for lots of people).
My songs get messed up again recently. Tried to add some vocals to a song and they get messed up pretty bad. Sections repeat or sound like a skipping vinyl record or CD. Audio quality gets muffled sometimes. Mashups still sound like they're mixed with too much gain and were squeezed through a compressor afterwards. Not all of them but most of them get a very high pitch "white noise" in the background.
I don't even file many bug reports anymore, because none of them got noticed in the past. I did three today though, but I don't have much hope for them.
[Bug / Critical Issue]
Account Plan Not Updating After Payment
Issue: I subscribed to the Pro Plan (Yearly) on Nov 25, and upgraded to Premieron 28th of Jan but the system shows my account as 'Free' tier.
I can access Studio but if I download a song I get "you cannot use ...because free account"
I've emailed billing@suno.com, support@suno.com, sent messages on Facebook and Instagram but have received NO response. Please help check my account status. invoice 4DZ72IVJ-0002
@sand depot ☝️
That's fantastic news! Thanks for letting me know it was resolved. I'm genuinely happy to hear that
Would be nice to hear anything from Suno about all quality issues that was introduced recently. Many generations have pronunciation glitches, often sound is very noisy and quickly degrade. I mean, song starts more or less ok, but in second verse it's getting very noisy and muffled. Stereo field also collapsing during song and it's gets almost monophonic to the end.
I understand that Suno native duration limit is around 2 minutes, and everything above that is extensions generated in 30 seconds chunks (that's why you can hear sound/arrangement change at 2:00, 2:30, 3:00... etc), and quality is degrading with every chunk. But sound degradation became so much obvious, it never been so bad before. I mean, I'm not telling that all generations have this problem, but it happens much more often.
Extend function not working properly - sound is different, arrangement is different, voice is different. Previously, I very often used Extend function, it was a much more consistent way to fix things in the song compared to generating replacements. Right now, Extend is not respecting original track well, it seems that original audio influence is too weak, but we can't change it since it's not configurable.
Would be nice to hear something from Suno, at least that you are acknowledge issues and working on fix. I understand you introduced new functions Mashup, Samples... but why damage existing old functionality? And not even a single word from Suno about this, just like it is an intended change?
What elder god do I pray to have my denoted singers keep their genders consistent? It even does a [both] when it's not supposed to.
Extend from X:XX time may be the only way.
Key is to get the song made, then try to cover it to upscale the sound.
I've had some success defining things such as Vocal:{male ensemble,verse prechorus chorus bridge},Backingvocal:{female ensemble,prechorus chorus bridge} But it's hit and miss if they stay separated or warp in and out of eachother. A single channel is easily to confuse, and why we can't have nice things such as 2 statically panned guitars where one plays rhythm and the other plays counterpoint. Check out the names each channel if you have a set of Stems laying around, if not I'll grab a screenshot when I'm done eating.
In regards to the absence of -in my prompt examples: Uppercase after the leading character because e.g. TimeSignature: becomes Signature:, and also using - as we expect to be able to use it, as it often is interpreted as a negative, meaning palm-muted becomes interpreted as -muted. So, yeah....
Extend is fundementally broken, please stop advising people to waste their money and our environment on things that are not fixes or work as intended. Thanks.
This! I have the exact same problems. If these quality issues are not resolved, the service is pretty much useless. I have already wasted thousands of credits. Suno should put in some effort in fixing the core featurs instead of wasting time and developer resources on gimmics, like samples and mashups.
All of my credits this months have exclusively been spent on testing, attempting to prove that what is being claimed about features and the quality of my prompts is true.
I'm not even mad about it at this point. What is making me angry is how unimportant we as customers seem to be, which makes me ponder if this is a short-term cash grab ran by a skeleton crew.
I'm sure their intentions were good, but inevitably got slapped by the industry. If this even resemble reality, it wouldn't justify ignoring paying customers pleads about getting what they paid for while being stuck for months as free users after buying a full year. We have a word for that.
Again, I press the team and staff. Talk to us. Tell it as it is and what your road-map actually looks like in terms of fixing basic payment and UX stuff as a starter... And I reiterate; you don't depend on your investors long term, but public perception - which is a direct consequence of - at the very least - things not deteriorating at an ever increasing rate as "new improved things" are introduced while core issues such as described yet another time remain broken.
Forgive me for having to say this supremely unhelpful phrase, nonetheless it is the truth as annoying as it is:
Extend works fine for me.
Then your expectations are exceedingly low, or your music doesn't rely on smooth transitions.
In this context, do you comprehend why it is at all broken?
If not, then that would explain a lot of the advice given by staff.
You don't hear it, or you don't care about it, while defending a platform promising entirely different outcomes.
What is your style of music? I do mostly hard rock and metal which relies on many layers and complex instrumentation, and no - extend never gets it right. EQ, gain and stereo spread etc. is always off. If you do sonically simpler genres like rap or accoustic, maybe it works.
Probably dubstep-inspired trap drowned in glitch elements. Which would probably sound even cooler with all these common issues coloring it.
Hi, i have same problem ----> #🪲┃bugs-web message
I've had the same problem for days. I reported above and I saw that other users have the same problem. I wanted to bring this to the Team's attention. I didn't mean to be rude, but it's not usable on Suno web browsers. I just wanted to understand if it's a bug or can it be fixed somehow. Greetings.
Multiple communities have confirmed that Cover mode is ignoring Personas since Tuesday. Resulting in generic humming and broken harmonies.
@solar grove Any suggestions? Or is my point crystal clear now? ☝️
7:59 bug still active. Any ETA on the fix?
My credits didnt update either but i have plenty from last month because i stopped using it due to some of the bugs like 7:59 bug which just burns through credits with unusable gens. I assumed no update fir me since the tech team did say i wouldn't be losing credits. one of my concerns was SUNO not being usable and rollover for credits was coming soon. i didnt want to keep using a broken product.I also dont want to lose my rollover credits either.
I'm not here to be your debate partner.
Hi I wanted to report a bug with the suno site when I try to subscribe to pro and put my card info in and what not it keeps declining my card now i checked my bank there not blocking it my account has money and the cards good and today also note its a visa card as well
PLEASE is it possible to have a statement on this Bug that I posted last week. Just to know if you're aware of it
##The inspo seems to be broken since the mashup update. I would usually select a playlist and inspo will randomly pick 4 tracks. Now the random selection doesn't work anymore and it would just pick the same first 4 tracks of the playlist.Can you please fix it?##
That's all well and good, I'm just seeking clarification if you comprehend the issue we are describing. Because if this is not the case, then there are grounds for further explanation past the detailed descriptions given so far at the very least.
I mean, if you guys don't understand the problem at its core, how are you supposed to effectively aid resolving issues? If this is also true of the developers, this would be a clear signal as to where this product is heading.
Keep in mind, that those who interact with the public, will shape public perception. And absence of meaningful responses grounded in a products limitations, does not foster trust. Do with this information as you will.
@sand depot, please. Take us seriously, if not for the products sake.
This doesn't feel right!? lol The version in the middle just continues like that for 7:43 minutes .... Beeeeeeping sound
It's a frequently reported issue which there have been zero commentary on from Suno. Started happening around 29th of January.
Yeah I've had the same problem in some other gens to, I think you're right on the date
One even beeped for about 4 minutes and then started the song...
I was sat there testing when it started happening and the initial reports started pouring in. Since then, what little interaction from the team also more or less stopped past sporadic payment issues. I'll restrain myself from speculating past this point.
Hmm that does not sound good
I can only speculate as mentioned in my previous response to staff, and will continue to do so until I am removed from a privately managed public space in hopes of some form of response or the issues the platform is facing is taken seriously and remedied. Currently this consists of blaming users for incorrectly using the platform.
To overpoplulated servers maybe? The GPU's glowing to hot
Do you have a feel of "when" it happenes? Like is it just a clean prompt from start or starting with a sample or remix/mashup or something like that? (My gen that faulted was when using a Sample from another generated song I made a few weeks ago)
Nope, there's no clear pattern to it past a lot of strange stuff starting to happen at an increasing rate after the new vocal-specific personas were introduced. I've had somewhat less issues after abandoning Personas all together, but it also happens extremely frequently when using things which color context such as Inspo and Covers. I haven't had energy to test further after realizing that the issue weren't only affecting me and thus not a consequence of my prompts. It also seems to affect plain-text prompts as well from what I've read.
Geez this does not sound good, I've only had it happening 3-4 times before and since I generate a lot of gens I did not really care about it then but. If it is becoming a more frequent happening it does not feel good.
I mean, slap on defaults, don't write anything anywhere and start generating. I would not be shocked if it's happens without any input from the user past clicking Generate.
In fact, I would be pleasantly surprised if it didn't, as this narrows down the likely source of this particular issue.
I don't want to contribute to a negative outlook on this platforms future. I'm just speaking my mind as I tend to do. If I'm wrong, it should be fairly easy to disprove my theories.
yeah that is what I was aiming about... If there is a chance to narrow the errors down to something specific.
I just love the new stuff the Suno team put in there with Sample and Mashup and so on. I've been hooked for a long time of generating ideas and this took it to being almost obsessed with it hehe
If I click on a song image in the playlist to play the song, and then I click again to pause, and then to play again, the listen number increments. It should increment only if the song is played from the start.
Indeed, the features are novel and would be great additions to the service. But it seems they don't put much effort into testing before pushing new and broken features to production, celebrating them as if they are working in any capacity past bleeding peoples credits for no good reason and doesn't affect other features negatively - which they clearly do.
A beta stamp doesn't make it okay.
Yesterday it wasn't incrementing at all, so I guess they're working on it.
Clicking on the Greek language on trending language selection breaks trending view, it stays on global displaying no results. Emptying the cache doesn't fix it, it remains broken for that account. This happens on all web browsers.
Hey everyone!
Just wanted to drop in to let you know that we appreciate you taking the time share what you're bumping into. Please know that all surfaced concerns are taken very seriously. Anything that can improve your experience is super important, we are building for you after all!
To help us resolve issues quickest, here are a few things to know:
Reach out to us directly through suno.com/feedback. Submitting through this method helps us correctly identify and route your report so it gets to the right people. I've updated the guidance provided here to walk you through the process with some helpful tips on what to include.
When reaching out here in Discord, two things help a ton:
- Keeping this channel focused: This helps us get what you share in front of the right team faster. Totally understand it can be frustrating when something isn’t working the way you expect. It matters to us too! We want Suno to feel smooth and comfortable so you can stay in the creative flow and have the most fun when creating your awesome jams. To help us sort things out quickly, please keep posts to just the issue you’re bumping into, the key details needed to investigate and assisting each other to debug them in a helpful way.
- Key details needed to investigate: If you’re reporting a bug, steps to reproduce are the biggest help. The clearer the steps, the faster we can confirm it and fix it. You can find some guidance on what to include here, but I'll also add it to this message and pin it incase it's helpful!
When sharing a bug, please include:
• Clear title: short description of what’s wrong
• Platform/Browser: Web / Tablet / Mobile + Safari, Chrome, etc.
• What’s happening: 1–2 sentences
• Steps to reproduce: numbered steps we can follow
• Screenshots or recordings: super helpful if you have them
Also, never hesitate to reach out my way if you need help with knowing what to share or where to share it best. Here to help!
Hi, @sand depot. Why would we use this when it is well known that turn-around can be anywhere from weeks to months, where you are the sole exception providing your much appreciated aid in regards to escalating issues? Also, why offer the same support by means of two channels in the same place one is expecting to receive as immediate turn-around as possible, and word things in a way indicating that one is faster than the other?
Second, I appreaciate your reply, but again - why would we aid in keeping valid criticism hidden from plain view?
What is our incentive is what I'm asking here.
To be fair and as I tend to say as to not burn people out (entirely, or so I hope anyway), I don't expect a detailed response curing all wrongs in the world. But these are points which I think are extremely relevant to almost all parts of this opertaion which should be put under thorough internal review. Again, thanks for making yourself available.
There's a lot more that goes on than I think is understood. The entire team is very busy helping. I witness this everyday. It takes a lot of hands to help a lot of people. And all of them care very much about both the members and their experience.
Criticism re: that's just not the intention of this channel's design. This channel is designed to surface bugs to help us spotlight them is all.
Feel free to reach out to me via DMs though to continue this conversation if you'd like and let's leave this space for bug reporting. Would love to better understand all ways we can improve. That helps a ton.
Thanks for your understanding!
I'm super aware of the scale of this operation, but it's as if you guys don't understand that one relies on the other. If everyone is to sneak notes under the table to you, we can't off-set the workload by co-operating to identify causes. The logic presented here is (and forgive me for saying so) counter-productive.
Same here! It’s absolutely broken. And we are paying for it
And again, I can't help myself returning to the feature bloat situation, which is breaking other functionality. An analog to this would be this channel's intended purpose, vs what it has become - our only meaningful place to collectively debug or at least understand that the issue isn't unique to any given user.
I mean, if I dare speak about anything in any other channel which doesn't resemble the perfect output people believe that they are producing, and is a consequence of their particular style of prompting; I'll get dog-piled to the point of ridicule.
I request you reread my original post. I believe what you are describing falls under the line of keeping the channel focused. Which is indeed what this channel's design is for. Criticism and this kind of conversation, not so much
So I can better understand, can you outline what in particular is problematic about my wording?
We aren't discussing bugs 🤷♂️
Alright, so it's not related to language per se, which was how that was initially interpreted. All good. My bad.
No worries, seriously. Thank you so much for sharing where we can improve. That matters so much!
I am compelled by nature, not much to thank me for. I want this platform to succeed, believe or not. But it takes a lot of people to tango.
Can we have a channel where we can actually freely discuss amongst ourselves, as to make this channel serve its function better?
That's a very super interesting idea! I've actually been thinking on how to approach such things safely. Would love to hear your thoughts on how to go about it. Please do feel free to DM me what would help best with this.
I'm going to reshare my original post, this convo has created quite a wall ha 🤦♂️
Hey everyone!
Just wanted to drop in to let you know that we appreciate you taking the time to share what you're bumping into. Please know that all surfaced concerns are taken very seriously. Anything that can improve your experience is super important, we are building for you after all!
To help us resolve issues quickest, here are a few things to know:
Reach out to us directly through suno.com/feedback. Submitting through this method helps us correctly identify and route your report so it gets to the right people. I've updated the guidance provided here to walk you through the process with some helpful tips on what to include.
When reaching out here in Discord, two things help a ton:
- Keeping this channel focused: This helps us get what you share in front of the right team faster. Totally understand it can be frustrating when something isn’t working the way you expect. It matters to us too! We want Suno to feel smooth and comfortable so you can stay in the creative flow and have the most fun when creating your awesome jams. To help us sort things out quickly, please keep posts to just the issue you’re bumping into, the key details needed to investigate and assisting each other to debug them in a helpful way.
- Key details needed to investigate: If you’re reporting a bug, steps to reproduce are the biggest help. The clearer the steps, the faster we can confirm it and fix it. You can find some guidance on what to include here, but I'll also add it to this message and pin it incase it's helpful!
When sharing a bug, please include:
• Clear title: short description of what’s wrong
• Platform/Browser: Web / Tablet / Mobile + Safari, Chrome, etc.
• What’s happening: 1–2 sentences
• Steps to reproduce: numbered steps we can follow
• Screenshots or recordings: super helpful if you have them
Also, never hesitate to reach out my way if you need help with knowing what to share or where to share it best. Here to help!
Thanks for your advise but i deleted at last the song, as I've tried different way to remove the bug, I did not arrive to anything, then I pub into garbage.
Hi I wanted to report a bug with the suno site when I try to subscribe to pro and put my card info in and what not it keeps declining my card now i checked my bank there not blocking it my account has money and the cards good also note its a visa card but still was giving me issues
No, it's not in mp3 format; the bug occurred when generating the cover version of the song. In any case, I deleted the file because I tried lots of different options and none of them worked.
Créate with grok a small video, it is free and the you can upload to replace suno image.
That_Was_Always_Allowed-Jerry_Smith.gif
Many of mine I tend to make with PixVerse or other generators because I know them a bit better and can get closer to what I want with them most of the time.
Hi David. Many thanks. THis is very useful to know. But can I just ask, will you and team inform us specifically when the covers/persona issue is resolved in order for us to save wasting generation credits on results when know will have voice errors and the audio noise at beginning and end in them? Thank you.
Recently, some songs come out perfect, but some have many glitches or dropouts in the music, making them unusable. We should at least get credits refunded for failed songs like that.
Anyone else able to submit feedback through the provided URL? (@sand depot I've mentioned this before, btw. :D)
Failed to provide some additional info. Quickly summarized: Straight up Chrome, no extensions, default config, blank user-profile, only logged in to Suno as prior actions. This has been a thing since my first week on Suno which I believe was October-ish. Reason for noticing were other issues persisting to this date and as described in my many rants related to audio quality and inconsistent instrument/amp/room dynamics.
Hey Fused! Thank you for this thoughtful response. I'll surface this interest in better visibility for bug fixes 👍
Hey Geir, thanks for flagging this! Trying to reproduce this on my end with same set up (Version 144.0.7559.110 ) and unable to. I'll keep trying. Very odd that this is occurring. Looking forward to better understanding what's going on here
Tell me what you guys need and I'll gladly provide.
Not that it should make or break a website, but I'm running Arch, btw (*ba-dump-tss*)
When making a song with Suno “Cover,” I’m running into a problem: whenever the vocal reaches high notes near the end, the audio quality gets messy and distorted—both the vocals and the instruments break up. The same thing happens with other songs too.
Do you know what usually causes this, and what I can do to fix or reduce it?
bunny-hopping off my last comment, i seriously think theres something going on w/ Suno's AI. the quality is either mediocre or even bad, or sometimes it just doesnt follow my prompt at all. at first i thought it was just lyrical songs kinda acting up, or maybe it fixed itself since yesterday, but even though a couple hours ago it was fine now the instrumentals-only songs are acting up a bit too. it's like the overall quality of the AI is degrading. i know v5 is still in beta technically, but it was perfectly fine and this seems to be a recent issue. if anyone knows how to, please push for this to be looked into by the ppl behind Suno because i'm seriously starting to think something is wrong w/ the servers or AI itself
Scroll through this channel and you'll find that you're not the only one.
if you have a create window open, and make a cover of another person's song creating a your new song, then select cover of the new song to further refine it with a cover, it breaks the re-mix link that is supposed to show the other user's original song being covered....At least that just happened to me....and I suspect that is what happened on previous covers that didn't link properly for me...anybody else run into similar issue? ....if that is that bug......pretty weak sauce suno...you need to start offering bug bounties to your users's if your not going to properly vet stuff before you release it.
@sand depot The model feels completely "lobotomized" since the Jan 29 update. (Affects Extend & Cover)
Extensions are disjointed, Covers are off-key, and the overall output sounds like generic nursery rhymes. It creates a jarring disconnect between the original audio and the generated output. It’s like the model has lost its Context Awareness.
Please stop wasting our credits on this downgraded version. Is this a bug or a permanent reduction in compute power? We need an ETA on the fix.
Supplemental: Extensions that makes no sense (transition, dynamics, even over all sound) have been a thing since I joined in Octoberish.
Generated 6 times. 12 songs. 11 of the 12 songs, some different prompts entirely and most tweaked to try and find the issue, all getting the 7:59 length bug. I just wasted 60 credits just this time alone, let alone the rest of today.
I have already done all the available "fixes" online, this is ridiculous and stressful. It doesn't matter what I do and don't do, it's entirely random and happening more frequently. The first few days I used Suno this wasn't happening. The past two days? It's nearly constant.
And to save the audio, I have to spend more credits to try and save the song, usually by trying multiple times.
I am using double to triple the amount of credits I should be trying to generate songs that weren't having an issue a few days ago, or having to spend credits trying to fix them for a bug outside of my control.
Two completely different prompts and this issue is persisting. There is no constant or clear trigger.
With pro plan you get 2500 monthly and a decent chunk is getting stolen by a extremely annoying issue.
Got told through a support ticket that this is monitored closely but almost every song I am doing is getting this issue, even songs I did earlier with no changes is suddenly getting this length bug.
And then on top of that, generated a set. Loaded for a long time, then simply disappeared but my credits were still taken even though the pair generated was not in my Creations, Library, or Deleted.
As an edit, this is happening both on IOS and browser.
I cannot make edits in the lyrics box anymore without the text cursor jumping to the bottom while I type. I learned a long time ago not to write lyrics directly into the text box, so I usually type them up first and copy paste, but I can't even make small edits now without it jumping around every couple of letters. It is absolutely infuriating and I don't think it is my keyboard or computer because I don't have that problem anywhere else, not even in the style description box.
Its not letting me make a persona for tracks no matter how many time I cover the cover of the cover of the cover of the cover of the cover of the cover. The song doesnt resemeble ANY KNOWN SONG IN EXISTENCE. I DEMAND A PERSONA FOR MY SONG.
The website isn't allowing me to download any songs, I have tried on my Pc, and on my iPad.
Make a mash up with two of your covers. Then apply a persona
So there is apparently an issue with system overuse. I have been experiencing forms of deteriorating audio quality and consistency since Christmas, which makes Suno unusable for me. I suppose the Suno team is working on this 24 hours a day. At least that’s what I would do after releasing a beta. When can we finally expect a change to the better? Regards.
Well, I think there are several different sound quality issues.
-
Prime hours issue. This one exists for a long time. For example, if you are in Europe you may find that sound quality at the morning is better. Because, when there are evening in Europe, it's already a morning/day in US. So, when two huge markets are awake, it seems Suno reduce number of inference steps used to generate tracks (assuming that Suno is a diffusion model).
-
Long track (longer than ~2:30) sound degradation. It seems that natively Suno can't generate longer tracks. So, Suno generate long tracks by using 30 seconds extensions (chunk). With more chunks an errors are accumulating, so sound is drifting. I don't think that this issue is solvable for now, this is just where a current technology is.
-
Recent (29 jan?) update sound degradation. Many generated tracks have bad sound - stereo field is narrow and collapsing to almost a monophonic sound to the end of song, high freq's are very noisy. Also, tracks often have pronunciation glitches. You can also hear that those tracks have bad prompt adherence, no matter what you prompt, generated track is influenced significantly by contemporary mainstream pop. But! Not all generated tracks have those issues, like for some generations another model is used (maybe a canary deployment of WMG model?)
Note, that covers are more sensitive to sound degradation, because model's context is already partially filled by original song.
So, when you are "lucky" to have all 3 issues at the same time - you will get exceptionally bad sound. Enjoy! 😁
As a paying customers (I'm on a Premier plan) we demand a transparency. Be fair with us, just tell us what is going on. I stopped using Suno for now and I will probably cancel my subscription, because my workflow is completely ruined by last update - bad sound quality (especially in covers), Extend function don't respect style and arrangement of the song, even voice intonations are different.
Hi, guys! Ran into an interesting bug... I had an ongoing Suno Pro subscription which I tried to upgrade to monthly Premier version. However, the process failed - first twice silently, and then with an error message saying sth like "cannot upgrade to Premier in the middle of subscription period" (or sth).
I was able to access Suno Studio once, but now after loggin in / out it seems like my subscription has reset into free version
... actually, even more interesting: it says "upgrade to pro" in the icon, but I'm still able to access Studio. However, when I try to export anything it asks if I need commercial rights, which indicates to me that my account in in somekind of interesting limbo
I am still unable to get into my account, so i said screw it, i have some ideas and some time, i'll just setup a new subscription and boom same issue. Anyone have any idea why i'm getting
"Authentication state is missing. Please try again."
Same issue for two weeks now. I pay for pro and can't use ths system at all.
For me, it's 95% 7:59 bug with the occasional clean song. Completely unusable.
There is NO ETA on the fix, and SUNO talks as if it's only an occasional overloaded server issue. I've tried it all days of the week and at all times of the day and night. It does not matter.
When I don't get the 7:59, extremely rare, about half of the few that "work" still sound terrible after the first couple of minutes. Fuzzy and distorted more and more the longer the song.
Ok. My sub rolled over yesterday, and i lost 1500 credits i havent been using because of the 7:59 bug. i was told by the support team that we wouldn't lose credits and they were working on the problem. My credits should have rolled over with last month's credits, i cant use because of this bug. I try generating and looking at discord every day for answers and crickets. The bug still occurs and no fix in sight and no word. I want my credits back until this is resolved.
Yeah, support does respond, so that's nice. However, there's no fix; I lost credits too, and I was told I wouldn't. They might as well send over someone from SUNO to just steal my money, haha
Supplemental: Loss of credits = Waste of real-world resources. I opted to stop after spending ~5.2k of my Premier sub solely on testing. Not trying to measure up against anyone's efforts here, just want to point out that it doesn't matter if it's 60 or 5k. If I get bored enough, I'll run the numbers to see if the stats align. :D
yea, i see that now (in public servers i just lurk and keep up w/ announcements and nothing more, so i dont read message logs). glad to see i'm not the only one...because i thought i was going crazy! everything was perfectly fine about a week or so ago (around Jan. 27th) as i was able to generate good songs that followed my prompts perfectly and didnt sound bugged out or low quality. i saw one message after you, pinging one of the devs i guess, that said a recent update may have bugged out the AI or something? Jan. 29th they said, which lines up w/ when Suno has been acting funky. i hope those behind Suno get to the bottom of this and fix the issue because it's practically unusable now. one moment it cooperates and the next it kinda just does whatever it wants, ignoring your prompts entirely
time for a break from Suno i guess lol! may have to discontinue my Pro and await an announcement, or lurk in here, to find out if they ever fix the issue before i continue using the tool
My conclusion is the same. No point in wasting water and power at this point. Hopefully they'll find a fix eventually.
I want to add one more thing to what was already said. It seems that Remaster function has degraded as well. I previously used Remaster several times to improve replacements and it helped, but now I see that Remasters sound very muffled and losing many details of original track. Overall sound is like 96kbps mp3.
Supplemental: It seems to have 3 outcomes.
- Wildly distorted audio akin to running it through a digital compressor with way to fast attack (1-2ms) and equally fast release, which tend to cause distortion artifacts very similar to excessive gain.
- Dynamics which resemble a pulse due to compression with an extremely high ratio and low threshold. Think akin to EDM where the beat will push everything else into the background.
- As described above, where an apt analogy is nostalgic memories from my early years, listening to an AM radio station while traveling through a valley with poor reception where the signal is consistently weak, but strong enough to not turn into static at any point.
All of the above too seem to get worse and worse over time, as I've described countless times where audio becomes more and more distorted for each finished bar/part of a song, and if the song is long enough (~3+ minutes, roughly half of the available context window) will distort itself to the point where it can not be used in any meaningful commercial sense past as a draft.
Upon further rumination, perhaps increasing the token limit to about 2.5x of what it currently is and enforcing a maximum length where the model is instructed to conclude the song prior to this limit. Essentially targeting 7 minutes while leaving overhead may have a snowballs chance in heck at resolving this. If I were to describe the sensation I get while listening to Suno's output, it feels constrained and in a hurry to finish - which is very human in this day and age. :D
(Important edit regarding abrupt endings and instructions: With strong emphasis of letting dynamics such as reverb, sustain, delay, and so forth play out in full and never fade out at the end. Also introducing Suno to the notion that feedback can be muted by a palm gently being rested upon vibrating strings or slowly faded out through as a guitarist would through the volume knob or foot-pedal. No need to chop off reverb, it sounds bad.)
not entirely wrong lmao!
#analogies4winz
the songs either start and/or end in a jarring manner or very abruptly, and the quality is swings-and-misses as of late
it's like ofc generations in general are swings-and-misses, thats AI for you when it comes to this stuff as the AI is trying to generate what youre asking but it isnt "smart", it's winging it basically even moreso than a real musician does when theyre making a song. a real musician at least has a general idea of what they want their song to sound like and has the musical knowledge to get there by playing around w/ notes, an AI cant do that, at least not quite in the same way. it's basically taking things out of its database that are related to your keywords, throwing it at a wall and hoping something sticks
I wouldn't expect perfect outcomes considering where we currently are with respects to NLP, honestly I'm surprised at where we are. It is sand in a weird state of hallucination after all, however, I believe there are basic changes that could be made (Ref. fixing the glaring parser issues I've described at least three times) to how we as users interface with the model itself, along with constraints introduced by the pipeline which is causing a lot of the artifacts that we are hearing and result in the over all sound quality to nose-dive. Ie. getting a model to spiral through text is extremely easy. And the confusion caused by what I suspect is poor parsing of our inputs, certainly does not contribute to Suno's "sanity" as it's drifting through simulated multi-dimensional space attempting to adhere to our strange requests. </supplemental>
yea, absolutely. ppl who already write lyrics and compose music probably have a much easier time knowing what to put into prompts to generate what they want when they want to enhance their works, but for the rest of us plebs who just use this tool for funsies and dont have that same education/talent it's not that easy lol! the AI is also still relatively "young" and "learning", so it's still adapting to our requests and building off them to further improve itself to make better generations for others
The notion of a KVCache which span users suddenly made an appearance. It would explain why we seem to be haunted by similar issues across gens and prompts, but can be "broken out of" (to an extent, not completely) by means of things like re-uploading exported tracks. Sorry, I'll stop speculating and attempt to let this channel adhere to it's intended purpose as described: #🪲┃bugs-web message
(Indirect protest to the notion of this being a productive guideline in absence of a bug tracker)[lingering delay and reverb,fade out on silence] .-\n.-\n.-\n-
I can't access my acc via google login,
i have the same issue, can we get an answer?
Hi Ho, any chance to bring back the original 'Remix' functionality? I cannot cover 7 minutes of my recorgigd because the model starts to halucinate ofter 2 minutes.
Same situation here ... after upgrade to premier ....
Am I hearing "guard" instead of "gard"?
Why is it a problem when i use 4k of characters in lyrics, it goes every time wrong with the end part of the song he sings text unrecognizable in Dutch language, he lose words or sentences, right no i get many creation with time 7:59 al bad songs because at 7:30 he is halve why my lyrics this cost me credits for nothing, pretty strange that i can put in 5k characters but the problems alrady start at 4k below 4k it mostly goes well
btw this is al v5.0
lol now make one in V4.5 and get the same destroyed songs and 7:30 halve way my lyrics this gets annoying, from 3k credit to 1.9k and still not make 1 song that is ok
What is wrong with SUNO? All covers i try to make are just Big mess. Distorted sound, sings wrong words, volume goes up and down. Makes it's own styles, even in the middle of the song. Everything is messed up! Trying to make 30 covers and every one of them are now in Trashcan... Can someone from SUNO tell us whats wrong and are you gonna FIX it???
I'm having a very strange issue. I'm trying to make a mashup between 2 songs, and 90% of the resulting songs are exactly 5:17 long, they cut off abruptly at the end instead of ending naturally (like they're cutting off in the middle of the line) and it's dragging the whole song out in general because the source songs are under 2:30 each. I feel like this has to be a bug because it keeps consistently happening. Who do I report this kind of bug to, and is it possible to get a credit refund on these bugged generations?
Same here...frustrating. 4:80 is the time here 🙂
okay so it's not just me... something strange is going on...
to try and do some troubleshooting here... by any chance are you using 1 song with lyrics x 1 song with no lyrics? cause thats what i'm doing
Try to make some covers....everything is messed up there. Song is different in style...distorted etc.
well mine sound amazing, but they keep cutting off
i had a similar issue, you have to use extend to fix the issue
2 song allmost the same. Same lyrics in both. Cover was so bad to i try mashup...but....nope
but yea, Suno's been having varying problems
not a good solution. i'm making a mashup of 2 songs... if i extend it, it won't sound like the original songs anymore
youd have to trial and error it unfortunately and hope one generates to properly complete the flow
if the song sounds slightly off, maybe try remastering afterwards?
maybe not now due to the degrading quality at the moment
it's occasionally generating fine songs, but its also consistently generating broken songs in the exact same length
Somebody says that extend having issues also...
degrading quality?
thankfully you havent had the issue from what you said
but yea, if you read up in this thread, everyone is having different problems
i know, but i'm just saying that fixed it my end
what's this degrading quality thing?
as in like ppl are getting muffled audio, jarring starts and endings, crunchy audio (as in like it sounds almost staticy and grating)
and even outside of audio, ppl like myself would have our prompts ignored as the AI does its own thing
hmmm
it changes the genre of the song mid-generation
what you had that the song suddenly ends
i've gotten SOME stuff like that, but i figure thats just AI being AI
i've also had another issue where if a line has repeating lyrics with the next line, it's skipping the line
well yes, ofc, but it's gotten worse since a recent update on Jan. 29th according to someone earlier
also wait are you talking about the songs getting muffled sounding during exporting? or when generating? cause i noticed they sounded more muffled on export
while generating
whats strange is that the inconsistency of the AI is inconsistent
one moment everything is fine
the next everything is just going haywire
it doesnt generate properly
yeah but i've noticed that a song i generated like a week or so ago, when i exported it, now sounds terrible
ah that i wouldnt know of since i exported my songs awhile ago and backed them up already, but it could be part of whats plaguing the AI right now too idk
hmm
right now cover generation is making one song fine, and the other song without audio..
Cover and Replace in Studio have become useless. 9 times out of 10 I get full band instrumentation when trying to cover/replace a single instrument track
Hey guys, how do I cover or remake my song while maintaining the same singing style and instrumental? Can you share a prompt?
Hi. All the songs that I created are distorted and muddy. What has happened? Whats the problem?
Covers are also messing up lyrics. How does “They’re not sorry” written in the lyrics and in the original become “Then a story” when cover is applied?
Covers are fundamentally broken.
@rocky ledge I want a refund because my account is now gone!!!!!!!!!
Hi Suno team / community,
I’m having multiple serious issues with Suno and I’m honestly stuck at this point. I’ve been emailing suno@support for over 2 months and haven’t received a single response.
Main issue (critical one!). In Studio I have NO sound, while the same tracks play normally in Library.
This is NOT caused by:
-volume faders
-muted tracks
-browser cache
browser itself (tested on Chrome and Firefox, other computer, mobile - same result)
There is simply no audio output in Studio, which makes it impossible for me to work — even though I’m on Premium.
Other recurring problems: Get WAV / MIDI sometimes does not work, but credits are still deducted (50 credits). After that, I still have to spend additional 10 credits per track just to regenerate MIDI. Frequent bugs and inconsistent behavior overall
At this point, I honestly don’t understand what I’m paying Premium for, since I can’t reliably work in the Studio.
My questions аre: Is this a known Studio audio bug? Is there any fix or workaround? Where can I get actual support if emails are ignored? If this can’t be resolved, what is the proper way to request a refund?
Any help from the team or community would be greatly appreciated.
Thanks in advance.
The video generation UI doesn't update the cost for generating video based on the preferences stored from last use, leading to it sayaing 50 credits for a 10 second advanced generation. The button only updates after it is pressed. I've noticed this but not thought anything of it, but a user came into the general chat upset about this issue
No chance, Suno. I'd continue burning credits to identify issues for free if I get to retain distribution rights. Past that, it'll be covers for songs I'm 100% sure are generated on a free account, as those would belong to Suno anyway. This is silly. Get to work on things that matter. Stop all the new feature dev stuff. Fix bugs. Jeepers. :D
Unrelated to ongoing chatter: I'd love to see commentary on why the Extend feature even exists at this point. What's the technical reasoning which justifies this past "pls spend more credits"? I'm having bad thoughts when ruminting upon this subject in light of certain contexts which have been discussed, but leave the rest to the imagination of the reader.
- Suno Jones 🤪
I'm sure we all would
Past clearly being heavily edited, there's so much of everthing being discussed audible when skimming through on laptop speakers. I commend you on your efforts to stretching out songs, but um. I don't want to make enemies today. :D
these are just straight extend work
there might be eq on some
I could tell you how extend works what it's good for
From September 2025
what you need to feel out in the model to use it or know when to use it
what date would you like one from
Jan 30th. Have at it.
I see messages regarding things that have been addressed only a day or so before; some users don't even bother reading any posts here before they launch into their issue(s). Further, too few details are given from them that might help the devs know what is causing their issues. For example, mispronounced words, wildly changing genres mid-song, not following prompts, etc. Devs need to know if it's a cover/extension/new generation, or if you've altered any of the sliders (always risky and unpredictable behavior doing this!), or if you're using overly wordy text in the styles field, etc. So many possible variables, yet you're only detailing the output end result. The only way they can solve these issues is if they know what you are doing to get those results. That said, most new posts from too many users are never going to read this msg, they're gonna jump right to the prompt and start unloading their issues with all the wrong details. (sigh)
I honestly don't believe for a second that this is v5 and straight extend.
There's no way.
Jan 29th, no edits, extend only: https://suno.com/s/mt67xqXDAQ06Mj9B
I call shenanigans.
No stamp.
We can hear in the voice that this isn't v5. This is an old model. I'm not here for those. The sample you're providing are valid in the sense that yes, they are extended. But all of the issues that have been brought to light above is present. So I can't say you're wrong. But I say, this is not at all relevant - even to my speculation as to why Extend exist and seek technical reasoning behind this thing. I did not see 45 minute songs coming, but more along the lines of as a mitigation to songs ending early and so on. :D
You can be upset and not believe, its not a skill issue the you can't do this, it's a skill I've aquired that I can. The tool has nuances and power
I'm so sorry, man. But I'm not at all upset, please stop assigning emotions to me. :D But there's just no way that this is plain extend without intervention. None. :D
Do you want screen shots of every intermediary
That still doesn't prove anything.
Madness
Yea, this server is an interesting place. On that much we can agree. :D
I mean, if you've cracked it. Do a writeup or video tutorial so that we can all learn from you :D
You believe I arranged on jan 29th a fake v5 with fake v5 intermediaries to hide the use of another model?
I have a detailed entry in workflow on prompting
No, I'm saying that I don't believe it to be straight extend without intervention. Literal.
Which doesn't have a focus on extending so I'll try to make a post on extending
Sweet! Looking forward to it.
"How to use silence to hide transitions and abrupt changes in dynamics 101"
I mean listen to the song named "↻𐑛𐑴𐑯𐑑 𐑕𐑑𐑪𐑐 𐑑𐑷𐑒𐑦𐑙 [0]". It's so distorted at the end that one can barely make out what's going on. And this is the outcome that is being presented as defense of a function which is proposed as a way to get around issues in generation. I withdraw. I don't know how to communicate without this ending up being interpreted as a declaration of war, etc. :P
It's exactly what I've been describing all along. And the attitude demonstrated by those who claim nothing is wrong, is to the letter. There's nothing special about the way this song is prompted. It's fairly straight forward, and following most guidelines. Where the result is progression not even close to what is being described in the lyrics. So I question a lot about the authority being insinuated with a sentence that leads with "It's not a skill issue", and ends with "I've acquired a skill". I yield the floor. xD
I mean, there's confirmation bias and then there's a seeming majority of the Suno community. Please forgive me, but this is funny. :D
there is a feedback button on the website where you can submit bugs and other various feedbacks
dude, you grilled me mercilessly got many examples and you don't like the breakdown on one song so you sing an epic about how it's all dumb
meanwhile I'm spending my time working on a workflow for extending
like maybe baseline contentedness with life is a contributing factor here
Good on you. I prefer to spend my time on things which efficiently produce meaningful results in the present, and do my best to ensure that issues which is counter to this result is remedied long term. Continuing at this point is akin to contributing to the planet being set on fire for little to no viable result. I'm not judging though, this is just how I see things. A view I've acquired, if I may. xD
The feedback menu has submit greyed out if you don't select a different platform then Web then switch back to Web even though Web is the default....suggestion...the what platform box should be blank forcing the user to select it if you are not going to recognized the default value of Web...this seems like yet another bug in the website on the feedback menu to report bugs ...no less....i could not figure out why submit was inactive the first time for the life of me since all the fields were populated....this is a straight up bug of having to toggle the platform to something other than web than back to web....selecting web from the drop down w/ changing to something different first results in submit still being greyed out.
But not inaccurate?
there are few people on this server as experienced with using extend to produce surprisingly good results as I am and I tried diligently to satisfy your demands but you accused me of using forgeries and heavy editing and then said the intentional distortion in one track is why the whole thing is nonsense
Sorry, you've had people point out to you that the issue you are having is not one that everyone has. Many many users of Suno use it without the issue you have.
Your response? Everyone using it without issue are doing it wrong.
Everyone, ok. Very objective statement considered what's found in this channel.
It's as if I'm the only one. Lol. :D
Going by the screenshot you posted above you've cancelled your account. You say the tool doesn't work for you. Remind me again, why are you here?
Because you're certainly not here to use Suno.
Right this moment my ape brain has me here paying attention to this instead of making a document for anyone to read to help with the issue you claim to want fixed
I extended this song nine times. No issue.
I have outlined time and time again why I am here to the point of suspecting even you knowing, including that this is counter-productive.
There's not more to say here, really. You guys enjoy your alleged success. I'll be sat here staring and hoping that someone competent enough to resolve the issues outlined time and time again will do so. Then I'll return to verify, and remain a loyal customer if indeed the quirks this platform is facing are resolved. :D
your camping inflates your utility as a feedback metric like a barnicle on a port meant to funnel water for actual purpose
but this isn't a place for conversations
i get your point, but the thing is it's inconsistent. theres no way to replicate most of our problems. so it's hard to explain whats going on in detail when most of us dont know whats wrong. all we can describe is what we're getting and thats that. besides someone did give something specific: someone mentioned they noticed these issue since the latest patch on Jan. 29th, and everyone else having problems agreed thats when they also noticed the issues, so it lines up. if it was only a couple users having these problems, thats one thing, but theres such a large pool of ppl having one or multiple issues w/ Suno at the moment, and it all started at the end of January. so youre calling all of us liars/overexaggerators?
there is significant overlap in the general learning curve and the balances impacted by even subtle updates
it is like getting sea legs while also learning to sail and the ocean changes physics a little
So there is always a component that is a new model tendency and then a component that is.... not skill... just time adapting
many highly structured approaches insulate people from feeling model changes
this leads to a gap between experienced users with niche methods and then experienced AND inexperienced users
who collectively have workflows that have been interrupted, in, as you astutely point out, ways it's hard to replicate
i personally didnt have these issues when v5 first came out. the only issue there was was that the AI didnt follow prompts correctly, but thats a given since that version was still gathering info for its new database (i'm not tech-savvy at all, so pardon me if this is false info and correct me please if v4.5 and older versions' databases rollover). everything else when using v5 sounded fine; as promised, the vocal and audio quality sounded much sharper and more authentic. as far as i'm aware as well, this isnt a new version, they just added a new tool. yes i know updates are also considered "versions", and thats probably what youre meaning, but even smaller update versions werent this bad. whatever happened during this recent update completely discombobulated the AI
Then you actually DIDN'T get my point, or you wouldn't suggest I was calling anyone "liars/exaggerators" (if I was saying that, I would have said that!) Further, I never stated to whom I was referring, simply 'many users', IIRC. I never said "replicate" your problems, simply that you give some useful details (ie, Weirdness 75%, using audio upload, etc.) as opposed to "I've wasted 1000 credits and my song keeps doing [this] and [that]...is Suno going to fix this?". Never said for you to detail "what's wrong", but instead, what you did to achieve this undesired result. When I have issues with Suno, I backtrack, change prompts, experiment based on information from various user guides, etc. to try to achieve better results. We (should) all know both AI and Suno does not deliver results via direct orders or commands, so we have to up our game to figure out what does get results. I save all of my prompts externally and use the ones that are most effective on similar songs, which works great. You have to be interactive and adapt to the ever-mysterious ways of this technology. And not be so defensive!
sorry, i didnt mean to sound defensive. my point still stands though: it's impossible to replicate our issues
youre asking what it is that caused the problem and we dont know
one moment it's fine and the next it isnt
And I will stand by my response, that I never said for you to replicate your issues.
no, not replicate perse, but you did ask for specifics and not understanding that doesnt seem to be a factor at all because our issues seem random every time