#📚┃suno-school
1 messages · Page 156 of 1
I guess I just need to find a good way to get a foundational idea for a song, that isn't generic. Rather than expect Ai to generate better lyrics from nothing.
You can just like, go find a song you really like the lyrics of, grab a piece of those lyrics and then ask ChatGPT to write something from those lyrics, but prompt ChatGPT to encompas a certain emotion, style, sound, or whatever
Thats what I do.
It will always be a little cheesy, it’s up to you to make the lyrics it writes burn.
Sometimes if you ask for variations it will also give you four or five options each with a bit of gold in them to put together
An example would probably be needed to help further. As others have said having that much problems with getting only the desired vocals is kind of weird
Ok that's definitely what I haven't done and could try, is including those descriptive factors.
You guys are giving me good ideas to try, thank you!
An example?
Of the full prompt. So others could check if it behaves the same. Or make better suggestions. I'm currently not at my pc so I myself can't help probably
Sure. I can provide it.
Just did a head count. 13 gens.
But then again, even if in any prompt that I may have missed, there was "no female vocals" written, it still comes across as extremely odd to me that when instructed, the AI completely ignores multple enforcing instructive rules. Even if the AI goes a little wonky because it detects the word "female" once in the prompt, it completely ignores multiple instances of "male vocals" both in prompt and in meta-tags to direct it even more. And as a paying subscriber I atleast expect that I should be able to tell the AI what I want, and what I don't want, and enforce what I want, where I want it, without it having it feel like I am rolling the dice. And as I previously stated - Out of 10 or so songs that actually turned out somewhat decent after 2-3 gens, exactly accourding to my specifications in the prompts - only this one song absolutely refuses to work with me the way I want it too
I burned easily 2/3 of my credits, this month, on my first month Pro, with it throwing piano in everything, despite no piano prompts, or male vocals despite female vocal tags, or failing to use sitar despite clear tagging for it. Though, I swear this is the first time it added what sounds like an acceptance/awards speech to the song.
Anyone else ever get that? Just some random speech at the end, some words seeming intelligable, but mostly gibberish?
Sometimes, after a crapton of generations, I'll get what sounds like gibberish introducing the song to a live audience.
Yeah it had an intro like that then a long speech of the same at the end. Almost sounded like someone in a podcast interview
only if it tries to go beyond the boundaries of the lyric prompt. usually when extending
Trying to do a song using my persona on 4.5 it takes the same melody and just replaces the words.. really is wrong and never had that happen before. ANyone else have anything like that?
Just write your own. Lyrics do not have to rhyme. Try coming up with a short story, break it up in to verses and see what Suno can do, you may be pleasantly surprised 👍
Are you using the Exclude Styles box? Try this string: "Female Vocals, Female Voice, Female Voices" in the Exclude box. Use capitalization to give the prompt more importance.
Typically happens the first time I generate using a Persona. Make sure to generate something severely different in style from the song the Persona was created from at least once to "clear it's mind".
I hate writing, if I could write, I would probably have been a music artist by now lol
Try using the summary for a book or movie. Anything is better than relying on gpt for lyrics, especially if you want uniqueness.
Thanks for the advice on that.. I will give that a try..
Every AI has a pattern that comes out in lyrics, and I can 95% of the time spot them right away.
Does anyone have info on the new "Copyright Infringement" option in the Report section?
So is SUNO just not going to fix the Compression of 4.5, or just let it sit there as is - Because the garbled mess the gens make is really starting to piss me off
4.5 is still in BETA... they are still actively working on the model.. If you have feedback, use the feedback form so they can be aware of what ever issue you have.
was gonna say its broke i wont use my persona anymore and waste credits..
Still in Beta...
please dont talk down to me.. I get its in beta..
Just stating facts... there are a significant number of users that are complaining as if 4.5 was in a full release state.
If you have issues with a generation use the Report Bug option and then click thumbs down on the generation. This. generates a report for the developers and can help get your credits back.
I get such angelic, smooth vocals when it hallucinates while generating an instrumental on 4.5.
https://suno.com/s/8kb1SoWrHybVqnfb
https://suno.com/s/e0qv9aeVIxw2zlje
There are no lyrics, but I can recognize a few Greek words. The flows and notes are mostly from famous Greek songs, or at least, it knows how to make a perfect song.
This may be good for the developers of Suno for research purposes.
Hey everyone!
Hope all is well! Wanted to take a moment to spotlight the form in this post #📝┃user-research-feedback message Since there’s been so much time to test drive the new model, I’m sure you’ve bumped into some areas for improvement! If you want to, please feel free to share your insights directly with the team there! Best way to make it better!
Hi @daring lava! 💖
Hi Ginger! Thanks for helping people always! 🙂
Its cathartic lol
I'm curious, any way to see what parts that make up a whole song - when it' composed of severals parts due to extending?
Not after Get Whole Song. I don't think there's a way to see which parts has been combined... if it was possible, maybe only devs and not available and visible to us users.
Personally I think this is something there should be, as you may want to have several variations of a track, it's logical to try different iterations - which naturally will create different combinations of part. Thanks for answering.
If you click on the full combined song, all the parts below it will highlight
I too remember something like that but recently It's not showing that, not on full song anymore.
They used to as I remember it, but not any longer.
Is there any way that Suno can find my account from a song I generated?
I made multiple accounts to bypass the 50 credits a day, and now I can't find an account with a third of my music
Bro.. Why would you do that? xD
I was invested every way but financially at the time lol
Isn't this against TOS ??
anyways... Send an email to support@suno.com and ask for help, they might be able to.
tsk tsk
is there a way to duplicate my song? I want to keep the original but make changes to a copy
Hey, Does anyone know why I can’t “make a persona” with v4.5? It says I need to subscribe, but I literally just subscribed to the pro plan. Tried different browsers and logging in/out—same issue.
When will they learn Audio Quality is by far the most important thing. Any new prompt features or improved vocals or UI changes none of that matters if the one thing your product is supposed to do sounds like crap. You couldn't fix the treble boosting problem in v4 so you removed treble altogether in v4.5 making everything a muffled less than demo quality waste of time. (Even with a ton of post edit EQ'ing.) I get it's not as simple as that on the backend to just fix something but damn that definitely was not the solution. I'll still pay my monthly to help out to hopefully get it done but it's in an unusable state right now. It's like if you bought a new phone and all it did was turn on but nothing actually worked on it. (Don't give me the "it's in beta" excuse, that's what was said for the v4 problem all through beta and then the entire duration of v4 and it never got fixed.) Rant over.
To me the most important thing is the average song setup and then quality, if the songs are boring I don't care for how clean they sound and that's the issue I have with v4.5, the outputs bore me
on top of that the quality isn't good, yeah
So am I the only one that's noticed an incredible dip in the quality 4.5 is giving now? Like before it was fire, now it's like a slightly better but still gritty, scratchy 4.0..
Its been a bit difficult to edit my songs because the vocalist doesnt stay consistent in the replacement. Any workaround for this?
hello everyone, I have a problem with the steams of a song created with suno.
I want the instrumental of a song but with the get steams function I get a very dirty and low quality instrumental, can anyone of you help me if I provide you with the link of the song to extract the instrumental in high definition?
I'm wondering why I always get this "An error occurred. A system error occurred." Is it because I'm using a VPN?
Hi Dasvon, thanks for the response. I realised my mistake shortly after sending that (of course) thank you!
hello! with version 4.5 when I try to use the “extend” function it is impossible to continue it. It adds at most 5 seconds more. I have tried it with different songs. How can I get it to continue?
Was messing around with some stuff and discovered Suno will count non-copyrighted, public domain things such as poems as "copyrighted material" (i.e. The Raven by Edgar Allan Poe)--so does it just flag anything and everything as "copyrighted"?
have you tried extending using the five seconds
you have to own outright the rights exclusively so public domain fails a rights check, also many public domain performances are not themselves public domain
vpn can get your account flagged
many external tools offer stern separation
you're not the only person to not like the changes every model brings people who think other models were better
No, I loved this model. However, I've noticed it's declined.
have you tried writing better songs
No 
Takes notes Things that Jesus would do... insult people's song writing abilities.
thank you for not being hurt by that joke
each model has a personality and this new one likes to provide era and room noise and old problems with high end quality under extreme styles still exist
the upshot is the negative promoting works better if you can be thorough with what you dont like
but we all know some styles are harder
Okay, what kinda kickback are you getting to always sing the praises of the current model.
i spend a lot of credits to get what i want using extends so i'm always full of it
basically
Okay. So when you do extends. Do you piecemeal your songs together or do you generate on big thing and replace the part you didn't like?
i don't like replace as you all probably know it's very gambly
whether i start big or not depends on if there's anything tricky at the start
like a skit intro you don't want to try to do the whole thing off first
I have one song which I adore of v4.5 so the ability to create stuff is there, it just is obviously not tuned to appeal to me since my taste in music is very likely not aligning with the taste of the majority of users, which makes it difficult for me to get anything I like, which is logical, a pity but I'm not much bothered by it since I have a model I like
If v4.5 will be adjusted to appeal to me and works better with persona it's a win and if not I lost nothing, that's how I see it
I just didn't "get"
they'll be working on fine tuning that o understand they come out swinging at first and then fine tune to user complaints in reported songs and conversations here
yes, and it does not work. I only create instrumental songs. Finally what I am doing is creating them with version 4 and extending them with that.
so some pointed complaining is good for the product
if you can well articulate the damage
what do you mean it didn't work
it extended five seconds no?
It would need some sort of "experimental" slider to be more daring and less hm, appealing, most of the fun I have with Suno comes from the model surprising me and v4.5 never really surprises me, it might be better at following instructions but I like to leave the instructions vague so that it draws from many influences, but on v4.5 that causes a more narrow and less exploratory spectrum
yes, but with version 4 it extends 1.5 minutes, it is no the same
i never know why users can't have more configuration where configuration probably exists
this one can do up to 4 minutes but changes or extending from a downward cadence or silence requires putting in enough dead space for, metaphorically, the musicians to tap their feet and jump in
the new model is a bear at closing your song down to not extend too long, the trade off is extends can seem like endings
but if you extend through short extends eventually they open into long ones
even if the band sits there a second, which remove works well on
not what I see happening, but thank you.
how far out in silence did you get extends before giving up
and how many extend
i always forget this there is a caveat
you can't use get whole song
I have created hundreds of extensions in every possible way, for days. With version 4.5 I have spent about 1k credits trying to extend and many hours in a bad mood. I prefer to do it with version 4 which in 2-4 extensions already did what I wanted. It may work fine with other types of music, but with techno, version 4.5 is really bad.
if you are trying to build out extensions
you use it after you get passed the hard part
so what you're now saying is i'm right you didn't try to get through the small ones or you did but prefer not to
i would take only one second of it and then extend again from that and continue doing so until they stop being free
no get whole song until you see long extends
because they will build up their own hyper local context
and forget whatever is stopping them
ok it works! Thanks! and now, how can I paste them wth the main part?
use get whole song on one of the later ones
create->get whole song and it will zip up from where you last had the beginning
through all the steps
can u me say some tools that make this much better and professionally ?
also no problem re: this problem, it has been sensitive since the upgrade
Hi! I want to make a kind of album for an imaginary artist, for that I would like to have the same voice on several sounds (logical), but is it possible?
'consistent voice' and 'same artist' work ?
Thanks!
I think most people looking for a free version use Audacity, but Audacity plugins for stem separation are very similar to suno's
With persona but not on v4.5 probably, persona works kinda bad there
@candid gull the persona feature is the closest thing to what you want
but it's not what you'd call great or consistent for what you want
but it could be!
but it probably wont be
but it could be
but it wont
yeah but it also "persona" the beat
i just want to have the same voice
If you use V4 or v3.5 it's very good but a letdown on v4.5
eyes the redbull on the desk
ah :/ i really need the v4.5 for my 5234 character long prompt ahah
That doesn't happen as much on the other models
Then you probably have to wait until they fix it
chatgpt says that i can put "consistant voice" or something like that, but idk if i trust him
Or try to get beginnings on v3.5 or V4 and then "remaster" them with v4.5 and extend or so, or cover, there's also persona+cover I think
GPT has no idea how Suno works, it cannot know when it doesn't know so it hallucinates a response
yeah but he checked that on reddit
yeah i don't thnk im gonna trust him
its sad that we can't create a consistant voice 😦
When persona is more compatible with it it will work again probably
but persona copy the instrumental/beat also, i don't think it will fix it (i want different song but same voice)? or im wrong
does anyone know what it means when it says the uploaded audio matches an existing work of art?
It doesn't do this on the other models and mainly maintains the singer
it mean what it mean
yeah
if you want a more in depth description it's probably a spectrographic fingerprint from a third party l ibrary of known works
ive never uploaded the audio before though
oh ok, last time i tried persona it just copy the entire sound and made a little different instrumental
if it was made with standard samples it can sound like things already copyrighted
okok ty for ur answers
I've around 4k different Songs with the same singer on v3.5, I only create with this dude because I like his voice
But doesn't work on V4.5
yeah but v3.5 is not audible lol
I think they say the personas are not especially designed to work across models but I've heard you're supposed to try and reproduce the original in the new model and persona off that
don't want to make 1996 quality songs
I'm talking about how persona used to work on 3.5 and 4
And it just doesn't do it on 4.5
This message usually appears if you are trying to upload copyrighted audio (which is against ToS, as you have to have ownership of rights to anything you upload.)
V3.5 quality is often better though than v4.5
I used it to unmuffle some v4.5 songs...
not for me lmao
really far
and the prompt on v4.5 are so much better
can make very long prompt with the help of chatgpt and i see a big difference
Depends on the persona over there, when you have a clean persona the songs are also cleaner, was creating some stuff without persona and it was not good
thanks
oh,that is bad.What can I do to avoid this? I can't connect to the Suno website without a VPN.
that could be a sticky situation and makes me wonder if you've done it all along if it might not be the problem
but it's hard to say, is it a private vpn or a consumer one
I mean I assume either an institution, nation, or family don't want you using it
i have a new persona for 4.5 but it has a vice versa effect and mainly keeps the melody, not the singer, it gives the same song over and over when using a persona over there, it's basically not usable right now
I imagine late in the life of a model is when they become useful to exacting people
yeah v4 has become able to use my v3.5 personas less than a month before it left beta state, they work now over there
i just don't like how the model sounds much but they work 
4 was like ear candy hard to contain because it was always so bright and scraping at the edge of clipping, 4.5 sounds more like when you put on vinyl and the first consideration is, this was mastered extremely conservatively and will need its entire tone tuned from a flat eq into something decent
there is a lot of extraordinary detail tucked in, but it's coming out ideosyncratic and not in the human-ear-friendly way 4 was
v4 wasn't ear friendly to me, it was so sharp it always gave me hearing fatigue after a few minutes
v4.5 has audio problems on my system and sounds muffled on top in many genres
v3.5 has robotic vocal glitches
choose your flavor 
i run with v3.5, the robotic glitches can be edited away but i basically live in the edit section screen
Hi, 4.5 remaster feature constantly creating worst quality results and just eating up credits by that. Just sharing my case.
I personally think remaster should be called smooth
Hmm? 4.1o doesn't have that problem so far, I know, well... as long as you tell it to "research how to properly make Suno lyrics", it will search the web and keep it in memory in the current context.
What I used to do is have it generate raw vocals instead, then split stems and mix+master the final myself in Ableton.
how do you properly make "suno lyrics" though?
what are suno lyrics
It's not always smooth as it changes the seed a little bit, if I am correct.
We both know it uses Bing to search, so as long as your search term is human, you'll get through it.
that's not the point i want to make
I think for me as a person who wants more than what is apropriate it would be useful if features included their method of action as tooltips so it could be said what a remaster is
Also, it probably learned the patterns now, because we get this.
Also the persona in 4.5, is pushing too agressivly through, even if you turn on "Intrumental" the halucionated lyrics are still coming through :\
is it a pre 4.5 persona
You can see the sources where it found the info from, so, yes, it knows how to generate proper lyrics for Suno.
yes
that's a cool story my man
I think they updated the persona in a way we don't want it.
I heard it now uses the same seed as the original song, so, yes, it may hallucinate old lyrics.
I now use personas as a work-around to use a song but change its lyrics, just like a cover, but not really a cover.
yeah, at least there is still option to use 4 version, which is a great that it is still available
For languages where you combine English and a foreign language. Because the Edit section won't work for it, it just messes up the entire project.
did they remove the Remaster feature for songs made with older models? I can't seem to find it
I don't see the button either, but I'm not currently subscribed
If I remember correctly, it was under "Create"
Does the button appear if you select a newer model?
no, and I'm not subscribed either. If it still exists but only for subscribed then I think I'll subscribe, it was a good feature
It seems worth it to subscribe. I'm personally just waiting for a billing issue to resolve
hey, idk if its a bug, i have 2500 credits, when i try to generate a song it tells me i have no credits left. anyone knows why this happens?
Is there a way to change the style of a song, say Lo-Fi Chillhop for Verse, then say Raggae for chorus?
to change as fast as possible use extend and only accept single second portions from any extend you get while continuously prompting for the new desired style
eventually this will result in a hard change
to maybe get a change just put tags int he lyrics and both styles in the style input
and then everything in between is what's in between the two
S/O to people helping people here! Always a big deal!
small question, still get free credit per day after i paid subscribe?
Only when you have used all of your subscription credits
uuhh, well i take it better than nothing,
Hey guys is anyone else having issues when you go to extend a song in v4.5 it keeps giving me the 2 second clips? like a lot 😦
You'll want to make sure to begin the extension prior to where the song feels like it's on it's way out! Good rule of thumb when bumping into this is to push it back a little. Continue to do that till something grabs hold!
Happy to help!
omg it works!! yay
If it shows you have exactly 2500, it's possible that your account has just refreshed? As well as the payment being auto-paid from your credit/bank account, have you received an emailed from Suno that your payment was posted. Email would be from Stripe/Suno.
that was a good concise version
Are you saying there is a better method for getting vocals without having to stem separate?
Hi!
Does anyone know if the SUNO team has any plans to improve the Remaster system in upcoming versions? Or if they believe it will ever be possible to create an exact remaster of the exact original song, instead of the current result which feels more like a remix or cover? @flint haven
so ive told chatgpt to master my song and he actually did a better job than bandlab.com (text on image is in polish but chatgpt wrote what he did with the song and turned it into a mp3 from a wav file)
chatgpt generates mp3 files??
Can anyone explain to me how it works with the subscription do you get the credits after the day it's paid for or do you get it on the day you pay for it for the next payment for the credits
you get instantly after paid subscription, if you pay for montly you get 2.5 credit after your sub active
Soon Suno will just outsource all its remaster buttons to ChatGPT without telling you 🐦⬛
Yeah because it's the 17th and it hasn't went through do I have to wait for the next day cause I'm confused how it works
well if like that, you must ask support team through email.
when i paid, i instanly get my credit
I normally just wait till the month I have to pay next comes up but if it doesn't do it automatically i'll just do it myself I don't have any payment problems it's just I'm not really the waiting type and getting help from support is almost impossible for me unless it's an emergency so I'll just deal with it myself but thank you for the suggestion much appreciated
Hi, i need help i guess. "This device was detected as suspicious and has been blocked. It will no longer be able to perform actions. If you believe this was by mistake, please contact support." i get this when i try to log back in. yesterday was working fine, and nothing has been done (imho) to trigger it, or if i did something wrong, i have no idea. I already emailed suno, but is there something else i can do?
Yeah the payment went through so I didn't get my credit so I don't know what's going on is it a one-time thing cuz I don't understand how it works
the only way, you have to contact the support team, I think there is a mistake, but if you are patient enough, maybe wait for the next day, just make sure.
What mistake
Yeah even says refresh monthly on it too so it's not a one time thing as I thought
that error on the system i think, if you paid for first time, you must get your credit
i have two paid account, and get credit on the same day.
but if not first time, maybe this will explain.
Ok so either I wait 24 hours or complain to an email
it says the payment has failed, but im trying to cancel it anyway and i cant
Support takes literally forever to reply
yeah but why cant i cancel it
i can only change the payment method, not cancel
why is there an option to cancel the free plan but not the paid one 😭
I think that button means that by cancelling the plan, you revert to free
yeah that make sense
might have to wait for daily reset if you just did it
I kind of thought you kept the credits if you unsubbed though
That would have been the follow up question, if your payment was declined or failed. With it being in a limbo state , Stripe may have a protocol to reach back out and try a re-process the charge after a failed attempt, as many companies do.
The CANCEL button is just minimalized and stays in the BASIC box but it represents all levels.
but the cancel doesnt work, i still cant generate anything, and it still shows i have credits. and also before the payment renew date i had like 1000 credits, so is everything just gonna get removed?
If its now switched to "Upgrade to Pro" you should check your profile instead. Does it say "Renewal date: XXXX"
:"Cancellation date: xxxx" or something else. If you've successfully canceled, it may also say "Renew Subscription" , but of course, the payment didn't go through and so past the expire date. So a couple of things are going on.
Are you definitely sure you want to cancel the account? or are you anticipating funds to be available now in your bank/credit so it can attempt a recharg?
click your credits to see your present plan and status.
this is what i see, and i found nothing with the renewal date or cancellation date
this is in the monthly subscribtion tab
I made an assumption. So are you Basic and clicked to Upgrade to Pro this month? or were you on Pro and it auto-refreshed each month?
i was on pro for 1 month, and this month payment has failed, i forgot to cancel it before the renewal date
got it. So now it appears 1) its in limbo for the payment not going through. it gave you the 2500 , but locked you from being able to do anything. 2) It would not , from past history, drop you automatically to Basic. But it says you are a Basic account now. So I can only assume that you HAVE canceled by clicking the cancel button and its even possible tomorrow the Credits will drop to 50 and unlock your account.
I've got a feeling that locked accounts must be unlocked by Billing directly. Its hard to know with multiple actions of no payment received and clicking cancel together.
At least get an email started with billing@suno.com to try to work this out.
Keep aware, Stripe may still try to reach out and charge your account once funds are available as well as your Credit/Bank may charge overdraft fees. Billing may be the only avenue besides trying to contact a mod and DM you the issue with the bank.
Sounds like a very confusing, unbalanced system
Click your avatar, click Subscription in the drop down menu for subscription status.
Oh, sorry mate. It looks like you ARE definitely in Basic status. Did you click Cancel twice (more than once)?
Yeah, since I wasnt sure if it worked
yea. No worries. it may even be a good thing. We hope that will stop any further attempts on the bank account, but I dont know.
Seems like its gonna take a while to cancel
Yes. Youre right. I dont know what communication would be between Suno and Stripe. Lets at least DM @steady kelp as this is a bit convoluted and involves a possible charge process still possibly open in Stripe.
- Check your account in 4.5 hours (refresh time for Basic accounts) and see if it drops to 50 credits 2) Stay on top of your bank account for any recharges and you may wish to alert your bank and possible ask to waive any present bank fees. Hopefully it just failed and thats it, no fees. 3) Send your billing@suno.com email out now. 4) DM MikeB to see if it can be expedited a bit giving the off chance Stripe might try a re-charge.
/refresh
Refreshes at 0000 UTC+0 (<t:1712880000:t> local Daylight Saving Time)
Alright, I will follow these steps. Thank you so much for helping me with this!
Normally. if you cancel a paid account, you can use until the next renewal date, and then you'd drop down to Basic. But because of a payment not going through, along with clicking cancel after the renewal date, the Pro account status is both still stuck in limbo and says will cancel June 15th. technically, if no payment is ever made, you would be stuck in limbo until June 15 , then drop to Basic. We can see what happens in about 5 hours, but likely its now stuck/locked in limbo unless a payment is made to carry you through to June 15th.
So, yes, do all the previous steps and try to resolve this as soon as possible. Good luck, friend.
Yeah it doesn't make sense for the other outputs to be accessible by default
I actually didn't know this. I think I'll quit publishing then.
uh oh. Not trying to cause a stampede. Just getting thoughts. 🤔 maybe... yea.. I think I'll delete and post in suggestions.
I'd rather know, than continue in ignorance.
I understand. just dont want to create a mob reaction. lol. OK. deleted and put in Suggestions only. moving on.
Am I imagining things or does this actually work when doing an extension?
[repeat the riff from 00:13 through 00:26 in the original clip]
I have some free credits. Might give it a go
is the ReMi lyrics model not working for you or is it just me
it knows the word "repeat", but Ive not tried to that extent. interesting if it works. HOw many iterations were done to compare? It may be just that a riff often gets repeated anyway.
I've only tried it 2 times so far and I was curious if anyone else had tried this before I spend a lot more credits testing it
no idea, last time I used that it gave me unusably-bad lyrics (~6 months ago)
Unless I'm doing this incorrectly (or because I'm using 3.5), it doesn't seem very reliable. 4.5 might interpret it better
5 generations (10 outputs) and only the first one seemed to try to actually repeat it
I'm talking about 4.5 specifically
I think 3.5 in general is slightly less likely to obey the suggestions
Sorry, I'm stuck with 3.5 until I sub, so I can't try a 4.5 gen. someone else might though
Yeah, after another handfull of tries I think it was mostly a fluke
Is it me or does the generate video thing not work? I click it and the little thing comes up that says "Generating video", but nothing ever happens. Also, the download video thing still isn't working for me. I know I have to wait for it to generate, but it's been days now and still nothing.
Are you a free user?
Yes
V4.0 still miss pronounce (Thai Language), V3.5 Solid (Generating the same time) 🌟🙏🌟 THank you
Hey Jason! When you get a chance check out this message 🙂 it will give you a quick break down of what is and isn’t covered with the Basic plan #📚┃suno-school message
if you have any questions let me know!
You can try the phonetics method to the pronunciation right in your lyrics and see if that helps
Hi all! Looking for the specs for the video loop new feature to try it out. Anyone have this. Information? I can’t seem to find it in what’s new. Thank you!
The video loop is 10 seconds long #📢┃announcements message the image itself the ratio for that is 9:16
Your amazing! Thank you!
You’re welcome 🤟 If 9:16 doesn’t work well for you, you can try 1000x1000 the image should be fine within the 9:16 frame too
I created an account with my gmail. Loved it. Decided to get a paid plan. When I tried to reconnect, I had to signup again. I proceeded with discord this time. It worked fine but when I started looking for the first songs I made, I couldn't find them. Only to realize that I have two accounts with connected to my email and the other to my discord. Before paying another month on my discord linked account, I wanted to connect both accounts together, or swap the songs from the first account to the second. Otherwise, I'd like to cancel this month subscription which was billed automatically to me without my knowledge and I'll get the subscription on the other account. Easier to fuse both accounts together but I don't mind complicated. We'll just cancel this month subscription to start. How do you suggest we proceed ?
Does 4.5 have the ability for me to upload a rough guitar recording for it to base its generation on?
Hey there 👋 thanks for reaching out! Best next steps would be to reach out to support@suno.com. They’ll be able to give the best guidance on if this is possible!
Okay, so with the prompting in the style section. There's some examples on the pop-up that we get when you first load. The page of course. Begin with rap, melodic singing with the hook or whatever. The hook or whatever.
But to what extent does this actually work? Does it recognize the prompt labels on the actual lyrics? For example
[Verse 1]
....
[Chorus 1]
....
[Verse 2]
...
[Chorus 2]
...
Style: The song begins with heavy drums and a powerful delivery of verse 1, chorus 1 increases the tempo and the delivery remains powerful, verse 2 suddenly pulls back the tempo slows and the vocals become more whispered, chorus 2 builds from the show tempo to a now perform delivery at the last half.
Or is it more restricted to
Start is Y, verses are is B, chorus C is D. Kind of low literacy prompting and
The style has a context limit of 1,000 or at least the character input limit, I suspect it's probably actually a 1024 context limit which is because it's an llm encoder. Eg like UmT5 or something, since they allow 1,000, that is either arbitrary or because there's 24 key tokens preceding it that's actually used by the style process or it's just a straight encoder. Which is what I really think it is
So I wonder how literate it actually is
Hey Insomniac! Just did this a couple days ago and it did not disappoint. Here’s a cool article that touches on the upload audio function (just in case you hadn’t seen it yet) https://help.suno.com/en/articles/2477633
There's a new way to make music in Suno, and it starts with your sounds! Upload Audio allows you to turn any sound into music. Whether you've recorde...
lemme read this carefully
in my non general chat glasses
What I'm saying here is that I suspect that they are using a direct llm encoder. Eg not processing.
There is some nuance to answer you question that depends on us finding common ground with my knowing how to program a NN but not how to tell you how to install comfiui and have it effing work as good as automatic1111, who has never got a llm to run
generally you cannot make references, you can't make assumptions about second layers of interpretation and that has been a hard rule until lately
where it is true that some level of clausess or relationships can be resolved, or appear to be
I can't speak as to the specific algorithm being used, so before I go into any more detail about intuition
I'd need to know where your curiosities lay compared to where I'm coming from in answering you
Well I mean I've done a lot on the underside of llm encoders alongside with image generation. On the back end mainly involving attention, mechanisms and guidance focus. I guess my point here is is the interpretability attentive to the actual style request? Or is it simply focused on temporality and direct application. Which is how the prompts in the example look, but I suspect it can do much more than that
it is by no means procedural, it is not clicking through slices of time
if that is your question
but it's not not doing that
Eg. A weakly applied training would see. Timing-location-style to apply, etc , etc
if you want an engineer someone else would have to step in and talk from their standpoint
I can tell you experientually how it behaves from observations
Well no I know it's not procedural what I mean in this sense
Begin with X, shift to Y , End with Z. It's not per se procedural in the model side. But if the training is only at this depth the literacy isn't going to go beyond that
Hello! I upgraded to the pro yearly plan a few days ago. I don’t get the pop up asking if I want the credits now. I only get this message. I’ve sent an email to billing support several days ago and haven’t gotten a response yet. I would really like to get this sorted out please
literacy is not limited to understanding, or else we need a finer point on the word literacy
that sounds so nonsensical
Not when it comes to prompt adherence...
well you may be beyond me, you can tell something juggling a multivariate signal to jump through the next hoop but if that's getting filtered by a bandpass what have you accomplished
and would you train one that way
sorry
I am smoking
and that's nonsense
😂
@tawdry scarab clearly.... Lol
I guess basically what I was asking though is if I tell it to do something with verse 1 is it literate enough to recognize that verse 1 is The first one that is labeled as such in the header in the lyrics.
I mean I can test it. I just didn't want to dump a bunch of credits into something that has already been done
Billing support takes about a couple days to reply back during business week days maybe wait it out a while longer and see if they’ll reply to your message.
well can I answer your question somewhat disingenuously by posing a thought experiment
actually short version real fast
it's not going to let you program it like that smoothly
is the short not playing around version
the long version is you can make it do crazy amounts of stuff
and that usually includes a creative version of your original intention
and then the thought experiment is, if you want a song that is like songs that were written such that someone knew when they wrote it that the first verse in the header was also the first verse in the body of the work
and if that were lets say part of a movement, with instrumental directions
Then your input is going to look exactly like the body of work for which it was true in training that the desing was well put together
it was coherent and it will resemble and give rise to coherence
so short answer, you can't program it but you can give it referential meaning and it can surprise you and you can learn to think like a procedural programmer instead of an object oriented programmer
because that will let you use extends and shape music over time
I hope that hasn't been clear as mud
I've been waiting for two weeks
They will probably never answer
So what I can see here is that it tries to adhere somewhat, but it can't change its bass music. Well. The problem is that it's already stuck in a heavily trained area. That is a country. Song does not become a urban hip hop song obviously. But it can try to push it together. It can't however shift it at the chorus level because I'm guessing chorus delivery is not something that it has a scene often at all changing. Since that's rare in the real world. The words may but not delivery.
But it did listen more than I thought.
I can show you some more fundamental changes if you don't mind the length required to make them musical
but you have to play with it, it's fluid
Yeah. I can see that
you can get what you want but you end up with unicorn things you wish you could get
I was prompting up like it was 4.0 for the longest until I actually looked at the 4.5 upgrade little advertisement and realized wait a minute... But I'd already burnt all my credits up. So I just bought some more and figured I'd try it out this way
Yeah. That's ai everything for ya lol
I just finished a 20 minute mix that starts with some pretty people repellent metal I'm anxious to find anyone wanting to listen through XD
Lol toss it here
Listen and make your own on Suno.
Ok . Let's see
13 minutes it is not bad at all, though the 13:30~ transition felt like an end to the song. If that break to silence was not there or shorter. It could transition really well not feel like a new song
But even with it's not bad
oh you know what that's a good point leaving the gaps in is kinda like broccoli in the teeth that could've been removed 🤣
There's some gaps in there that aren't really that bad. Actually, the only one that got me was the one around midpoint 13 minutes in. That was really long. And it felt like the song ended changed tracks and began anew
the model does that i don't know why
but it's obvious where the album cut points would be
is how i think about it
Yeah it can be plucked out can't it even within Suno using the edit song stuff
right
that's a daw problem
oh ha ha
i misunderstood you
embarrassingly
suno can but my reaction came from feeling like it's my baby i just made it 🤣
Well I mean it works. It's not a complaint. I don't care if it sounds like a new track or not. I just mean it feels like it's two different songs after 13 minutes. Midpoint. Because the transition. Eg. If you snipped it there and created two different songs, I wouldn't notice if they played back to back
totally
people often say that the extends aren't working because they only give you a few seconds
sometimes it's building that little gap
If I were to actually do something with it outside of suno. I would probably draw the pad harmonic in the background across it.
well, anything you do with it in a studio is like using a toy camera in a photography studio just a little
but also
pretty awesome toy camera
Well okay. So with extend what I've noticed is your best bet is to actually chop a completed verse in half. Or in a last quarter of it, and use that as the beginning of your next. It works pretty flawlessly for me that
yeah i like to have it stitching some continuity
there are different circumstances where you might like different approaches
whether you want to ensure it blends and gets feathered in blending like it's right in the middle of something it can pick up on
Are you serious?
at least that's how it seems to me
Yeah. https://suno.com/song/f8b582a9-70a4-45bc-9d13-1b1972eaa054 this here for example. The song isn't great. But it changes a good bit throughout it, with very few transitional points that can be identified. If at all. Which is what I was trying for, the song itself is more experimental in the sense of the model, not experimental and trying to make it sound good.
Older models had issues with some of the positions I extended from. But 4.5 handles it like a champ. But other spots (not included) wrecked.
there seems to me to be some finesse to the tool
Yeah
some trade off between your management of past context in terms of the most recent things you came from, has something to do with it
No it's just examples. You don't need to listen to them as entire works lol
that's semi rude i guess i do like it but i was like gina be a minute
I really meant you skip through the psytrance one. I chose that one, particularly because you can hear the distinct transitions even as you skip through it. Because being electronic and repetitive, you won't miss the transitions so much
This one before is actually extended around 30 seconds and 50 seconds I think. I can't remember where I split it, to change little parts of it and the tempo at what not
It's short but used multiple extends to make it more chaotic
i just figured they were worth listening to
Gen on 4.5 / Instrumental Track is very good but there are noises, distortion and doesn't sound crystal clear 🔥🙏🔥
Hmm. Yeah 4.5 does some things. Clippity cricket sounds, crunchy from the redline. Maybe some samples weren't normalized in training. I dunno what's up with that.
But doesn't do it too often thankfully
Can someone kindly confirm that Generating Instrumental Track is good using 3.5 ??? Thank 🔥
Depends on genre...
I've said it before but I'll thank everyone here again for the work they've done to prompt more precisely and share that with the community.
I still find 4.5 infinitely frustrating but the work here helps a great deal. It just costs a lot more credits to do things now if you want to stray even a little from the blessed song structure.
I think it’s so crazy that it’s this hard to purchase more credits
Yes, it certainly does. I have spent 2500+ on learning how to wrangle 4.5 and I feel I have only scratched the surface.
I feel like that might be part of the plan. Monetize the most obsessive people who have the taste in music that the owners like the least
Dont get me wrong, I love Suno, and learning the new model has been actually enjoyable. I never got close to using even half my monthly credits before 4.5, and got to within 100 doing 4.5 stuff. As I learn how to get it to do what I want my costs will go down...
And I did write a song about burning Suno credits 😄
Just doing my due diligence but the extend song feature appears to be broken or not working. Using an android and I have tried clearing cache, cookies, and restarting phone. Nada
is it giving you short extends
Nope
No option at all
It's going into "create" mode
I click on extend and it goes straight to create
create is where you establish what will be in your song's extension
There's no clock to set the extension time
the time any generation or extension plays for is determined by the generation process, where if the song gets close to an ending, it will gravitate to ending
most of the time that ends the song before it requires an extension
Dude
yes'm
You obviously pick the spot on the track where you're extending from
It doesn't have to be the end
Now the timer is missing so you can't select
oh
i think bugs and then whatever you use
Yes did both just to be safe
ty
Oooo I’m about to fing lose it with Suno right now.
Why… can’t… this…. Fing ai just do a CHOIR ONLY SONG WITHOUT MAKING GIBBERISH FOR LINES??!??
OH… oh I’m now even more pissed. This discord is going to censor me?! SUNO, GET YOUR SHT TOGETHER
This auto moderator needs to go
Oh my god I’m going to lose it with this piece of CRAP
Get rid of this censor crap p
You only enrage me and others more with this censorship BS
“Woah watch the caps we hear you just fine” APPARENTLY NOT PAL, I wouldn’t have the same problem for 3 PLUS DAYS. But god forbid I get upset about it some stupid auto mod automaton is going to come in and censor me? And tell me not to cuss when I have spent hundreds on this program only to be given SLOP and GIBBERISH? I have a right to be upset, a right to be angry, when we are at 4.5 and I STILL can’t generate a classical orchestra with a choir. Ridiculous.
Changing display formats can literally make it disappear depending on your browser and format settings. At least try changing format to as small possible font and display and see if it shows up in the Create panel.... emmm.... on mobile. I dont know about desktops. Get that free test out of the way, at least.
It hurts more because I have such amazing songs in my repertoire but I can’t get a simple national anthem sung by a choir for a fictional nation. Drives me insane.
Did you say your prompting program seemed to help with influencing male or female vocals?
Oh.... follow up. Have you seen it improve on creating audiences or applause / or a smattering of applause? 4.0 wasnt that good, 3.5 was great from the beginning, but 4.5 is a goose egg.
I'm also getting a gibberish problem where it's clearly vomiting training material into my songs
good example here
Is it something technical if Suno debits my account but I don't get any credits...?
HOLY CHRIST I can’t get this moronic ai to stop putting solo vocalists in my chorus only song for the love of god someone tell me how the hell I can get this idiot to stop messing me prompt that is perfectly provided to them
I need someone a developer or something to step in here and help me fix this URGENTLY because I am not spending a god damn week on this
I am steps away from actually breaking something and I am not kidding someone help me please and urgently
@polar spade no specific luck on either side those are tricky to me still
Hi ! Do you guys also have issues with personas ? When I generate a persona (based on 4.5), if I reuse it there is way to much similarity, exact same track start, same instru and melody :c
@rare oak To avoid that I usually chop the music I want to use my persona on, for 30s-1 min and I use this chop as a persona
Why not have the normal video thing for free users and the video generation thing for the premium ones? I feel like that's a good idea.
Okay
You still manage to get the same voice with a sample of this lenght ?
Not always the voice is identical, but its similar
ookay, thanks a lot mate
my suno keeps reusing the song i made from before even when it has a new song made ,keeps remaking the old one how do i clear its cache
It sounds like you are using a Persona. That sometimes happens with them
nm my prompt had the word metallica in it
Emotional metal ballad in the style of Metallica.
this was the prompt
Extending songs is not working. It's not even going to the right menu when I select "extend" and I've tried everything. Extending is integral to using Suno so can someone please look into this?
Does anyone have some pointers on how to edit the lyrics in a way that will work "better" with video downloaded from Suno. I see in mine it line-breaks in the middle of a word and leaves punctuations on a different line. and the section labels (eg, [chorus]) goes on the same line as the lyric when in the original text they were on separate lines. any suggestions welcomed. thanks.
What does the next screen look like if you were to select Extend?
That looks like a Cover screen instead of Extend
I've cleared cache/cookies, logged in/out, and restarted my phone
See on the top of the first screenshot
Yeah exactly right
Is this on mobile browser?
Android
I just checked mine and it was working fine, I use Android too
browser
I'm in a beta for some features so it could be related but unsure
one thing I noticed
that your options look different than mine
Extend from... and the time...
your browser is Chrome right?
Yea, the Extend option on yours is under Edit/Remix instead of Create
That looks like a different version to me. Strange.
Thanks for confirming all this!
that might be it. you are accessing a beta test version and it might not working as expected
I'm running a released version since I'm not part of any testing group.
Are you even allowed discussing in public? 🙂
no
Discussing what?
A bug?
There was no way for me to know unless I asked in this group. I'm a bit tired and bored of the snarky attitudes
Don't they have a closed testing discussion among beta testers?
if you test features that ar not public. you cant discuss them in public.
yes we have that and its under nda
Extending songs is not new
Try reading the full scope of the thread before interjecting your pointless comments
dude
Is this group even have a moderator?
yes. me
Nothing secret was discussed in this thread bro
It was a tech/bug question on a pretty important Suno feature
Extending tracks
yes. if it looks different from what regular users see, as it was discussed. you probably have features that you havent. and to minimize damage i usually deleta ll of that because some users dont respect the nda... wouldnt be the first time
It should have nothing to do with any beta testing because that would mean it's completely broken
We literally didn't find that out until the end of the thread again. I encourage you to read the whole thing.
@muted summit does this user above have assess to alpha stuff? if yes we should reconcider that descision.
I'd like to report this moderator
For harassment and being entirely unprofessional
@muted summit
I have extended my Suno bill, but why hasn't my credit increased?
dear sir... you cant blame me for your oopsie.
@wet minnow I've reported you to Discord for inappropriate behavior and harassment. Suno must be proud
Bro, calm down... no one is being unprofessional here.
Mr. Tree simply said you shouldn't be discussing any new feature, bug or not, that isn't released to public... as it was part of the NDA that you agreed on when joining the beta testing.
You can discuss but within the testing group only.
Are you off?
As you clearly understand you have broke the rules big time and now try to blame it on me to distract... they will ignore that.
cool
I'm a paying premium customer of Suno since the beginning. To be treated this way is PATHETIC
I also run a company and would never employ a moderator like you
I'm sure you can leave the beta group if you want to.
You've also been reported
and use the version everyone is using
you just make it worse... please stop.
@half dragon @wet minnow you have
I've pinged you in the appropriate channel.
Please take the discussion regarding the Suno problem you are having there.
We are trying to help you not break an NDA.
Anyone can read this thread. What feature was discussed that violates an NDA?
You've put yourself in this legal soup. So now swim out of it.
yes. pease dont discuss the stuff we cant speak about here.. thats all we ant.
we cant tell you which feature you leaked... because we would do that too.. and we respect the nda
Guys I have questions,I have extended my Suno bill, but why hasn't my credit increased?
ok cool... if you are ready with being mad at yourself ... calm down and discuss that internally with us.
might take a little time to show up. Have you tried logging out and refreshing?
My lawyer will. No NDA was broken but you've now slandered and harassed me
I think you should just chill dude. You are overreacting, quite severely.
I have tried refreshing and logging out, but the result is the same.
Have you read the thread? Or just kissing up to a Suno mod?
its the strategy to make other things look worse to distract. thats normal behaviour for some types of users... we expect that to some extend... i am not mad or so.
you on mobile or web browser?
When a paying customer has to block a moderator, the company has failed
´see
Let's not comment on it further... just let him vent and cool down eventually.
Mobile and web browsers are the same, I've tried it
I've not been the hostile one
we gave him all the cances to go to the right place, gave all the persons to contact... user is still aggro... timeout will follow soon if it gets worse.
You are now.
How?
have you checked to make sure your payment posted?
maybe post in #🪲┃bugs-web too
may i suggest to a: not get in the aggro spiral and b: discuss that in the right place and not here... please...
Hello, does someone know how to make songs that sound like Alan Walker one's like Fade, Spectre, Alone... i really like the style but i tried someting like this: electro house, edm, main catchy gliding melody. strings, plucks, synths, pads leads high pitched, stereo effect, light and aired, lite verse and big drops. uptempo, mainstream. cool but its kinda bad
Hello, I'm Marco from France. Can you tell me how to deal with a voice that's too loud? Is there a way to balance it, like an equalizer on Suno?
Hey HyoDash! Thanks for reaching out! I personally haven’t tested out things in that style but maybe someone else has and will chime in. What I can say though is that it can help to start simple and then build out as needed. That will help you better identify what is working and what isn't. Very interested to hear more about your exploration and testing into this though!!
I have found out that sometimes - Less - is more.
Hey Bravo! Thanks for reaching out! There are a couple of ways to approach this! The restyle function can help a ton when trying to retain the original music/meldoy/etc while adjusting the levels of the song. Here's an article that will help explain: https://help.suno.com/en/articles/3591617
Additionally I've heard a lot about people using persona and cover to help repurpose a song they previously made. Keeping the details in the Style of Music would help keep it from steering away from the original! If you need any links to those things please chck out the #1294418290348921007 channel! Lots of great resources there!
Also, if you have any other questions, please don't hesitate to ask!
A new song/cover usually has louder centerfold vocals since that is standard in today’s music. I found that by remastering a track, it evens out the sounds more albeit with some distorsion. Depending on genre this could be used for artistic effect. Otherwise I recommend changing some values in an audio editor.
How do I make an audio editor my friend?
Sorry, not sure I follow. Would you mind explaining more about what you need help with?
I think he's just asking how to make/use the audio editor that was mentioned
forgot a very critical comma before "my friend"
If you want to use an external tool, you can try splitting the vocals from the rest of the track and just lowering it. If you want to keep it within Suno, you should follow David's message, maybe also search around for specific audio-level tags to add to the prompt if there are any
I had this issue after I started upvoting songs and making new ones with a similar prompt/structure. The intro would feature the "Ah-ah-ah"(?) from the generations I upvoted, even if I didn't prompt it. Not sure if the upvoting was related, but I removed the upvotes and slightly modified the style/prompt, which fixed it.
- The upvote/downvote system is used when training a model. It does not affect your day-to-day usage.
- Indeed, I have heard from people who had similar issues with Persona that inserting a new line at the start (sometimes having a sacrificial verse) and/or changing the style prompt helps decrease the chances of the old pattern/lyrics reappearing
- If it seems that the peraona always gives you your old song, despite doing the aforementioned steps, you may want to consider making a new Persona (I would highly advise cropping the original source to like a verse or two so it decreases the odds of reappearing structures)
I don't have access to that feature, but that might be what caused his issue
Hmm? You referring to the crop tool? It is part of the Editor which you should be able to access as a subscriber.
What I suspect is that if you use the same style and/or lyrics as the source of the Persona, the AI would accidentally pull up/go into the learning pattern it was using from its source, which is likely why identical parts from its source reappears.
When the new song has a modified style (e.g. new keywords in the middle or removing/changing a part), and/or having a new line at the top (e.g. [Short instrumental intro]), it should have enough for the pattern to make something new.
The up/down-vote has no impact on the outcome
I was referring to Personas, which are also a premium feature
If it wasn't upvote-related, it was coincidental
Ah, my bad. I typically see such comments regarding persona because they just love to repeat what you had from its source
I’ve noticed that one of my song’s play counts have stayed the same for over a month, despite listening to it multiple times a week.
The play counts are unique users, you can't flip the count by listening to it over and over.
But I have private songs that have almost 100 plays that only have listened to.
It doesnt matter, you only count as a single play no matter how many times you listen to it.
The one I am speaking of though is one that I made public, and it seems like it stopped counting my plays once I made it public
It stops counting your plays at some point, likely to avoid this exact gimmick
I have a lot of tracks that are private, and some that are public, none have counted my plays more than once.
And Dotpuppy has it right, its to keep people from gaming the system
I have private songs that have over 90 plays counted
but that just leads to a new game on the system: play your stuff over and over Private, then Public it for a big play count boost to start...
Hello, I keep trying to upgrade from the free plan but whenever I put in the correct information for my card it keeps saying declined i've tried multiple cards (They have money on it) and it keeps saying it was declined. What should I do?
Ok, thank you.
Please be more specific
The lyrics that I used myself that are my own, can I register them after I used them on Suno?
??
I create the lyrics, they are my own authorship, however I did not register some of them, if I use them on Suno, are they still mine or not? I heard that Suno uses the data to train the algorithm (I understand that this serves to improve the algorithm itself), but what happens to my lyrics? If they use them, do I lose the rights even after using them on Suno? My question is, do I still own the lyrics? That is my question.
Most of my songs (I mean the lyrics) were written by myself, in fact I practically don't like generating the lyrics using Suno's chatgpt, in fact I just like to get the melodies so I have a base to create my composition.
I understand that I only have rights for commercial purposes when I subscribe to the paid plan (in this case, the melodies). Now, my question is regarding the lyrics that were created by me.
it depends on the country you are in if you have natural copyrights for them or if you have to register them. i think the best idea is to have a look at your local copyright laws.
is the contest up? and did they pick winners?
^
Just for future reference, if someone is looking back to the past for this info, the link to the #soundseasons volume 1 winners is back on 2025-05-15 #🚀┃updates message
thanks
in this case I want to register the lyrics of the song
Please look into or contact your federal Intellectual Property Office (e.g. US Copyright Office). They likely have an association or branch dedicated to copyright of the lyrics itself.
Typically, when you submit an application for copyright registration/protection, they ask you about each component (lyrics, composition, etc.)
Alternatively, you can ask a lawyer that works with intellectual properties, but that route will likely be much more expensive than contacting your IPO
Suno You finally put out of the play button! I couldn’t be happier thank you so much. All I have to do is stop using the Safari browser on my iPad and open it up and chrome and Walla. It works.
I will now stop my complaining about this
I need support with a billing issue .
Email billing@suno.com
i did , nothing but the confirmation email its been 4 days
they said come here for faster help
have you checked out #1294421750397796435
yes.
guys all of this is on the website which is common sense , if you do not work for suno disregard my comment because I have no desire to speak to a random person.
Then you will be basically talking to no one here. This is primarily a community-ran server, there is no official support team present on Discord.
Typically, billing responds in 2-3 business days (during Eastern Time hours). They do not work on the weekends as far as I know.
So if you sent an e-mail on Thursday after ~5:30 PM EDT, you will likely see a response by Tuesday since that would have only been one business working day thus far
cool thanks
4:05 and 4:07 wow a whopping 2 seconds longer. 🤬
What exactly is the problem/concern?
why only 2 second between the two? that twice now it happened. I was thinking no less :10 or :20. But :02.
Between two fresh generations, or two extensions?
fresah gens
They can end up being the same length sometimes, especially if your prompt doesn't allow much leeway for the AI
I'm a noob so how much leeway might be needed?
if you extend from the small extensions several times without getting whole song it will allow you to break into some new sustained music but depending on how it ended it may resist starting something new
but the more times you extend on extend only taking a second or two the less context it will remember
I'm not sure because I've never particularly cared about the difference in length between my generations specifically, but I would imagine just being less specific in your prompt would give the AI more leeway to produce different outputs
so it will start with a new style
okay Thanks
Is it down right now?
Is there a reliable way to have different voices for verse?
I tried:
[chorus - voice: ghostly children's choir]
THE TOMATOES BLUSH WITH BLOOD SO RED,
THE ONIONS MAKE YOU WISH YOU'RE DEAD.
THE SPINACH STRENGTH IS NOT YOUR OWN,
IT'S JUST THEIR WAY TO CLAIM YOUR BONES!
[verse 4 - voice: trembling survivor]
The garlic bulbs will glow at night,
A sickly green unnatural light.
They hum in tones that break your mind,
And leave your sanity behind.
The turnips grin below the soil,
Awaiting next year's human spoil.
[chorus - voice: demonic auctioneer]
SOLD TO THE BEET WITH PULSING VEINS!
SOLD TO THE SQUASH THAT KNOWS YOUR NAME!
SOLD TO THE PARSNIP'S IVORY TOOTH!
SOLD TO THE VEGETABLE TRUTH!
[bridge - voice: eldritch god]
WE WERE HERE BEFORE YOUR KIND,
WE'LL BE HERE WHEN YOU'RE UNDONE.
YOUR FLESH IS BUT A PASSING FEAST,
WE GROW IN DARKNESS, NEVER SEEN.
This didnt work and resulted in the same voice for all verses.
Suno 4.5 if that makes a difference
I normally do a seperate tag instead of including it in the [verse] tag
Like [Verse] [Whispering]
Doesnt work 100% of the times, but v4.5 seems to pick up on these tags more reliably than earlier models
ah, will give it a try
thank you
alas it didnt work for this one (judging on the first 3 verses of each version) doh.... an interesting tatic though
hmmmm a regen and it is doing somewhat different voices..... I love suno but there is often a huge dollop of random lol
Hey for some reason I have no trash bin icon on library page on two different phones
To double check, they both were in the library correct? Because I know that the Create page (which shows your workspace) does not display it but the Library page does to the right of the Upload Audio button.
(For future ref: if you need to access the trash but cannot access the button, visit https://suno.com/me/trash)
I heard it was recommended that we use personas made from 4.5 songs for 4.5 songs if we wanted the full benefits of the improvements. Would remastering a v4 song first count for that? And then remaking my old personas that way.
I suppose your new persona is technically from a 4.5 song even if the 4.5 song came from a v4 persona. I just did that this weekend and I think the new Persona seems good so far.
The original v4 Persona still had a very good voice and was still usable but it had shimmer problems and degredation problems. I think the 4.5 who came from her has less of that issue.
Has anyone found a reliable way to ensure that vocal stylings remain consistent through a song? As opposed to a mix of (metal, for e.g.) harsh vocals and clean vocals that seems so common?
What I tend to do is setting up certain details and parameters at the top of the custom prompt.
In short, I use section tags with inline descriptive or parameter tags with comma separation.
Using this format:
Tone: clarity, texture, pitch range, resonance, voice type;
Technique: breath control, articulation, consonant intensity, vowel shaping, resonance focus;
Emotion: emotional tone, psychological impact, thematic intent;
Projection: volume, support, intensity, power for climactic moments, dynamic range]```
I've done something like this:
```[Vocal Details: Powerful Male Baritone (PMB) -
Tone: clear, slightly raspy, midrange chest voice;
Technique: controlled breath management, powerful consonant strikes, open vowel peaks, intense resonance;
Emotion: supportive, determined, fiercely protective, empathetic, cathartic release;
Projection: full chest, deep breath support, intense power for climactic moments, sustained notes for emphasis]```
Then to call it and modify it do something like:
```[Verse - PMB | clear, slightly raspy, midrange chest voice; deliberate, focused phrasing; moderate dynamics, restrained power, controlled breath management; supportive, determined, protective, empathetic emotion]```
Interesting, thanks for laying that out. Will definitely give it a try. Have you found success/reliability with this? I'm skeptical that the level of detail you've provided gets overlooked by the generation is all - unless it's enough to steer the generations into the direction you'd like
I've gotten fairly consistent results, from my experience, it seems to be better to use this method as a setup up for base songs to create covers from, focusing on musical content that uses more simplistic section tags, then create covers from that where the format used, lyrics and vocal parameters are introduced, and musical content that needs to be reinforced could also be done at this stage. It works really well with cover songs as long as I keep using the same formatting, syntax and context - because straying from the formatting and syntax produces some odd pronunciations.
Would the vocal details be a good starting point to build a Persona from also? Or is it better to just include the data at the top of the song and call it in throughout the prompt?
Can't say for sure, but for the method I use, which I described above, it seems you should create the persona based around the cover once you've nailed the vocal sound you're after. The way I look at it, you always make the song structure and music first then add lyrics, then tweak, refine and polish.
Though depending on work flowers it's possible to do several things at once, the main issue is character constraints.
But if you can get the musical intent and genre from the style prompt, then definitely you could go for a vocal setup for your base.
Personally, I feel it would be a really good idea for Suno to actually add an option to build a voice persona based on hardcoded but flexible parameters, rather than one build around a specific song.
Is this the right place to ask for advice?
I've been finding v4.5 has a tendency to start at a decent volume then constantly get louder until it has pretty bad clipping, to the point it clicks. I don't think I'm doing anything obviously encouraging of this, and basic directions like 'quiet' don't seem to tamp down much.
Has anyone else experienced this? It's causing me problems more often than not.
This seems to be new in 4.5 but I've not found anyone else talking about it online.
I would say, it's definitely the right place to ask for advice.
I haven't experienced this as specific and recurring problem with v4.5, this may have something to do with styles / genres you're aiming for.
Also, it's possible that you may have tags that are not enclosed, you may be using section tags that could do this - like build-up or something similar.
im getting no where near 8 minute songs after experimenting. any advice? I have tried a few things like telling it what i want it to do after how many bars and also time per bpm and key. its getting the key and bpm right but still only goes 3-4 mins.
im absolutely facinated with 4.5 it sounds so much better
when i extend a song too i only get a max of 10 seconds no matter where i place the extention
Hmm, I don't think I've done that specifically with tags, but the last one doesn't have any tags. Maybe it's worth a regen with a traditional tag structure and see if that gives them a bit more restraint.
Thanks for the advice
You're welcome. 🙂
@ripe bobcat
How does you're prompt look like?
Generate a hard-hitting future house track at 127 BPM in A Minor, 7:00 length, with a grungy, futuristic edge. Start with a 1m intro of distorted FM bass and gritty kicks, adding a clap at 0:30. Include warped A Minor synth pulses for tension. At 1:30, launch a 2m drop with a gnarly, detuned lead and shuffled hi-hats. Add vocal chops (“break”) every 8 bars, processed with distortion. At 3:30, hit a 1m breakdown with glitchy, overdriven synths and a dark pad. Drop at 4:30 for a 2m section with heavy percussion (shakers, toms) and a harsher lead. End with a 30s outro fading the bass. Ensure a raw, aggressive mix. If Suno supports it, amplify glitch effects and boost bass distortion, delivering a club-ready track with a sci-fi, industrial grit that slams the dancefloor.
i mean its making exactily what i ask it but less than 4 minutes
and only extends up to 10 seconds
You could try segmented generation:
- 0:00-0:30: Distorted FM bass and gritty kicks, building intensity.
- 0:30-1:30: Add warped A Minor synth pulses, rising tension with claps.
- 1:30-2:30: Gnarly, detuned lead with shuffled hi-hats. Vocal chops (“break”) every 8 bars, processed with distortion.
- 2:30-3:30: Breakdown with glitchy, overdriven synths and dark pads.```
Then Extend the track from 3:30 to 7:00, continuing at 127 BPM in A Minor.
3:30-4:30 (Build-Up): Layer in ambient effects, intense risers, and subtle percussion to build anticipation.
- 4:30-6:30 (Second Drop): Heavy percussion (shakers, toms) and harsher leads, maintaining a driving, aggressive feel.
- 6:30-7:00 (Outro): Strip down the elements, fading bass and distorted effects for a sci-fi, industrial finish.```
Also, try your original prompt with v4.0 and if it works well, then remaster to v4.5
10-4 will give that a shot
every song has been unreal since the update though, just sayin
bravo to the team
up to 20 seconds now on the extention
so only about 4:30 max time
Was that with 4.5 or 4.0?
From my own experience, I feel 3.5 has better musical quality, but 4.5 has better vocal quality, 4.0 is between.
Remaster to 4.5 from 4.0 from my experience is not that great, it's better to create in 4.5 if that is needed.
the dynamics for electronic music at least has changed dramatically, you could make ok songs before, now you can actually pull off some fire stuff
with some remastering in FL studio they turn out insanely good sounding
Making some "fire stuff" with FL Remastering - You could do that even with a simple midi arrangements and using a decent VST, though that is manual.
Not sure why the AI won't make it to the specifications.
Many of my tracks made with 4.5 isn't that long either, think the longest in one go is like 6.30 minutes.
But have high level of control with my tracks, I don't do simple description prompts.
i am using your segmented generation, it seems to read it a lot better
too used to talking to MJ and other LLMS
I have prompts that work way better, the prompt I gave you was a quick and dirty prompt based on your prompt.
I used the same method we used back in the day when separating a filtered vocal from the original song and then mastering it, not so hard.
Try this: Just fill in the Mood and Key.
[Mood: ?]
[Key: ?]
[Tempo: 127 BPM]
[Instrumentation Setup:
Drums: Distorted FM Kicks, Shuffled Hi-Hats, Gritty Claps, Heavy Percussion, Shakers, Toms;
Bass: Distorted FM Bass, Pulsing Bass Arps, Driving Low End;
Synths: Warped Synth Pulses, Detuned Leads, Glitchy Pads, Overdriven Sequences, Dark Atmospheric Pads;
Vocals: Processed Vocal Chops ("Break") every 8 Bars, Heavy Distortion]
[Production & FX: Wide Stereo, Gated Reverb, Tape Delay, Deep Saturation, Bass Distortion, Glitch Effects, Dynamic Shifts, Industrial Textures]
[Song Structure:
0:00-0:30 (Intro): Distorted FM bass and gritty kicks, building intensity;
0:30-1:30 (First Section): Warped A Minor synth pulses, rising tension with claps;
1:30-2:30 (First Drop): Gnarly, detuned lead with shuffled hi-hats and distorted vocal chops every 8 bars;
2:30-3:30 (Breakdown): Glitchy, overdriven synths and dark pads, setting up the second drop;
3:30-4:30 (Build-Up): Layer in ambient effects, intense risers, and subtle percussion to build anticipation;
4:30-6:30 (Second Drop): Heavy percussion (shakers, toms) and harsher leads, maintaining a driving, aggressive feel;
6:30-7:00 (Outro): Strip down the elements, fading bass and distorted effects for a sci-fi, industrial finish]
*"[Descriptive Prompt: Generate a hard-hitting, high-energy future house track at 127 BPM in A Minor with a grungy, futuristic edge. Use distorted FM bass, gritty kicks, warped synth pulses, and detuned leads to create an aggressive, club-ready vibe. Include glitch effects, tape delay, and deep saturation for a raw, industrial sound.]"*```
youre an absolute hero among mankind
it is over the 1000 character max though i will edit it and see what i can come up with its great so far it hit the mood just right
That isn't for the style prompt, but custom prompt.
oh i see now
You can use style prompt to reinforce or add elements as you see fit.
so i turn off insturmental then?
Yes.
You should be able to extend it.
Basically where it ends, gauge where it is with regards to the song structure, then with regards to song structure in the prompt remove the parts that it made, and hit extend.
Curious (and forever grateful) to see if anyone can assist me with finding a prompt (or confluence of style/lyric elements) for generating a songs with this same kind of "feel". The feel being a consistent, punchy, and groovy song throughout that doesn't introduce disruptive versus (style change, inconsistent with intro, etc) that seems to be so prevalent in 4.5. I, since 4.5, can't for the life of me seem to find the right formula (be it genre or otherwise) for this type of metal: https://suno.com/s/xTmjr09z6GUGBiJf
I've tried to build out prompts, starting with the 4.0 "tag, tag, tag" approach that gave me this in the first place, all the way to more detailed, expressive, and specific long-tail prompts with the style description boost and more detailed lyric sections.
A lot of the tracks on the album/playlist it belongs to are similar in style to The Fly, and it seems like that kind of song style is now impossible to generate
I'm not sure, but I've seen others in the style field use both suno usernames and track names, which may also influence.
hi guys, sorry, i just wanna ask, I chose the wrong package, I intended to only pay for 1 month but I chose the wrong package and paid for 1 year, what can i do to refund ? anyone knows?
You need to contact support.
ok, im already contact, but the answer like a bot i think, he's dont gimme a solution 🙂
Hey Abdu Fathan 👋 Thanks for reaching out! You’ll need to reach out to billing@suno.com to get this resolved. If you’ve already reached out, hold back from sending additional emails and give them time to respond to prevent delay. Going to ping some guidance on what to include just in case you haven’t already reached out! Let me know if you have any questions!
/billing
For billing/payment issues, email: billing@suno.com
Include:
- Your Suno @handle
- Description of the issue
- Transaction ID(s) or payment receipt(s)
Before mailing:
- Check if your card was charged
- Verify funds were available
- If Stripe rejected your card, try using the iOS or Android app to pay via Apple/Google Pay.
Do not send non-financial issues to this address — it may delay your response.
ok thanks
i hope they can solve this issues
in this case, i just waiting for response? it's true?
Welcome to the Suno community 💬🙂👋
Correct, it is an auto-reply message that acknowledges that they received your message and providing links to resources to those that were not aware of it
Billing should reply within 3 business days (Eastern USA business hours)
anyone else all of a sudden getting 7 minute plus songs from Suno on 4.5? one of them is like...an acid trip I wasn't looking for what I got, but i can't lie. I kinda love it
Do you mean like...
https://suno.com/s/dENiEoCUqOWX94ek
Share it. I would like to listen.
Check it out...its ...different https://suno.com/s/8EEeJN6XvM6rlM1p
Like...I do really like this song but it is soooo off what I was looking for #magicofai LOL
I don't see anything that would cause it to extend so long. I'm just halfway through it, but it's a little harder than I have listened to in decades.
and yet its not bad right, like..it has something about it that is cool
I work on trying to extend my tracks to get a live concert feel.
It's not bad, but it's music I listened to as a teen.
Back in the 80s
lol I am far from a teen but heavy music still vibes with me. then again, I also write Drum and Bass professionally and the Cocteau Twins are my all time fav band ...i'm kinda all over the place.
I tried this song to generate several times. But I don't know why SUNO is not generating. I mailed them but didn't provide anything.
[Verse 1]
Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay - - --
[Chorus]
Bhaarath Maatha ki jay, Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay, Bhaarath Maatha ki jay - - - -
Veer Sainiko ki jay, Veer sainiko ki jay - - - -
Veer sainiko ki jay, Veer sainiko ki jay - - -
[Musical interlude, 4/4 beat]
[Verse 2]
Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay - - -
Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay - - - -
[Chorus]
Repeat Chorus
[Musical interlude, Balkon inspired beats]
[Verse 3]
Veer Sainiko ki jay, Veer sainiko ki jay - - - -
Veer sainiko ki jay, Veer sainiko ki jay - - - -
Veer sainiko ki jay, Veer sainiko ki jay - - - -
Jay, jay, jai jai veer sainiko ki jay - - - -
[Chorus]
Repeat chorus
[Musical interlude, Street Style grime beats]
[Verse 4]
Bhaarath Maatha ki jay, Bhaarath Maatha ki jay - - - -
Bhaarath Maatha ki jay, Bhaarath Maatha ki jay - - - -
Veer sainiko ki jay, Veer sainiko ki jay - - - -
Veer sainiko ki jay, Veer sainiko ki jay - - - -
[Chorus]
Repeat chorus
[Musical interlude, rapid beat cycles]
[End]
I tried to post my Spotify link to my DnB but the Suno police stopped me
Its Geezah, the art is a guy in a suit holding a gun. Folding Space is one of my albums. check it out if you want. I kinda like what i'm doing. I'm expanding on what I started in 2000 with Pimp X
Try posting a link to a webpage that links to Spotify and anywhere else your music is.
maybe this will work https://geezah.bandcamp.com/
I like Chaos Rain so far
YAAAY thank you. I'm all over the place with what I do. Suno helped finish a good amount of songs I had just sitting on a hard drive and added vocals to songs. It helped me want to write music again. then I branched out and did death metal with Goliath and Kinetkill and then Idol Pop with Luna Flare but Luna is gonna get a bit darker now.
When I post my newest song link in showcase it doesn't play and says it is zero length, is this a new bug of some kind?
I already did unpublish and publish again, but that doesn't help!
Check the song: How deep is your love #🐣┃suno-showcase message
It plays fine to me.
It plays fine for me too.
It does play for me too, but only in my library
It plays fine for me three.
Restarted the browser, but still the same, maybe I need to restart my PC!
same plays fine
@rose quiver I wrote this song for my girl. I missed her when she went to europe for 30 days. https://geezah.bandcamp.com/album/miles-to-you I like the 80's throwback version the best.
why is the liked button disabled for searching?
Right. You can't filter search results in your Library.
Maybe in the future, you can. Just not right now.
Right.. once you typed in the search box.. you can't choose a filter.
The search and filters are bad across the whole page, I guess they will address it in some huge overhaul
Restarted my PC, but still (0:00 / 0:00), all other songs work fine btw! What can I do? See screenshot below:
Maybe create a full crop so that you have a copy
I use free plan!
It plays fine on showcase room... but yeah, here it is unplayable.
Though, I don't see the link... is it public? private? maybe the video got issue so doesn't play correctly when embedded? Maybe regen the video?
This is a screenshot
ah
I published it, like I usual do
It maybe just an issue on your local browser... have you tried in Incognito mode? or maybe a different browser?
Well other songs work just fine, only this newest song has this issue
Not sure but it plays fine to me and others. So the song is fine
Yeah, it plays on my tablet too, still weird it is only my newest song
That is indeed weird, if that's the only song that can't be played for you.
First time I have this issue, the other songs I posted in showcase all play just fine
The song is fine... it plays perfectly for me on showcase room and others also confirmed.
Just your browser can't play it.
Seems like it. But only my newest song, older songs work, weird!
Don't worry about it for now... as it also plays fine in your Library.
Maybe it'd resolve itself when you tried sharing again the in the future.
Just tried a diffent browser and then it also plays fine, thanks for your support!
The feeling of going through old flutter ridden songs
and covering them
and just
listning to the flutter disapear
its insane
absolutely amazing
It has a tendency to make the vocals a bit off, but it doesn't take too many generations to improve that
yeah damn
if it wasn't for the fact that it tends to screw up vocals, usually by adding too much distortion and/or vocal fry, and robotic undertone
it'd be out of this world
Cleaning the browser cache solved the problem 😅
Hi, I recently paid for a SUNO membership (on [insert date], via [insert payment method]), but my account is still showing as a free user. I’ve tried logging in and out, but nothing changed.
Has anyone else experienced this?
Would appreciate any help or advice. 🙏
Thanks!
Outside of USA? That can cause bank transfer delays. 0-4 days
Receipt in email by Stripe/Suno of posting payment is usually the sign you should be active.
Check bank/credit account for any issues or declines.
Hey @civic heath! You want me to reshare this in #🐣┃suno-showcase?
of course mate.
Okay great, going to share there and tag you and remove from here. Any specific text you want me to include alongside it?
that's enough. no more txt, thank you very much
Happy to help!
Hey, just wondering why you can't make cover songs of other artists lyrics, obviously you can credit them etc but why can't you do it now, you could before on 3 & 3.5
The reason is obvious.. just make your own lyrics.
I doubt you can with 3.5 anymore, the system is more strict now.
I do make my own 😊, just used to make some lovely remix covers of my fav songs sadly can't do that now
I think the reason is Suno's trying to avoid potential legal headaches down the line. I'd love to make covers too, but I would understand, if I got the reasoning correct.
That's fair enough, shame tho cuz I made some lovely songs before and a remaster would be very cool but can't remaster them either 😔
Yeah I remember making covers lol it was silly and fun but it's totally understandable for Suno to not want it
Hey the Canadian national anthem is public domain! so why am i being slapped with a " lyrics contain copyright material" sticker. this wasnt a problem when i first made the song.
You can freely:
Use or perform the anthem in public.
Print or reproduce the lyrics (traditional or current).
Create your own recordings or versions.
Not here on Suno apparently. You just can't.
well i wasnt slapped with that when i made it a month ago!
Apparently not anymore!
I'd like to know how to tell Suno that my music without lyrics does not have lyrics. I've only been able to share one thing so far because it thinks my instrumentals have voice.
any one know the hotkey to move songs to trash
It might be sometimes.. it detects the instrument sound in your upload as vocal sound... as it might sound like a vocal; that'd be my guess; I don't think you can tell it that it's not though.
There isn't.
ok ty you guys should create one,it would be helpful
Feel free to make a suggestion on #💡┃suggestions-web
Is it just me or do songs still degrade in quality after about 2:30 even in 4.5
yapp.. they are 🙂
damn too bad, iv'e tried to replace the song after 2:30 like they recommended but have issues with making the chorus sound the same
What's going on?! Terrible things happen in section replacement mode! the keys jump spontaneously, there are some wheezes and creaks in the words.
the tone jumps sharply spontaneously
if i cancel my subscription and i have left over credits will they stay or disappear?
The left over will stay until the end of the cycle, then they'd be gone.
Is there any background noise, fans, or hiss or noise in the recording itself? Presuming you recorded some instrumentals in your home/studio.
Yeah, you could say there's noise. 😄 I have about 20 years worth of sound files I am sorting through.
Yea? Hard to tell really. As far as AI, vocals are just another instrument, but trying to distinguish vocals from , say, a double-reed oboe may still have issues. Any extra noise droning in the background could make it harder.
No worries. I have other ways of sharing stuff I make, and I don't need to use the website for anything but production.
It's just terrible! What's going on?! There are 32 generations and all of them don't match the tone, rhythm and level!
in replace section
Can you be more specific as to how exactly you're using the Replace feature? This sounds like a methodology problem
Agree the replace section is now totally useless. There is no more fluid flow of the music. It is really really bad.
I’ve tried to increase the length but the result is always bad. Can’t use successfully this feature anymore on my side.
Anyone else concerned by this issue?
considering people are complaining about the replace section,
I'm assuming there isn't a good method for creating an intro?
so finally playing with 4.5 after back from vacation, and it feels very "muffled" compared to 4. am I imagining things?
You don't need replace section tool to create an intro for a song.
Or are you talking about replacing an intro section of a song?
is there no way to see what song of mine is the most played, that is not published
There's no such filter within Library currently.
hmm..
Is there a command where you can instruct the AI to only play instrumental for X bars? I've tried typing in various forms but none have seemed to be successful.
I have to insert the generated section into Adobe Audition - this is insanity!
there is no such command that will clearly fulfill your prompt.
Okay thanks!
Hi! Anyone with terrible sound quality? Like washed old school Walkman tape quality.
A few but I just binned them
this is like a 19 second part of a song i love but haven't rly found a way to restructure the other part of the song effectively.
editor seems like the best bet but its like it starts confusing the lyrics??
well how would you do it in general? extend with "intro" as the style?
I use [Intro] in the beginning of my lyric
Is the Persona thing to save the vocal style?
Look here: https://help.suno.com/en/articles/3484161
Do you want to capture the vibe of one of your songs, and use it to make more great music? If so, Personas will help you get there! With Personas ,...
Thanks!
does that work for instrumentals though?
I think so.
On Create, disable "instrumental" button.
Put Instrumental in the style.
Enable custom mode, add
[Instrumental]
[Intro]
...
[<other structure section goes here>]
..
[End]
Then hit Create.
https://suno.com/s/rjZCl4DhmppzOeYe
Hi ! someone has a tip to force longer generations zith the 4.5 ?
Having more lyrics might defiitely help.
I agree, but for instrumentals ?
Might be possible via custom mode, then add more section structures in the "lyric" box (but without any lyric)
Hey guys, I was wondering how many times does it take to generate the video from a song? I used it last year and it was quite fast to generate the video with the lyrics so if anyone can help me about that i'd be grateful, thanks 😄
Do you have a subscription? It's a paid feature now
Oh I didn't know that, I thought it was free
That will probably solve the problem then lmao, thanks:
!
Where is this option ?
At the top , you should see a Simple/Custom toggle.
On the app or the website ?
You didn't show who you're replying too, I thought you were still on custom tags to extend instrumentals, right?
Ohhh my bad
Video generation is under "Download"
Ohh nice, thanks !
Why aren’t all my playlists in the app? Only the 12 that most recently have been edited are accessible. If I want to access another one within the app, I must add or remove a song, or change song order.
Any of you had issues with removing parts from a song, where it doesn't actually do it an you get an error that basically says something similiar to that it couldn't load the song?
It could happen in other functions too, like Crop.
Sometimes, it seems like it was a refresh issue, as I ran into sometimes while cropping, after a refresh, the cropped clips finally shown.
But.. I also see it didn't work for others.
I tried with the refresh, still same error - tried several times even.
Then tried closing the browser, still the same problem.
Report it if you haven't. #🪲┃bugs-web
I'm sure others had also reported but more reports might help it got fixed sooner 🙂
I haven't yet, kind of annoying, because I want to fix the track with Suno's edition, rather than bring into a DAW, because that won't fix the issue for the track on Suno.
I dont know why but it's quite tricky to get the AI to do a 3/4 time signature lol. It keeps trying to revert to a 4/4 lol
If you edit a song, does it create a new file or does it overwrite the one you have?
It'd create a new file. Edit is non-destructive.
Okay, I found nearly good version of my song but noticed on the small replacement I made there was an audible seam? Is that a common issue?
It could happen sometimes.
Remaster it would often smooth it out.
It'd be good if it didn't sound like the singer gagged on a note at the beginning XD
Okay thanks
Support is aware of it, I i know, and I've seen it "working" more often, but yea, about 1 of 7 times for me it refresh doesn't help. It is worse if I refresh too fast. So I let it simmer about 10 seconds and then refresh or sometimes check my other browser Library.
Still not working for me.
Nor is it in the browser library.
Hopefully they'll fix it soon.
meu suno parou de criar musicas desde cedo ,nao consigo respostas vou canselar e um absurdo nao ter suport e derepente para de funcionar assim
vou ter que colocar no reclame aqui ja que nao obtenho respostas
What?
Support takes awhile to reply
If that's what you're asking
I have so many failed gens thumbed down at this point that it's daunting to think about trashing all of them. I worry I'll trash a good song by accident. Is there not a way to filter for all my thumbs down songs?
looks like you can only hide disliked or show liked
suggest feature #💡┃suggestions-web
Why i receive this message ?This device was detected as suspicious and has been blocked. It will no longer be able to perform actions. If you believe this was by mistake, please contact support. What sould I do?
i can log in on my phone but not on my computer
No idea what triggers that, but the solution is relatively obvious
First thing to do is email support
The longer you delay emailing, the longer they take to respond
The email is : support@suno.com
how do you generate a song using your own voice without it sounding robotic or autotune like
or multiple layers that makes it sound like that too
Yes, in your workspace... click Filter and de-select "Hide disliked" option --> that would show your thumb downs songs.
I'm looking for an only thumbed down, so I can't see my thumbed up songs anymore
Then don't select anything else on that filter list.
EDIT: ok, it doesn't just show thumbdowns... it shows the ones that NOT thumbed up nor down as well... that filter doesn't work that way. Sorry.
You could maybe try uploading the audio and fine-tuning a persona?
Unfortunately, uploads with vocal-like sounds detected are restricted from being set to public and you cannot create personas with them last I checked
On that note, maybe generate a vocal song and mess around with coding to get something close. #📚┃suno-school message from yesterday will probably help
Can anyone tell me how to make the audio not sound so murky?
Does it give an error saying network issue or something like that?
it says "upload Failed Please try again later"
Please see points 5–7 at #📚┃suno-school message for some general tips on how to improve output quality
There may be a #1252267444773326848 entry with more tips in regards to this, such as https://discord.com/channels/1069381916492562582/1367549634536542389
Hmm don't think I have seen that issue before. Is it a large resolution? Try resizing it to be smaller, the size may be too big otherwise
The only requirements stated is that the image needs to have a height of 720px and must be in jpg/png/webp format
no ive used it on other songs, i know it works. im not sure why its not working hmm
https://suno.com/s/Dxf7uzlRXKaNxWpx see the video works on the original song
Huh, strange. I will just blame it on technology in general XD
Perhaps crop your song (selecting the entire track length) so it creates a new instance? That might fix it if there was an internal variable issue or something.
Just in case for the future: remember the network error (when people uploaded their image) I mentioned earlier? The solution was stupid: you had to zoom in, even just a little bit worked.
Hi everyone... I was on paid plan, then cancelled it for this month.
However, once I did that, I am no longer able to create any songs, because it keeps saying:
An error occurred. A system error occurred.
Any suggestions of the cause and solution please?
- Did you cancel it through the website or in app?
- Did you try logging out and back in or clearing your cache?
- Did you try toggling custom mode and/or the model version in case a variable was not set correctly?
Wow thank yoi
-
I was logged into the Website.
-
Yes, logged out, cleared cache, still keep getting the "An error occurred. A system error occurred." message whenever I Create a song.
-
Cant toggle model, its stuck on 3.5 on free account since I cancelled subscription.
Any ideas?
Have you tried another browser? Just to see if it's a browser thing or account related.
Also, what about mobile app? have you tried there too?
Hi everyone, I need help – something seems broken with my Suno Pro account. So unlike the guy in the post before i didnt cancel it, i paid and i cant do nothing now ...
I’ve been actively using Pro (paid through Apple, confirmed), but suddenly I can’t access version 4.5 anymore. It looks like my Pro status is gone.
I tried logging out and back in, and even cleared the browser cache (Firefox), but nothing works.
Same issue on both mobile and desktop.
Important detail: This happened after my friend briefly logged into their own Suno account on my devices (both computer and phone).
Ever since then, my Pro features seem to have vanished from my account.
I already emailed support (including billing@suno.com), but I really need to work and this is blocking everything.
Has anyone else experienced this or found a fix?
Thanks in advance 🙏
Yes, tried IE and Firefox...same error when pressing Create.
Yes, same error in mobile app
this happened immediately I cancelled my subscription and went back to the free account.
I just wanted to cancel for 1 month, and I am too worried to even pay now seeing I am getting this error.
I emailed support, but no reply yet
hey..ty. i gave up after sending support email to no reply, i steer clear and trying to delete ALL of my suno accounts so i can make one final one and stick with that. ty again for your help. u did help. i was too bitter and frustrated that's all.
hugs
i don't think they have the people to attend all that
mine's not replied either
Can we get a post-hardcore genre? Or an emocore? Or a screamo? Or something in that pocket? Well, if anyone bout that underground shi from 2000's on, check this and let's link up. Let's go. https://youtu.be/pq082py4lDk?si=Rt2LlqQhX6eAz-GN
Fucked with tha stash. Slammed my hand in tha door.
"J. R. Makin' Waves"
Fucked with tha stash. Slammed my hand in tha door.
Countin' blue faces but grindin' for more.
Fucked with tha stash. Slammed my hand in tha door.
Countin' blue faces but grindin' for more.
Ops caught sleepin', got tha jump.
GET THE FUCK ON THA FLOOR!
Movin' that weight....
Can I please get more credits? I’ve upgraded but it won’t renew for two more days.
If you upgrade mid month, I think you should be able to choose whether to start a new month right away, or wait until the normal end of your old monthly plan.
The Remaster feature is one of the worst features Suno has ever introduced. It has never worked for me in any version. Instead of improving the audio quality of the original song, it completely ruins the first version. You should either improve this feature significantly or remove it entirely. My credits are being unnecessarily wasted every time I try to use it.
Guys whats happening with the search tool, library etc... its kind of mess now !! i can't retrieve my songs like before... i've 20k+ songs how do i do ???!!!!
But that option never pops up for me
bring back the search by lyrics
(please upvote this)
Hi,
I completed the payment on May 14, 2025, and received the receipt from Stripe/Suno (attached). However, my account has not been activated yet, and I haven’t received any further updates or access to the service.
Could you please check the status of my subscription and let me know if anything else is needed from my side?
Thank you for your help!
Linda
Hello, please contact billing, they will be able to help you out. There is no official support team on Discord
For billing/payment issues, email: billing@suno.com
Include:
- Your Suno @handle
- Description of the issue
- Transaction ID(s) or payment receipt(s)
Before mailing:
- Check if your card was charged
- Verify funds were available
- If Stripe rejected your card, try using the iOS or Android app to pay via Apple/Google Pay.
Do not send non-financial issues to this address — it may delay your response.
They should respond within 2 business days
Hi Rabe, I have the same problem with Suno. Unfortunately I don't have any advice for you, I would need it myself. Maybe it will help you at least to know that you are not alone in this, that there are more of us like you. I lost access to my Pro account DasakaCZ@aquarius 14 days ago and Suno has not solved anything yet, except for 2 impersonal replies to my emails. It is a shameful treatment of customers, in my opinion the support does not have enough capacity to handle the problem, or they simply do not care. I have a prepaid account for a year in advance and I had to pay for a new subscription on a free account to be able to work. The only advice I can give you is to constantly remind yourself. But there is no guarantee here either. ☹️
What do you mean with "lost access to my pro account"? As of you cannot login with it?
Because that's what my issue is right now. I'm locked out of my account with the message "(my email) is not allowed to access this application" and I did not receive any information about that, also paid a year in advance. Sadly, also no reply from support til now 😦
Hey everyone. Hoping there are some 4.5 experts out here who can give me a quick hand. How do you actually get an actual deep male voice for vocals in 4.5? Was super easy back in 3.5, harder in 4.0 but doable, but now, not only can I not get the really deep, rich bass voice style that many of my songs need, but I have to fight tooth and nail just to get it to give me a voice that sounds like the singer's balls have dropped yet. I am at the end of my rope.
tried baritone?
Wouldn't have thought of that, since a baritone voice is actually a higher range than a bass voice, but let's see what that does...
Nope... Still doing it's level best to sing soprano. 🤣
what's the full prompt?
The proper voicing for a lot of my music should cap at B2, ABSOLUTE TOPS, and I have had trouble even keeping it to the third octave; it's not been uncommon to have a song where I explicitly told it to use "deep bass vocals" reach up into the fourth or even the FIFTH.
"The instrumentation should consist of only a grand piano, and the song should be sung by a man with a crooning, operatic, deep bass voice, who favors sustained notes and vibrato in his vocal style".
The rest of the prompt is all tags in the lyrics box, giving notes on specific sections, but that's the style prompt.
I'm noticing a lack of gengre
sure the voice is operatic
But maybe give it a gengre to hold on to
You are, and that's by design. I am trying to make a fresh song to base a Persona to base off of for 4.5, since trying to port in my personae from 3.5 results in immediate octave-jumping, despite having genre tags. Basically just trying to find ANY combination of stuff that will lead to the right sound. What the heck, I will give adding a genre in a shot again... Maybe it will work. Thanks for trying to help; here's hoping!
Use the creative wording 4.5 likes too maybe
A *genre here* track whose instrumentals consist exclusively of a dynamic grand piano, lead by a deep and crooning operatic voice. The bass vocalist carries the song with sustained notes and vibrato, giving it a warm and smooth atmosphere.
as an example
progress?
is anybody else having trouble saving personas with V4.5?
Not here, no 
Why I paid on mobile app via Play Store but why can't I use Version 4.5 on PC (Website)? Please.
Hello, I'm reaching out to you because I currently have an annual subscription that I’m paying for, and I can’t 'make a persona, thank's !
Hi crysis 992. When logging in via phone, I didn't receive a verification code, a constant message: try again. After several attempts, it reports too many login attempts and blocks. Still unresolved and no information, only that the request was accepted and someone? will deal with it. It would be more acceptable for me if someone wrote to me that it will take longer until xx dates., than that I don't know anything at all and so I don't know whether to build a new account from scratch or hold on and wait for renewal, the uncertainty is the worst. The account that I manage now, I had to pay again for because of the sound quality and ownership of the author's rights to my texts. So I pay twice, get once! This is really not a good advertisement for Suno, although I otherwise love it very much for the options it gives me for my work.😕
Hi mixtrix, try asking ChatGPT, I often use it for advice on prompts and it worked for me. It recommended everything directly to me for input into Suno. If you define well what you need, it will guide you how to request it. 🙃
I'd like you to improve something that I think is a problem with "personas." When you create a "persona," that persona is associated with a prompt. When I try to change the prompt for another type of music but choose a specific persona, the system ignores me and doesn't create the track with the instructions I ask for, but instead with the instructions associated with that person.
For example, I have a "persona" I call Annita. She has a slightly digitized voice that I like, but it's associated with EDM music. When I ask for that voice to sing pop or rock music, there's no way. I click Create and it always comes up with the same EDM song. And it's always the same melody... 😦
Never tried yet, but what if you cropped the original song to 1 min or so of the part that sound best then create a Persona from that... then use it in another song with totally different genre. Would that work? would it still keep the vocal preference but allow to work on the new song?
Awesome idea, thanks!
Are duets/multiple vocalists easier to do in 4.5
Hi LiraBot, I completely agree with you on this, remastering usually cleans up and smooths the sound a bit, but it takes away the expression and rather highlights the errors in the pronunciation of the text if they are not corrected. I am also throwing away credits unnecessarily on this. Recently, I tried a 2x remaster experiment, the first time from 4.5 to 4 and then back to 4.5. Oddly enough, it helped, but it could have been a coincidence. :-))) 😉
do you receive any kind of error message when trying to make a persona?
Subject: Urgent Account Access Issue – Username @ceoadonisking
Dear Suno Support Team,
My name is Adonis King, founder of Black Goat Life Music LLC, and I am writing to urgently request assistance regarding my Suno AI account (username: @ceoadonisking,
I have sent multiple emails over the past few days to support@suno.ai, hello@suno.com, and contact@suno.com. Unfortunately, I have only received automated AI responses from "Mike and the Suno Team" that have not addressed my issue or provided a solution.
Here is the problem:
My account dashboard states that my next billing date is June 8, 2025, indicating that my subscription is currently active and paid.
Despite this, I cannot create new music—I’m receiving a message that I’ve used all my credits.
This is preventing me from using the core functionality of the platform, even though my billing status shows I should still have access.
This is disruptive to my creative and professional work, as I rely on Suno AI to produce and release music.
I am respectfully requesting that a real support agent contact me directly to resolve this issue immediately. I need:
Confirmation that my account is paid and active.
Access restored so I can continue using the service I’ve paid for.
Clear communication from a human support representative.
Please treat this as a priority.
Thank you,
Adonis King
Founder & Artist, Black Goat Life Music LLC
I hope you also sent this to support@suno.com?
/support
Heads up: Discord does not have any official support team members.
For all non-financial issues, please send an email to: support@suno.com
Include:
- Your Suno @handle
- A short summary of the issue
- What you've already tried
- Any relevant screenshots or references
You'll get an automated reply when your email is received. Sending multiple emails about the same issue may reset your place in the queue.
Im also not able to contact suno, accidentally bough a 1 year subscription when my intent was to buy a montly subscription
