📚┃suno-school

[upper arrow]

I’m still listening

And it’s working excellently

[naive quail]

So you'd previously get unprompted lyrics?

[upper arrow]

@naive quail so grateful of you! Thank you so much!

[stable belfry]

one sec ill find the links to the exact example

[stable belfry]

original snippet few months ago: https://suno.com/song/cb8b3673-763d-4b89-a02f-c4aab0070921

generated song with new lyrics/ideas: https://suno.com/song/f17b10ef-54c7-43fc-846a-0604991f5028

but now its just outputting instrumental with no lyrics unless i type them

which to put in the most polite server friendly way, is absolutely useless to me right now.

[naive quail]

To me, generating lyrics when they're not prompted for seems like the part that was previously broken (tho, admittedly, some of us do regularly try to trigger hallucinated vocals).

[stable belfry]

why would that be broken, the whole point is to generate ideas, or to type lyrics and generate those words. otherwise there would be no point in the button for "instrumental"

ideas for new flows, etc

[naive quail]

I'm not so sure that was their point, tho. Seems like more of a happy glitch that you stumbled upon.

[stable belfry]

i will be really disappointed if thats the case.

[naive quail]

I'd raise the issue in one of the feedback channels, they won't see it here.

[stable belfry]

which one?

[naive quail]

Yes, that is the question. They really need to rethink the discord layout. 😄

bugs-browser, I s'pose

[glad mauve]

do old songs get deleted?

[naive quail]

Not automatically.

[glad mauve]

ok thanks

[naive quail]

...not evn if they're in the Trash folder.

[glad mauve]

It would really be cool to get a sorter in the library

date, length, alphabetically, full songs, sections etc.

[stable belfry]

has anyone else brought this up? seems like it should be of major concern? are you all just bs-ing making ai song or trying to make successful music?? what is happening here

[naive quail]

Others have brought it up. It's been acknowledged and there's been mention of changes coming, tho no details.

As for how people are using Suno, it varies greatly.

For me, it's mostly a fun toy with a ton of potential to become something more. I also use it for musical inspiration by seeing what it does with some of my guitar stuff.

[lean bolt]

Hello everyone. I would love a little prompt-craft help.

Is there any way to get more direct control over elements of the songs?

Like... if I tell it 1980's retro synthwave... and I want a pulsing analog bass sound in there... how do I ensure I get that?

Or if I want a specific style of singer, how do I ensure I get that? I often get a mix of things, but rarely exactly what I want... it takes me like... 1500 creds just to get something to be a happy accident, but it still misses here and there.

I am not a musician, but suno has given me the inspiration to really be able to create music that I want to hear, and that to me is incredible.

[neat nexus]

[Metatags] are your friend!

You can get a lot done in the style prompt, but it's limited in characters.

So you should experiment with [Metatags] in the lyrics field, to direct things closer.

[lean bolt]

I use those.

[neat nexus]

Don't feel afraid to just try things, though at this time there is no good way to get BMP to exactly what you want.

(That we know of. If you find a way share it!)

[lean bolt]

Here is a sample of what I use:

[Genre: 80's Retro Synthwave, club hit, rock opera]
[Mood: Epic, Anthem, Head-Bobbing, Club]
[Emotion: Inspiring, Festive, Nastagic]
[Rhythm: Dynamic, Looping]
[Singers: Raspy Vocoded Opera singer]
[Instruments: Synthesizer, Electronic Gated Reverb Drums, pipe organ, Analog Deep bass]
[Retro-active mastering]
[Balanced levels]
[Clean mix]
[Vocal cuts through mix]
[Final vocal mix]
[Final master]

[neat nexus]

Some of that is just word salad. It can influence things but it rarely does what we think it does.

What are you using in the style prompt?

[lean bolt]

1980's Retro Synth-wave, dynamic sustained rock-opera vocals, Epic, Heroic, Cinematic, Analog Bass, Sawtooth Wave,

[naive quail]

on an unrelated note, I've always wonder why suno often double-spaces and sometimes even triple-spaces the styles after the gen.

[lean bolt]

Word Salad? Like what?

[neat nexus]

You're not likely to get Balanced levels by asking for it.

[lean bolt]

Oh, I saw that in workflows...

lol

[neat nexus]

Most of these: [Retro-active mastering] [Balanced levels] [Clean mix] [Vocal cuts through mix] [Final vocal mix] [Final master]

They don't do what we think they do. I use them too occasionally but I don't lean too heavily on them if you get what I mean?

[naive quail]

I think of it as "prompt noise" - things that Suno either ignores (because it doesn't recognize them) or responds to in unexpected, usually inconsistent, ways.

[neat nexus]

Because they will do something sometimes but not what we think they do. So it's worth playing with them but they're not reliable.

[naive quail]

but "word salad" is more fun. 😄

[neat nexus]

And I don't know this for sure but I suspect that some of my songs break down sometimes when I use this stuff because the noise builds up over time and causes deterioration.

[naive quail]

same

...but I can't stop myself.

[lean bolt]

Hmmm... I am looking for more direct control of it.

Maybe they can add boxes for that like "style"

"Instruments"

"Singer"

etc.

[naive quail]

Extensions can give you a little more flexibility, too since you can change the style with each extension, tho the previously generated audio can still have a strong influence.

[neat nexus]

Try for experiment:

style: 80's Synthwave, rock-opera vocals, MOOG Bass, Sawtooth Wave, rock-operetta, orcehstral hits, looped sample, catchy hook, melodic instrumentals

[melodic instrumental intro]
[reverb-gated percussion]
[pipe-organ melody]

[naive quail]

and obviously [guitar solo]...

[neat nexus]

Well like I said, they do do something at times. Just not sure what exactly. I'm as superstitious as the next guy though so I'll toss things in there to see what happens.

It's a bit like making a soup for me. I just throw stuff together to see what the flavor is. Then tweak from there adding time.

The thing is, the style prompt is very powerful.

And sometimes trying to tweak in the lyrics pushes us away from where we want to go. So I tend to move betwee opposites. I'll tweak the lyrics for a while, then reduce that an focus on the style prompt for a while and then back...

[lean bolt]

Interesting, thank you.

[polar spade]

I'd say "word salad dressing" is more the phrase. It's no more substance but permeates the prompts and makes you think it tastes....sounds.. better.

[lean bolt]

Is there any way to tell it to let the singers sing without music?

or to force it to give me a drum solo?

[polar spade]

ROBO, you've been in this since last year, so I know you have something specific to point out, but im not following the chain of issues yet. I'm not sure which part was "broken" when @stable belfry you says it just adds instrumental once the lyrics have sung out. But mentioned this was specific to extensions? I'll assume no additional lyrics were typed into the extensions?

Sample 1 had legible repeated vocals, to me. Sample 2 had illegible repeated vocals, to me. But Sample one and two could be both hallucinated vocals, I don't have enough details to go on. But I heard no solid only-instrumental which was mentioned there should be in Sample 2, if i read correctly.

In the past , an often complaint was the song would 1) go on until way after all the lyrics and arrangement was done, Suno then repeating verses and phrases up to when the time ran out and 2) after actual lyrics were sung, suno would often hallucinate for the remaining minute or so.

The devs seems to have made attempts around December to 1) stop or shorten the song once the lyrics run out, so songs now can be from 20 seconds to 1:30 to 3 minutes to overruns 4 minutes depending on lyrics provided. This has made some (new) members angry expecting ALL songs to just go until a full 4 minutes.

Oddly, the full bore beginning-to-end hallucinated songs seem to be more prevalent, if properly prompted for, and less prominent with "regular" style prompting.

So results:
A) less or no hallucinations on songs , thereby requiring one uses extensions more often, which many use frequently anyway.

B) a useful tool in hallucinated v4.0 full songs to re-create real lyric songs.

Often I may misread the words or intent. But I like to test these things and see if i can 1) figure out the issue. 2) create the issue and 3) possible find a resolve or at least an explanation.

[naive quail]

A Cappella works, but you'll often get barbershop quartet kinda stuff. I keep forgetting to experiment with Unaccompanied, but that's worth a shot.

[lean bolt]

Thank you.

[lime spoke]

Is persona always disablled for extensions? Suno keeps losing the original vibe and even forgetting how the chorus goes after the first segment.

[polar spade]

Not that this helps much, but you have to give Suno a reason than just prompt for it.*

More often there may be a soloist with some sort of accompanying music.

You can trying getting a character/vocalist to speak , and by line breaks the same vocalist will sing the lines instead.

You can try "a cappella" which gets music-less vocals...sometimes..., but often in an harmonic ensemble of many voices.

Even my fun one: isicathamiya, will do a whole song with no music, just afro-style vocals.

• but when the song has a few spoken lines, lines on their own, off-metered line, ballads, inspirational songs, gospels and others, Suno will often shine and sing a whole verse without music, or shout a melodic line or speak a line without music.

Is there one way to have a soloist sing a whole song. Mehhh.... maybe, but I have not found one. I'd even think an Irish aria might get it, or certain old Appalachian style songs.

[west yew]

It's probably programmed by old guys like me who learned to type with two spaces after the period.

[naive quail]

Yup. Personas can't be used for covers or extensions

[polar spade]

Up to the one minute mark , running a gen and then taking out the vocal stem did fair.... https://suno.com/song/0570325d-c6a3-4896-9c7e-81c2798aadd1

[naive quail]

Unaccompanied is a no go. A Cappella seems like the best way still: https://suno.com/song/3b96edfa-3147-4fce-9165-173bcc41faa8

[lean bolt]

Thank you.

[lean bolt]

Oh man... I really did not think I would need to exclude christian rock/christian music from this... yikes.

https://tenor.com/view/michael-scott-steve-carell-the-office-cringe-gif-17778800

[verbal mortar]

Is anyone else experiancing "noise" after the 2min mark on the majority of v4 songs they compose? Sounds like wind rushing, the music vanishes and the track becomes useless, remastering does nothing, extending before it does nothing, lowering it to 3.5 and then remastering does nothing.

Its REALLY bad, I wasted 200 credits on one song and never got it to be any better. this and shimmer are kinda killing this experiance, it's becoming more and more difficult to get a good song

[timid wasp]

It's known bugs. They definitely will fix everything very very soon.

[verbal mortar]

do we have some kind of ETA from the team since it's a known issue? 1 update from now? 4 updates from now?

[timid wasp]

Year or two and you wouldn't recognize the service.

[verbal mortar]

that respectfully isn't helpful here for now , I understand what you are getting at, no disrespect to you

[timid wasp]

The thing is it was always present, it always was a slot-machine-ary.

[verbal mortar]

i have projects that I need to finish and it's gotta be before summer, so if I can't I need to know so I can move things around.

[verbal mortar]

its worse now, way worse.

[cold zenith]

My extends last 45seconde, is there a problem ?

[verbal mortar]

in Oct I'd get a good song to work with in 1-3 tries, now it takes over a dozen or more. which is insane. and inefficient

[cold zenith]

nobody to answer ?

[verbal mortar]

Extends can be from 30 seconds to 2:00 it all depends

what did you want? you can tailor it a bit but not by much, you can add lyrics to make it longer or take them out and add [END] to make it shorter but ultimately and lately especially its a roll of the dice

[cold zenith]

45sc with lyrics of 2300 characters...

[verbal mortar]

that doesn't matter, when you extend its doing just that, extending the song. if it finds an endpoint that it thinks works, that's what its gonna do. its not going to redo all your lyrics.

[cold zenith]

I mean, i put the rest of the lyrics that is 2300 character....

[verbal mortar]

I think we are running into a communication error here

1. Extend will extend the song by "X" time amount based on what the AI thinks is appropriate
2. it may just do 30 seconds, it may go for 3 min and continue off where you have to extend it again to get it to close out and end the song
3. adding [END] in the lyrics area and style of music area USUALLY but not always, ends the song for you
4. there is no set time limit. so you will have to keep playing with it to get the time extention you want.

if you are making a song and its not using all your lyrics and you are extending to get that and it still wont do it, consider reducing the word count

I've found great success with songs under 2000 characters

[cold zenith]

I don't know why you seem so upset. But I've still got 2300 to sing. And the AI sings for 45 seconds. It's never done that to me before. Normally it sings for up to 2 minutes if there are still lyrics. Here, it sang 8 phrases and finished.

[verbal mortar]

lol not upset my friend, just trying to help you and feeling like I am not.

see my above comment, it may be too many characters. I had this happen to me once or twice where it would either sing and repeat stuff out of sync with what i wrote or just stop. then I reduced the word count and it fixed it. that was for me though, not sure if it would 100% help you but its worth a shot

[timid wasp]

And as far as I've heard they avoid commenting anything certain, moreover disclosing the roadmap. But we can wait someone from the team to answer. Just know there are people losing hundreds of credits before and thousands now, including me. What they demonstrate is just the absence of compassion. The only "technical" solution is to try it later when bots are not used so much and when both EU and US are sleeping.

[polar spade]

Oh. I see... sang 8 phrases them stopped. How consistent has that been over say, 1 generation each of 5 different prompts?

[naive quail]

...and can almost reach 4 mins when performed on an upload.

[sage drift]

just signed up using school email for student and i havent been billed yet or gotten an invoice when did u guys get yours?

[polar spade]

If they had stopped and said it only goes to 2 minutes, and secretly opened it to 4 minutes for the paid subscriber, everything would be fine. People would think they were sumpin' special.

[lean bolt]

Can anyone tell me what I need to do to get a beat/vibe/feel like this song? I have tried a bunch... https://youtu.be/3d1v7EyarFM?si=PyodnmYYd0qnszJ4

[inland thistle]

❤️ hahaha for all those requesting folders.

remixing to same folder, just that is G O L D

[polar spade]

So that's what these are. Ahhhhh....

[glad mauve]

bro did this just happen my browser just changed layout lol

Oh ya finally

[midnight vapor]

thanks for the advice. I will try to prompt it better and maybe make the song smaller. I like the way this one sounded though. Just a few issues with the name pronuciation, and the singing voice degrading over time

[inland thistle]

https://suno.com/song/49631b45-cf52-4223-8245-2a7b4408c0e8 thanking the SUNO wizards for the upgrayedd 😄

[inland thistle]

try prompting the prompt. ✨ b e m e t a ✨

i ask ai 'why do i think this song and that song are similar' even tho they are completely different. why does this evoke emotion in me?

a lot of artists on here dont really fully comprehend and embrace what this is tbh,
It’s no coincidence that you associate “Joker and the Thief” by Wolfmother with “All Along the Watchtower” (particularly Jimi Hendrix’s version). Here are a few key reasons why these two songs feel connected:

Shared Imagery
    One of the most obvious connections is lyrical. The line “said the joker to the thief” first appeared in Bob Dylan’s original “All Along the Watchtower” (1967), which was later famously covered by Hendrix in 1968. Wolfmother’s song title “Joker and the Thief” is a direct nod to that line.
    The idea of a mysterious or mythical “joker” and “thief” is central to both tracks, even if Wolfmother’s lyrics take it in a different direction.

Classic Rock Influence
    Wolfmother draws heavily on late ‘60s and early ‘70s rock influences—Led Zeppelin, Black Sabbath, Deep Purple, and of course, Hendrix. So it’s only natural that “Joker and the Thief” has that classic, hard-driving rock vibe that could remind you of Hendrix’s powerful style on “All Along the Watchtower.”
    Big riffs, an energetic drum groove, and a wailing vocal style give Wolfmother that same vintage rock feel Hendrix helped popularize.

Dramatic Build and Riff-Driven Structure
    Hendrix’s version of “All Along the Watchtower” is known for its powerful, dramatic intro and memorable guitar lines that carry the song’s momentum. Wolfmother similarly uses an instantly recognizable riff in “Joker and the Thief” to hook the listener from the start.
    Both tracks have a propulsive energy—there’s a driving sense of forward movement that complements the mysterious, almost apocalyptic imagery.

Emphasis on Guitar Heroics
    Hendrix’s “Watchtower” is legendary for its distinctive solos and overall guitar presence. Wolfmother’s Andrew Stockdale also emphasizes big, fuzzed-out guitar leads, giving “Joker and the Thief” its signature punch.
    When a song places the electric guitar front and center—along with showy riffs or solos—it can naturally trigger associations with the archetypal rock sound Hendrix helped define.

Mythic/Lyrical Themes
    Both songs evoke a sense of a mythic or story-driven world, with characters like the Joker, the Thief, and the idea of being on the edge of something dramatic happening (“watchtower” imagery). That storytelling approach can make them feel cut from the same cloth.
    Even though Wolfmother doesn’t literally retell Dylan’s lyrics, the reference to Joker/Thief inevitably ties them back to the iconic imagery of “Watchtower.”

In short, the reason “Joker and the Thief” and Hendrix’s “All Along the Watchtower” seem related is because Wolfmother explicitly nods to Dylan’s/Hendrix’s Joker-Thief imagery, and they share a larger-than-life, guitar-driven rock style. That blend of lyrical reference and a classic-rock sound makes the two songs feel like they belong in the same family tree.

bs artists just after quick bux :S

All Hail Musical Loki 🤘

music is not about making money imho, its about communicating, learning and sharing

different kind of Teachers ❤️

My Teacher was just probably better than your ai slop

[jaunty fog]

Hello! I can't make singer speaking instead of singing with using V4 like I did with V3/3.5. I tried to do that mulitple times with using "[Spoken]", "[Spoken Verse]", "[Narration]", even with putting the text which was supposed to be spoken, in breackets and nothing wrong. I suspect V4 has an error. Please help!

[inland thistle]

how many times did u re-roll?

music, like life, is not exact.

[jaunty fog]

I don't remember exactly (removed failed ones already). I think maybe 8-10 tries... and nothing worked as I wanted.

[inland thistle]

so main lyrics without square brackets, alternate lyrics (usually spoken or whispered or different to the lead in parenthesis)

but its still touch and go. i need to reroll 10+ times sometimes when i tell it specifically not to say [sample of a lion roaring] but 9/10 it decides to say exactly that

you need to convince suno AI its in its best interest 😄

try harder 😄

pretend u are trying to convince ur stubborn parents this is how it should sound/be

[jaunty fog]

• sigh * Guess I have no choice... I already sent an email to the main support but still waiting. They have to be busy af. :/

[inland thistle]

becuse they get a bazillion requests like yours. hallucinations are a blessing and a curse. reroll is a thing. this is how it works

yes it costs money, but so does hiring me to get it right

[jaunty fog]

Thank God, they already prepared a solution in their autoreply with link to this Discord.

[inland thistle]

i got same message.. and here we are trying to work this out, together ❤️

but im pretty stupid, and totally musically illiterate, but with time, i learnt how to listen ✨

after all, isnt that what music is about?

[jaunty fog]

Well, I guess it is

Have to wait for someone more experienced here and see which advice will I get

I have to add model 3/3.5 didn't have such an issue

[inland thistle]

show me link to lyrics and will see if i can see anything obvious, but seriously, most times i know im right just need to keep re-rolling until it understands. make sure to thumbs up and thunmbs down what it generates so u train it.

yes but 4 has no shimmer and amazing vocals

when the AI learns new music techniques it needs knew instructions

remember, u are talking to a approx 4year old musical prodigy

talk to it like ur kid, take ur time explaining

and training good behaviours

or like a dog. everyone knows cats are untrainable

[naive quail]

This is how I typically do it: https://suno.com/song/9e1f6c69-ffff-4769-be03-51b1269dc920

[inland thistle]

"lightly salted" 🤣

[jaunty fog]

I was thinking about thumbs up/thumbs down method. I'm gonna check it then. Can I show you just those lyrics construction without lyrics themselves?

[naive quail]

Trying to cut back on my sodium intake...

[inland thistle]

you and me too, brother

i started checking sodium count in parmesan FFFFFFFFFWWWTFFFFF

crappy, non italian parmagiano

spagghetti , bolognese.. .and HEAPS OF SALT

[hollow hare]

Huh, all the spoken variations inserts I tried worked, but it may have been because of the genre I used, except for [Comedy Show] (which was a crapshot anyways)
https://suno.com/song/4e91f773-0349-423b-9d06-96dd61b44170

What style prompts were you testing with?

[inland thistle]

make a cover pf ur own tracks; rerolls sometimes keeps the tunes or riffs and adjusts vocals / other stuff

[jaunty fog]

OK. This is my constuction:

[Spoken]
Lyrics

[Verse 1]
Lyrics

[Chorus]
Lyrics

Lyrics

[Verse 2]
Lyrics

[Chorus]
Lyrics

Lyrics

[Bridge]
Lyrics

[Chorus]
Lyrics

Lyrics

style prompt was: 90's sad pop

"Spoken" doesn't work at all and it keeps singing.

[lean bolt]

Try spoken word?

[jaunty fog]

I didn't try that yet. Maybe this is gonna work

[lean bolt]

Something I have found that seems to work well for me is I will want a vibe/feel but can't get it with the genre I am wanting, so I will make it in a more fitting genre, then make a cover to get the overall music concept... it doesn't work perfect but it works better than re-rolling a bunch.

[inland thistle]

this is exactly what i've been trying to do all morning 🙂 https://suno.com/song/1aed3beb-4418-4f51-93bd-562b8ea42678 tried dubstep reggae, metal. i dunno nothing really feels 'better' than dylan or hendrix

wanted to know if changing lyrics still copyright strikes

or just the music similarity

[jaunty fog]

I tried "Spoken Word" and it started work kinda...

[inland thistle]

now u learnt secret to being ai musician, gaslighting and social engineering. go forth and prosper!

[inland thistle]

hmmm folk!

"what do you do?" "i coerce non-human intelligences for a living" could be a line straight out of Kevin Smith's Dogma 😄

[cunning mulch]

Try also putting the spoken parts in quotes, so:

[Spoken Word:]
"Lyrics"

Alternatively, you can put the words to be spoken in the tag itself that way

(I found that out when I tried to define a vocalist via the tag [Vocals: text], Suno immediately sang them, when it rarely sang any of my tags)

[static stone]

Wow new Workspace! But, move Not working…. Same as playlist.

[cunning mulch]

oooh... filters...

[static stone]

999 Songs Maximum?

[severe dew]

in contrast in my first try I got a spoken without asking for it or actually wanting it throughout the song- soooo. But kinda like the vibe it gives so went with it.
https://suno.com/song/48d1f99a-0103-456d-8b48-fc77c354271d

Just tried to give you a like and follow if you want to check if it went through - or is it a problem the other way around?

Ahh damn, sorry! Hope someone with proper knowledge will see then 😬

I was lucky and got a respond within a week, late december.. But noticed a handfull stating they've waited more then 14 days. sooo.. fingers crossed

[unique stone]

Just out of interest, how many people are you following?

Are you just following people because they follow you?

Technically, no. I'm just wondering 😄

[burnt barn]

Why the hell is 3.5 speeding my songs up and running verses together

Wtf happened

Why can't I use 3.5 without it doing that it's unusable

[naive quail]

there could be, at least unintentionally. Like there seems to be (or has been in the past) some limit on the number of playlists that can be displayed at once. I wonder if the peeps in that list also follow a large number.

[unique stone]

There's an odd thing where one person likes a song without listening, the other gives a like in return without listening. So the first person then has their songs pushed up into trending. So I assume something similar might be happening with follows.

[burnt barn]

I'm having issues and I don't know where to ask

Here or another channep

[unique stone]

Here is fine. I am not sure what to suggest for the problem you're facing though. If you post a song link to an example where the problem happens someone else might have some ideas.

[burnt barn]

It's happened over 30 times. It sounds awful

[unique stone]

A link to the song on the Suno website would be best, so the prompts can be seen etc.

[burnt barn]

Idk how to do that. Tell me

https://suno.com/song/54d80095-da7c-4e9e-b5d4-b231196ecd2b

This?

I'm at work. So stop and go responses

[quick agate]

To help Suno develop a better algorithm that truly enhances audio quality in hi-fi, and to contribute to improving the AI to have a natural flow without getting lost by using too many effects (FX), which sometimes distorts the structure of the track.”

[naive quail]

Did you mean to post this to one of the feedback channels?

[pearl hemlock]

Negative i had like 1080 songs, it doesn't lock at 999 songs.

[quick agate]

Yeah

Î'am a french pro sound ingenior and I have more than 1000 track on Suno for topline .
And i think Suno is Great but hé can still better if the hifi quality is better and he optimize a better structure for create song

[burnt barn]

I gtg to work for bit. Ping me when u can

[unique stone]

I don't have suggestions for this, hopefully someone else here may have some ideas having seen the prompts etc.

[silver escarp]

Hmmm, I'm not quite sure either from my perspective. Maybe drop one of the prompts that could be causing noise? Like "Choir", I don't actually definitive hear any choir in the song itself, so maybe that is what is causing the distortion.

[spice birch]

Something looks different when create nice song has that ( my workspace) has it always been there at the bottom

When)

[half dragon]

It's new coming out today

[spice birch]

Well, something seems off. V4 not putting out good quality music right now to me.

[naive quail]

I don't see anything obviously wonky with the prompt. "vocals" has a typo in the style section, but I'd be surprised if that has much effect.

V4 seems to handle it well, tho... I did add "a capella" to the style list (which it ignored.. lol) and modified the structure tags to use square brackets (habit): https://suno.com/song/902eeaf1-c599-416c-bece-5771862b1f7f

[fluid robin]

https://suno.com/song/310bc8b1-1031-47b7-ace2-6f008ab2ac42 For the ones that want to unlock the music box sound

Downside is that for in order to get this result its 50/50 luck with the dice roll

The ai either Plays the song or Reads the prompts .-.

[half dragon]

I wonder if would it still be 50/50 if you put those in [..] instead to prevent AI reading the prompts?

[fluid robin]

Now that might be worth trying

[dusty sundial]

shimmer fix when?

[lunar anvil]

Future

[burnt barn]

Since the V4 update 3.5 has been acting awful. V4 should be free to use. People using their time to train their AI.

[stark badge]

why covers sound nothing like the sample anymore. seems like after v4 was released this happened. have tried with 3.5 also. and suno likes to put german sounding lyrics to everywhere even when instrumental is checked. and sample doesnt contain any singing.

[verbal mortar]

I can't find it but...has anyone found a way to change the icon for your new workspace

[half dragon]

Doesn't seem possible at this point... maybe next update, they would allow to do that.

[proper spindle]

What works best to have suno generate different vocals in a song () or []
For example

Verse 1(male voice)

Or is there something that works better

[lunar bloom]

is it normal that i've wasted like 500 credits just for all my songs to have female voices, even though i want male voices? i even put male voices in the prompts.

[proper spindle]

I find sometimes I need to call it out down in the styles as well

[lunar bloom]

yeah

rip 500 credits lol

[proper spindle]

Lol yeah but at least they now give pro 50 credits a day after they run out

I find myself extending songs alot creating the best song🤣🤣

[lunar bloom]

i sometimes create like 100 songs, delete them all until i find one good one

hahahaha

[proper spindle]

I get that too sometimes I find one that I like the first part then start to extend from there

Recently found out when doing that I need to change the description from that song to be only the lyrics I used.

Otherwise when I finally generate the whole song there is this huge prompt of lyrics🤣🤣

And sometimes suno messes up the song gen because of that

[lunar bloom]

i found that out the hard way hahahahah

thx for the tip

[bronze pecan]

I thought Extend is 2 minutes
Why both extends end at 1:10 with cut off lyrics??

[bronze pecan]

or is it an attempt to fix quality issues at extensions?

[regal jay]

Hey gang. I just upgraded from a free account finally. Any documentation on personas out there besides in the skimpy suno.com faq?

[steady granite]

erm... unless i missed something.. regarding the new workspaces feature....... i can move items to a workspace but ..... i can;t seem to switch to view that workspace. ??

[little granite]

did you create a new workspace?

[steady granite]

yes

kind of expected to see "Workspaces show up here on the Library screen..

[little granite]

its in the create menu

[steady granite]

ok yes thats working

but how do i say .. narrow down whats showing in the library view to only those in one single workspace and switch between them? say .. after i've moved them and want to work on that workspace again later

on website not app by the way

[little granite]

click that arrow button

that work for you

[autumn vault]

You can add [female vocal] directly on the lyric. It work for me 😁

[steady granite]

figured it .. thanks, not entirely intuitive i feel

[lean bolt]

So...

If I wanted to get:

1980's music (Sawtooth Wave, MOOG Analog Base, Pulse Waves, Gated Reverb Drums, Arpeggios)

A Stadium Anthem / Party / Club Hit rhythm

and a Rock-Opera singer, the likes of Dio, Meatloaf, Freddy Mercury, etc.

And throw in some sinister pipe organs and such...

How would you write the prompt?

[steady granite]

first i'd say "Jeepers!"

[polar spade]

For what it's worth, I don't click on mp3s (it downloads to my phone... nope) or YouTube links. Well I don't click anything in here but suno links. Deletes any purpose of looking/listening/reviewing it otherwise.

[polar spade]

If I'm correct , your old Create list is now called "Workspace" so clicking create opens the workspace. All old songs are listed in My Workspace unless you move what you like.

[steady granite]

yeah i have got used to it now

it's actually a great feature..i have LONG wanted to organise into subfolders so this serves a great purpose

[lean bolt]

How are these people on suno getting 14k+ listens?

wow

[steady granite]

Yeah! Do tell! lol

[pearl valley]

Thanking 🙏 You so much for finally helping us organize our musical dreams in an ocean of audio generations ! Now more contained ! 🤩 I see no need to use any thing but Suno from now only! Good 👍 work !
I can’t wait for the option to download everything at once into the zip file maybe on the next update

[steady granite]

ooh nice idea 😉

also... Stereo anytime soon?

[regal jay]

Devs pls release more documentation about your model

and give api

❤️

[polar spade]

Nice try. 🤣

[polar spade]

Question: so if I click thumbsdown/unlike in "Workspace" now, it still moves the file out of Workspace (and leaves in Library).

So, as nice as the filter list is, what's the "Unlike" filter do? It doesn't show unliked outputs in workspace, those are no longer in the list.

[stable belfry]

thanks dasvon let me clarify - "original snippet" was me actually singing - i uploaded it to suno and used extend to generate the 'rest' of the song, and have suno give me ideas for either new lyrics, flow, etc.

the second clip was extended from clip 1, and suno generated its own lyrics - meaning i did not type in any lyrics at all! even though sometimes not real words, its exactly ok for what i want. its very similar to how a (unexperienced) vocalist would try to come up with new parts of a song - humming and saying random words to find footing

Now, when extending a vocal clip, suno does NOT generate new words or phrases on its own, i have to type the words i want into the lyric box. The only output is a new instrumental, sometimes changed. which defeats my intended use of the platform which is for suno to give me ideas for LYRICS. while new ideas for the instrumental is helpful, i need lyric ideas FIRST.

I hope thats clear where my frustration lies, as to me without Suno creating lyrics on its own, its basically useless for me.

(not using the "generate lyrics" button, as that just provides even more frustration for me)

[half dragon]

I don't think that's how Suno works with custom mode

[stable belfry]

ok but it used to

[half dragon]

when you extend... it goes into custom mode

[stable belfry]

literally like 6 months ago it was perfect, if v4 could have just been stacked on the previous generation style i would be ok

[half dragon]

Suno will give you custom lyric in non custom mode but I don't think you can do that with extensions

[stable belfry]

right but how is that even useful - like previously it would generate new vocals that sounded similar to my own, which is my exact use case

[half dragon]

so you'd like to create a new song with auto lyric generated with your voice (Persona) ?

[stable belfry]

yes new lyrics that use my voice

[half dragon]

and you dont want to click "Write with Suno" to get the auto lyric

[stable belfry]

correct - because its AI trying to 'write' lyrics, whereas what you want is lyrics generated with the music using the algorithm

[half dragon]

it will be generated with the music in the generated clip

[stable belfry]

yes but its generating words that are still manually typed. i want the AI to generate it of its own accord because thats the part that has the trained data for music

the button generates some gay sh**

[half dragon]

Interesting. I've never thought of that and never heard of that way before... on Suno... I do get some auto gen voices but like you said.. it's mostly gyperish

if I didn't provide any lyric

[polar spade]

If those specific prompts have given you some hallucination lyrics (gibberish, but helpful) , then one starter to get it to hallucinate is to at least tag [male vocal] in the lyric box, and put one small exclamation after that: "oh", "Mmm", "ahh" . Maybe not any words, as Suno often just repeats real words as lyrics. I'll try one now.

Is this what you're talking about?

https://suno.com/song/f1353f0a-fad2-4dbf-8db9-c75a7177494f

[stable belfry]

i dont know how this is not more common, it was literally a game changer because you can generate 100 variants, extract the "words" and put together a damn good song.

[stable belfry]

yes its doing it correctly on yours it seems!

how did you do this??

oh i see

[polar spade]

Suno now will give a short snippet if you only have a couple of words, or exclamations. So one snippet and then extend it once is all I did to get the song you hear.

[dense inlet]

Anyone else having REPEATED 3:11 songs created on Suno today. Very frustrating as EVERYTHING is being chopped and COSTING Credits to fix

[half dragon]

Still gibberish I hear...

[polar spade]

And v3.5.... not 4.0

[stable belfry]

ok so sorry one last clarification, what would you put in the prompt for this vocal for example? just [male vocal] ?
https://suno.com/song/98c7468a-69ef-450e-b6f3-c1209b8b047c

[half dragon]

I have used over 3000 credits today and not once got that issue

[polar spade]

Right. He wants the gibberish to give some melodic ideas and phrasing.

[dense inlet]

Thanks for the response, good to know. Maybe I should flush the cache again !!

[half dragon]

Oh! I thought he did get gibberish singing but he wanted clear lyric sung automatically without providing any

[polar spade]

In the styles prompts? Well you want a vibe that you like. So it would be whatever you want. Let me try something with the test prompts I used. Hold on. - Im going to try for a "producer tag" followed by an instrumental, but used your lyrics for a damn long tag. Hahha... cross fingers, may be a total fail.

[polar spade]

That would be cool.

[stable belfry]

yeah i tried copying your lyric prompt and still got instrumental only on mine lol

[half dragon]

fyi, I didn't use any feature (Persona, Cover).. just straight up new creation

[half dragon]

Yeah! but I don't think Suno works that way, as best as I can remember! Maybe in the future, they could offer that!

[dense inlet]

My issue is with a Cover of one of my own Suno Songs, that I am trying to perfect.

[half dragon]

Yeah... that's why I want to make sure just bc I didn't have the issue, it doesn't mean the issue doesn't exist 🙂

[dense inlet]

Much appreciated. I will flush cache anyway....👍😁

[stable belfry]

see like it works great when i type lyrics manually but like we said before i need it to fill the gaps with its own gibberish https://suno.com/song/107d4e9d-28f5-4a1a-9ac5-21fa189bda76

[polar spade]

So. You gave lyrics, so that negates the part I thought you were trying, but I ran an instrumental.... with a little trick to get a "producer tag"- like intro..... then did a Replace Section and put your lyrics in place of the whole instrumental area

You would get basically the same by just just running your lyrics and prompts... lol. But it was fun going the long way around too.

https://suno.com/song/b0fc8ed5-7b72-4dcd-899b-676e1a9af465

[stable belfry]

well, it should be the original beat with my vocals, then extending, and then gibberish generated over the next parts. i've tried giving lyrics and no lyrics

[polar spade]

Well, I added "folk" "storyteller" to nudge suno to throw vocals in as well, if you have enough of your own prompts, it won't go folk, but might help get some gibberish. Try the "mmm" or "oh" in the lyrics box.... and start there. See what happens. Give yourself a few gens to get one, and it it sound like he's starting to sing but stops short, run an extension, or extend anyway. It often take 20 seconds to 1 minutes before suno starts singing gibberish, if even.

[polar spade]

Yea. I didn't know your original beat and your vocals? You got a gibberish one with your vocals? Then using Replace might be the way to test. Literally, just replace the whole initial song with your new lyrics... the beat and vocals should stay the same.

The output you heard , I did just that. I took a 2 minute instrumental with just the initial tag. Ran it through Replace, with your lyrics, and got the same beat, music and vocals.

Often, you can get a gibberish song, separate the vocals and instrumental, remaster the instrumental, and then do a cover of the instrumental and add the real lyrics in. Sometimes you get a whole song back with real lyrics. If that's fails, Replace is an option.

[polar spade]

Just saying it's not out of reach. But can be a 5 step process. Lol.

[stable belfry]

yes, check this out. this is from a few months ago using the extend from my 13 sec "original snippet" from my original comment last night.

as you can hear, it slightly starts with "my vocal", then is singing on its own the rest of the time!!!

https://suno.com/song/93b7b089-c96b-41ac-b21c-f06b680a0731

this whole playlist has my favorite ones that were generated in order to give me directions for the song. first one is the original uploaded snipped
https://suno.com/playlist/d7fe8899-7825-4abc-aab8-e0a19f808daf

[polar spade]

Often, you can get a gibberish song, separate the vocals and instrumental, remaster the instrumental, and then do a cover of the instrumental and add the real lyrics in. Sometimes you get a whole song back with real lyrics. If that's fails, Replace is an option.

I can't cover your song, so you'll have to try on your own with these two tricks.

[half dragon]

How often did it happen? maybe it was still gibberish but a bit clearer 🙂 maybe it's still luck.. like any other thing with AI

[stable belfry]

a few months ago it was 100% of my generations

[half dragon]

did you get it on every generations at the time?

I see it was v3.5..

[stable belfry]

yes every generation had new lyrics, the ones in the playlist are just my favorite ones, but they were all generating new lyrics

[half dragon]

did you try again with v3.5? or just with v4.0 it doesn't do that anymore

[stable belfry]

even extending the original using v 3.5 only generates instrumental now. you basically HAVE to write your own lyrics to get anything. so its very clear they changed even the v 3.5. i think they should revert it otherwise whats the point of being able to choose it.

[polar spade]

It may be a placebo, but I feel.once I find a set of prompts that get hallucinations, that set tends to do better than a new set. So I'm often pulling up and reusing prompts from a gibberish song.

You the same?

[stable belfry]

ill be honest man, i never used prompts , i dont know how. the only one i used a few months ago was "trap" as the style.

is there a way to have them fix this or add a "hallucinate lyrics" option?

[pastel moat]

Hi

[robust mauve]

can someone suggest me what to write in the negative prompt ection to aovoid the Noise in the songs

[timber stream]

Hey everyone.
Just a question, probably already quite known and asked a couple of times:
Is it known that with song generations, the further the song goes, the quality degrades quite noticeably?
We know about the shimmering (which has been reduced all of a sudden honestly lately), but I have a generated song here which goes more into the "powerful orchestral" where the vocals in the first 1,5 Minutes or something are quite clean, but the further the song goes, it more and more sounds like some nasty, heavy audio compression kicks in, with additional smaller artefacts.

Thats with the current V4 Beta tho. Just wondering if thats something common atm

EDIT: That also doesnt really get better with remasters or covers tbh. Granted, the current version I refer to here is a "cover" with the very same setup to see if it does anything to improve, which it does, but not consistently. Remasters were quite a bit more terrible tbh.,

https://suno.com/song/8603948a-4ac7-44bc-a212-5bf926f0ff26
Song I am referrring to.

[fair adder]

hello, i'm not sure to understand how work spaces really work. are they just another version of the playlists? cause i tried to create one, but nothing changes once i'm back into the library

[hollow hare]

Basically a playlist in the Create tab. I mentioned how to access this at #💬┃general-chat message

If you need any further help with it, don't be afraid to reply/ping me:)

[static stone]

Move to Workspace still Not working

Why everytime i make a extend, Suno Change to v 3.5? I lost many credits.

[random reef]

Can anyone help to share steps to download video from suno with only lyrics and not the background image

[opaque canopy]

is it me or every chorus kind of sound the same you can always tell .... oh god suno did it again chorus sounds way too similar even after thousand of generation.....

[regal bolt]

Hi anyone know if its possible in "Recreate section" to delete lyrics part and have just the beat ? any prompt can do dat? like Silence or Vocal cut? dt know...

[olive hornet]

Hi Guys. Im having trouble with Suno not finishing the track. its stopping at like 3:21. but the lyrics have not finished.. any idea?

[fallow musk]

is there any walk around to avoid the difference in quality between verse 1 and 2 ?! its too annoying really

[celest heron]

Can try doing a replace, but yeah you will most likely have to do multiple to find one that sounds the same.

[fallow musk]

is there any kind of fix coming soon ? the replace is not matching too

[celest heron]

Man, I hope so. I'm working on a similar issue right now. I've already created close to 20 trying to find one that matches. Frustrating, but it will get better with time I'm sure.

[fallow musk]

that time cost credit and money mate so its not the best thing to do !

[celest heron]

I get it! I'm annoyed I'm wasting so many credits for sure lol. But I'm kinda OCD haha

I know what it's supposed to sound like, and I won't be happy until I get it right

[fallow musk]

its not just the case, when u create something u need to see it as u want, by time you will lose your passion

[celest heron]

Not sure I understand what you're saying

[fallow musk]

by time people will lose the passion to use suno, because its not giving the usual service as it was, so its a kind of emergency issue to be fixed as its still exist since v4 released.

[neat kelp]

im trying to create instrumental music using instrumental toggled on. why doesnt the AI know when to stop? all song ends in 3:59, and doesn't stop properly

[celest heron]

Yeah I feel you. A lot of it is pretty buggy, pretty annoying. Like right now, when I do a replace, it's not asking what section I want the new replace to replace. So it's doing all kinds of weird shit.

[celest heron]

Yeah, I wish there was a way to specify how long you want the song to be. Or even a range would be nice.

[neat kelp]

not only that, there are many other things that sometimes upset me because it doesn't work right away

well im a free user anyway so i cant complain much, but if i were to pay money, i certainly want the service to be good (if not perfect)

[celest heron]

Ha, you and I both!

[celest heron]

I'm paying, so it makes it more frustrating for sure. I feel like I waste a lot of credits because things are buggy. Just like I wrote above, the replace feature is shit right now, amongst other things.

[olive hornet]

I would love the stems to be functional and not a rip stem. even if its just music and vocal like it is. If you can "Recreate" why not have an option to do "Recreate" without vocals lets say. do you think this would be pos in the future?

[celest heron]

That certainly seems possible! Would sure be a nifty feature to have

[plain fable]

How to enable “repeat song” on ios

??

[ancient flower]

Yo guys, I paid two days ago, and I have like 2500 credits, but I still can’t generate anything. Anyone else having the same problem?

This is the error message:

Need at least 10 credits to create.
Upgrade your account to get more credits!

[stable belfry]

no but i am stuck on uploading

[hybrid jolt]

Hi all,
Normally, we are supposed to be able to create songs up to 4 minutes maximum. But in most cases, we rarely exceed 3:30. Today, it's worse, I almost never exceed 3:08. All my attempts result in songs being cut off prematurely, sometimes 1 or 2 sentences from the end. Is there a way to fix this?

[naive quail]

A lot of people have been reporting this lately. I don't think there's anything that we can do other than wait. You should report it one of the Feedback channels where the Devs have a better chance of seeing it.

[elfin isle]

This is the most requested feature. The trouble is they’d need to retrain an entire model on individual multi tracks and then also have it output as multitracks… Not sure we’ll get it anytime soon but would be amazing.

[honest edge]

If someone wants to publish their album or their songs on the different platforms, I leave you a link with a 7% discount on the service in case it can help you. https://distrokid.com/vip/seven/8563836

[delicate bloom]

I've had to use audacity to get clean stems

[unique stone]

Are you using cover or persona?
If so these currently cause songs to be roughly around that length.
You can continue them by using Extend.

[delicate bloom]

question: what is the best method to use to change one word or 2 or a line or 2 of lyrics. Replace never seems to work for me

[olive hornet]

I have no idea how it does what it does already..hah but you can get it to not sing parts a bit in the "replace" function. So I think just being able to produce just the music (not individual tracks) without the vocal should be possible yes?

[delicate bloom]

you can do the instrumental stem in suno however it always leaves a bit of vocal

[olive hornet]

Its a "rip" stem is why.

[delicate bloom]

ok.. that's why I use audacity with openvino add on

[olive hornet]

We are asking g if it can create the same track but with out vocal. So it's a clean music stem.

[delicate bloom]

but then maybe that's a rip too?

I would love to have that .. I 've wanted that all along.

so it creates it not simply split?

[olive hornet]

Correct

[delicate bloom]

that would be so great

[olive hornet]

I understand that doing separate stems for the music would be hard. But im sure being able to just have the vocal "not play" could be pos

[delicate bloom]

thats why I"ve ad to use audacity because it cleanly splits into multiple stems/tracks

[olive hornet]

I ha e lots of singers who I could use if it's possible..lol

[hybrid jolt]

Yes, I usually use a persona.
I just did some tests without a persona and I managed to get to 3:59 🙂
However, I try to use the extend function as little as possible because it degrades the sound quality, unlike the section replacement function, but there are times when I have no choice but to clean the audio with Adobe Audition.

[olive hornet]

I use ripx. Not tried audacity

[delicate bloom]

How do you all change a word or phrase in lyrics?

replace does not seem to work for me

[hybrid jolt]

use the Replace Section function

[olive hornet]

I use persona and change the words that way

Some times works.hah

[stable belfry]

is it possible to talk to a dev or get feedback to them directly i really need the hallucinated lyrics back

[hybrid jolt]

Select the section in which you want to change the lyrics (at least 10 seconds). Select even one or two seconds before and one or two seconds after, and enter only all the lyrics of the section to be changed with the changes

[delicate bloom]

thanks

[hybrid jolt]

for exemple, if the original lyrics are
"The hills sleep beneath the blackened sky,
Each winding turn holds a hope nearby.
The scars, they sing, I hear their sound,
They teach me to live, with every round."
just enter
"The hills rest beneath the midnight sky,
Each twist and turn holds a dream close by.
The scars, they hum, I feel their tone,
They show me to live, I'm not alone."

[radiant chasm]

Heyas, I'm trying to upload some audio about 1:05 and its taking an incredibly long time. How long should it take? What could be wrong?

[polar spade]

It's hard to say use Replace, when you say it doesn't work for you. But Replace, "working" for me, is the cleanest means of replacing words or phrases.

Much of my mistakes is not going back and viewing precious steps whenever I make a change in the timestamp. That will reset the lyrics without our noticing.

And if the sample output isn't the same time metering as the chosen segment, then don't use it and try again on another. Or start over and check all the steps. You still are charged credits each time you run a gen, though, if that's what you call "not working" . But I tweak around with it, get a segment that fits and I'm done -- between 1 and, I'd say, 8 tries on most. In the end, quicker and cheaper than other options.

Maybe a process should be shown in workaround by someone. I have not gotten around to it as it can look tedious to do the ""check and double-check before clicking" routine.

[polar spade]

Uh oh.

But did you click on "permanently delete" button? That's the death of them.

[stable belfry]

dude i just want the old suno back and i dont have time to go build a competitor right now

[hybrid jolt]

"if the sample output isn't the same time metering as the chosen segment, then don't use it and try again on another"
Yes, to change part of lyrics in a verse, chorus or other sections, you'd better keep active the button "Make same length as selection".

[polar spade]

Oh you didn't lose a file, you just mean to get the hallucinations to appear as they "used to". I got it. Well there are the suggestion feedback channels for just that. As any feedback, the more feedback about the same thing, the more they'll look into it.

They are, after all, hallucinations, so that's a random thing as artifacts of trying to arrange and create the real parts. I hope they don't go away either, but I've seen an improvement in them for my uses. So, I don't know what to say.

[gilded falcon]

Does Suno not have a way to list the workspace in. different orders? Like by name accending?

[radiant chasm]

My upload is still "going"... How long should I expect it to take? What could be the cause of it failing?

[polar spade]

Still no sort feature as far as I can see. We still have to remember title names to search for them. That's all I can do. If you haven't noticed, there is a filter menu, so that partly helps.

[stable belfry]

if you upload copyright music it freezes, i just tried this lol

[radiant chasm]

I made the music myself

[stable belfry]

i think i found a similar workaround, just use COVER feature, dont know how it compares to few months ago yet still experiementing

[unique stone]

Don't do this. It's against Terms of Service to upload audio you don't own the rights to.

[stable belfry]

ah, ok

[gilded falcon]

Okay. So far this workspace feature has been very useful and I am happy for this update, as I have over 2000 generations that would of been a nightmare to organize!

However this still means theres going to be plenty of workspaces. I think a list arranger and a "download all in workspace" would do absolute wonders!

[delicate bloom]

all songs stay in the library correct?

[gilded falcon]

yeah

[polar spade]

Of course, I love folders, so I'd want a workspace called "holiday songs, then a folder under that called "easter" and another called "christmas" . And so on... otherwise, as nice as Workspace is, thers going to be lots and not sortable, just searchable. So files will be doubled up the same in holiday as in easter, for instance, unless I just lump them in one and be done with it. Gift horse , I guess. Can't complain.

[polar spade]

Between you and me, so don't tell anyone, I'm sure there are still protocols for trying to assure the nefarious bloke won't use a mass-group-download feature as an easier means of stealing songs. It's a balance of giving honest members nice features and keeping securities in place.

Already since day one, they've done a number of back-pedals to limit what was once available because some folks can't play nice.

[polar spade]

Yes. It appears the Library is still a mass way station for everything not deleted.

[gilded falcon]

really? What back pedals?

[polar spade]

Nope. Enjoy what we get. There are those that lurk. 😇

[gilded falcon]

okay

as regards folders I'm not sure how that would work

[polar spade]

I would say similar to what you find in a File cabinet or folders section of your computer. Just one subdirectory under another, cascading from a general category to specific.

[delicate bloom]

yay it worked

[polar spade]

I'd want to separate songs by bpm even. Or mood. Putting that in the lyric description is about it. But no worries, if a completed song (and a few others ) , then it's downloaded and sorted at home.

[polar spade]

Yay! Whatever it was. A positive is always good.

[gilded falcon]

I think BPMs are an ignored aspect, as it seems you can't specify the BPM when generating.
I guess folders could be similar to dropbox. I think it could work how playlists work but you can have a playlist in a playlist.

Speaking of playlists... WHY NO SHUFFLE BUTTON?

[polar spade]

You can always suggest it. https://discord.com/channels/1069381916492562582/1260263379243761776 One button click: one time/repeat/shuffle/continue

[gilded falcon]

good shout

[dusky heart]

Hi! The new feature is powerful with the the ability to organize and prioritize the work by creating folders. I was wondering how to organize everything, and now it’s possible! That’s amazing—thank you!

[gilded falcon]

Although Threads would be incredebly difficult, it seems the suno can catch different melodies I have noticed. Would be cool if there was a feature that could try and recreat or interpret (similar to the remaster feature) some of the instruments and play those alone.

[stable belfry]

oook , so NOW the question @polar spade , how do we reverse our way back to mixing the vocals?

say i get the god tier suno generation with a voice that sounds similar to mine. how do i get the vocal effects from the AI voice to my own raw vocal?? 👀

[naive quail]

Suno doesn't currently have any functions/features that'd do that. You can export the wav, stem it out and replace the vocal stem with your own singing in a DAW, though.

[polar spade]

I think you are asking what can only be done now on third party AI programs (sites). Not in Suno.

[stable belfry]

yes do you know the process?

[polar spade]

Not without researching sites. I have only watched different reviews on YouTube. Very interesting. But I've not worked with it much. Other members would know more.

[inner sedge]

I waste a LOT of credits trying to make a duet. Either it doesn't do duet, or it mixes the voices Male = Female, female = male. Thisvis ridiculous.

[naive quail]

The most simple way would be to use "Get Stems" to split the track into a vocal track and an instrumental track. Download the instrumental track as a .wav file. Import the instrumental track into a DAW and then record your own vocals on top of it. Done!

The Suno stem splitter is not particularly good, tho... so you may be better off using an external splitter (others can chime in on the ones they prefer).

[stable belfry]

rigt but the one problem is i dont know how to mix the vocals, so better solution would be get the vocal stem, and my matching raw performance, then use an ai tool to apply the vocal effects to my take

i guess it would be some sorta blending action

[pine wharf]

If you become a helper+ do you get free acess to premium suno?

[polar spade]

What vocal effects do you mean. Just pitch correction and maybe reverb? Or some changing of your voice through AI to give a sweeter or warmer voice?

[naive quail]

That sounds like a can of worms. Some of the DAWs probably have built-in AI mixing and mastering.

[stable belfry]

@polar spade so it wouldnt necessarily be a normal mix and master, it needs to use the AI stem as a TARGET to change my vocals to, such that it sounds like my voice but with the AI stem's vocal chain

[polar spade]

Have you used audacity? You could start there and play with some of their basic effects for a free program.

[stable belfry]

no no i have Logic Pro

im trying to avoid having to figure out all the settings and all the plugins i would need to use

i basically need to use another AI tool to morph my voice similar to the AI stem as a target

but it should still sound as if i did the performance, you know?

[timid wasp]

Well, even Audacity has AI plugins to heal and master tracks

[polar spade]

Again. I'm ignorant. So it sound like we're back to just becoming familiar with both presets in Pro Logic and the different AI voice augmenting platforms presently out there.

[polar spade]

Well, yea, but to back up a bit, you may get enough from just pitch correcting and Pro Logic to get the same result.

But there are sliders in AI programs to choose how much weight to choose more your voice or more AI enhanced. So sure, go that route if you wish. I'm just drawing a blank as to which I've played around with, since I've last looked as they came out last year. So I'm at a loss without researching.

[stable belfry]

yeah basic pitch correction wouldnt be enough. i have SOME mix experience to know that i need ai to get a good sound haha

[polar spade]

I need new chromosomes and voice lessons before I could get a good sound.

[stable belfry]

https://app.kits.ai/voices/trained i think this may work but im too tired to deal with it tonight so ill check another day

[polar spade]

Turning this back to Suno....For what's it worth, uploading your voice and a demo/throwaway song, using cover and remaster to possibly clean it up, and if AI quality, then make a persona to at least do lead vocals on new songs.

Otherwise, yea... a good dozen or so AI sites to play with. The good ones likely to cost money. 🥲

[stable belfry]

hmmmmm interesting, sounds like it would work, although id argue i should use the persona on the same song because of the way i might be singing in that particular song might be only good for that instrumental you know?

[polar spade]

OK. It's a "...yes, but...." world out there. Lots of choices. Pick your poison.

Have a nice night.

[cold zenith]

I don't understand why the extends last 44 seconds. Instead of maximum 2min, there's 2400 characters...

Since yesterday, I got that problem

Nobody to say :" there's a bug or not "...

[hidden surge]

Hi, I'm trying to get my student account validated but I'm not getting a response from SUNO, where can I go?
Thank you

[polar spade]

As you know, anytime a generation gets stuck on the same time meter length as a few of the other recent ones, then something likely needs to be adjusted in the lyrics or prompts.

As one mentioned to you yesterday, as you approach the lyric limits, it's possible Suno has a harder time at fitting things in and gives up, so to speak.

As other mentioned, it could have to do with the point of extension and try 2, 5 or 10 seconds (and choice of a couple of seconds) before the ending or your previous one.

What have been the results when trying these two tests?

[hollow hare]

Just give it time for Suno to respond. They have a high volume of email (like an additional week worth over the winter holidays). From what I heard, sending subsequent email will reset your position in queue.

You can try to DM our community manager @steady kelp . He has his own queue line but he might be able to help you?

[dim spoke]

Tell me how to forbid the network to use the piano and all keyboards in general. just got fed up with the piano playing everywhere.

[hidden surge]

Thanks!

[severe dew]

Try out 'exclude style' (negative prompt). If it doesn't work entirely be specific in the style box (positive prompt), as to what the instrument focus should be. Might still take a few tries for the AI to understand.

[gritty badger]

While trying to salvage my V4 songs with great arrangements but plagued by shimmers and degradations by covering them with V3.5, I had a thought—maybe we could have a feature for a V3.5 "Remaster" (or maybe "Demaster" 😅). It’s honestly heartbreaking to see so many great songs full of potential just sitting there, wasted because of these issues. Could we at least get a feature like that as a workaround until V4 improves?

[tidal sequoia]

Ran into an interesting (for me) problem this morning. Needed to do a cover of song to "fix" some issues. Of course, the cover ended up being shorter than original (seems to always happen). So then needed to extend the cover to finish out the song lyrics. Multiple tries and what I get each time is some type of background noise being added. Hard to explain but seems like some type of ambient noise like dishes rattling or something. Once starts, it increases in volume until end of song; even sometimes past the song when the extend tacks on filler at the end. I had noticed this toward the end of an earlier version of song, but I had already rejected that version due to other issues so didn't bother with it and assumed it was just a glitch. But now it is adding this background noise on every extend for this song version. I did another extend on different version of same song without this problem coming up. Any ideas?

[glossy zephyr]

How can 2 words (You're Dead) get flagged as copyright in lyrics for a song, but this entire account of Taylor Swift songs is sitting here getting likes and clicks?

https://suno.com/@saylorthrift

Not 'technical discussion' really, but definitely wild that it's this blatant on copyright thieving, that is able to happen here.

[dusky heart]

The sound of songs is getting worse. A veil in the background sometimes noisy and loud.

[half dragon]

Taylor Swift? really? why... anyways, you do you.

[glossy zephyr]

grrrrross eh? The things you find on reddit, makes me realize why I tend to stay off there, lol

[half dragon]

Yeah.

[tidal sequoia]

Since I don't listen to Taylor Swift, I have to ask. Is this Suno remakes of her songs? Or is it someone doing their own songs in her style? I see that they appear to be using actual photos of her on the songs (which in itself would appear to be a copyright violation); but I don't know enough of her music to know the answer to my own question.

[glossy zephyr]

I had to look up lyrics, lol, I don't know her music at all, but could recognize that it's lyrics sounded familiar from a tune. All I can say, is it is blatant, copyright use all over, from lyrics, to the pics used.

The more I checked the more it was obvious copyright issues

[tidal sequoia]

I hate when Suno allows something like this. Keep it up; and it is going to get the whole site shut down. This is why we can't have nice things . . .

[naive quail]

They're definitely using her copyrighted lyrics.

[quasi summit]

Where are the upgrades we were promised last year?

Asking for a friend.

[tidal sequoia]

When I see something like this, it always makes me wonder if it could be a competitor doing it or even the music industry itself trying to get the site shut down. Conspiracy theories? But that is no excuse for Suno not taking action when this happens.

[opaque canopy]

ok so lol i found the trash icon.... click on it and realise that all the song i was sure it was deleted are there and now i have 138 pages of stuff to delete one by one permenantly come on LOL

yes i can shift and hold multiple bla bla bla... my main request is when I deleted a song idont want to have to go to the trash so i need to delete again ??? i guess its for people who make mistake

[torn spoke]

puts "solo" in tags
Promptly gets a duette that switches every fkin line

[marsh prawn]

Hello!
I’ve been trying to create meditative music with Suno by specifying a tempo of 65 BPM in my prompts. However, many compositions are generated with tempos ranging from 90 to 150 BPM.

Here’s an example of my prompt:
"Generate a meditative piece with a slow piano melody and cello accompaniment, ideal for mindfulness. Use a tempo of 65 BPM."

Can anyone advise why this happens and how to achieve a steady tempo of 65 BPM?

Thank you for your help!

[severe dew]

Not sure it helps since I don't do instrumentals much, but maybe short the sentences. Less prose, more tags or full stops - ex: meditative. slow paced. 65 bpm. piano focus. piano lead. mindfulness or use [ betweeen central themes such as genre, pace and the like.

[west yew]

So I'm getting flagged for "prompt contained inappropriate material." I'm describing a one-on-one battle, but I'm using words I've used before without incident. Did somebody just decide Suno needed to be PG?

Apparently death is a problem now for Suno?

Nope. Nevermind. Apparently a character name in the story is offensive to some. Huh.

[naive quail]

but why would you name a character "Lite Beer"?

[spark cargo]

Hi there, anyone know off hand what the optimum px size is for the song-image to display correctly in thumbnail view?

[west yew]

Apparently "dink" is offensive.

I thought it was Double Income, No Kids.

I didn't make it up -- it's a character from a book.

[tidal sequoia]

Try it as "deenk" and see if it gets through.

[neat nexus]

I don't know that you can ever really get that specific in a prompt. Can you try testing metronome 2/4 in your prompt to see if it helps?

[dusky snow]

Are there any plans to make it possible to do covers longer than 3:30? Or create covers with a persona? I have a 5-minute song I’m trying to get to sound better, as the audio gets muddier and more compressed as time goes on, but when I remaster it it maintains the degrading audio quality as you get closer to the end of the song. Doing a cover works well to fix this, but it only processes 3:30 of the song. So then I thought I could work around this problem by covering the song in two sections. But without being able to use a persona, the resulting songs sound completely different with no way of keeping any sort of consistency between the two. I’m kinda stuck.

[pliant idol]

Try this; Slow, piano concerto with cello, melodic, sad, slow and peaceful, piano, slow and soft, piano forte

The model, used like this with repetition and over lapping constructs should get you pretty close

[surreal dew]

Is anyone else having problems purchasing extra credits? I recently subscribed through my android phone, and when I try the purchase credits on the site, it just loads forever when I try to make the purchase and nothing happens.

[proper spindle]

I noticed that too, I'm glad we now also get 50 credits a day but buying more would be even better. I tried the other day to get more credits and had no luck as well.

[low stream]

I hate when I am trying to edit just 12 seconds of a song to get the wording pronounce right and suno is like great here's a 1:30 replacement, no matter what I do for this one song it can't seem to get it right lol

anyone have a good resolution for this?

[jagged sand]

im curious how do you watermark the music

[hollow hare]

Watermark or copyright?

[tidal sequoia]

I have in the past done a cover and then done an extend on the cover to get in the full song. (As indicated in earlier post today, I am experiencing a problem with an extend on a particular song; but it usually works.) Also, on very long songs, I have done an extend on a song, and then taken the whole song from that extend and extended it. I have one song that is something like seven and a half minutes, so it took multiple extends to finally get it all in since I would lose some of the end of the song at the next extend.

[shy spear]

Okay, so for some reason, I'm getting this new error where the first extension of a track ends at 53-53 seconds of generation, and entirely ignores the rest of the song. It ends way too early, no matter what I do. I'm in v4, if that helps.

[tidal sequoia]

Conspiracy theory: At various times in the past, I have had songs generate and terminate early even though there is time left on the limit. I have also had extends terminate early. And it is clear that covers are made to be shorter and not capture the whole song being covered. I don't know how many times that I have had songs terminate only a few words before the end of the lyrics. The "conspiracy": I think it is either the AI program being programmed (or the AI program on its own?) forcing you to use extra credits. However, over all, this appears to only happen sporadically instead of being constant. Though if I were programming it to do that, I would make it sporadic to throw off suspicion. Again, purely a conspiracy theory - because when something does not work the way I want it to, then it must be a conspiracy.

[shy spear]

I in no way indicated it was a cover lol This is an original song that I wrote.

And it's never done this to me before.

[tidal sequoia]

Cover for suno has a different meaning than cover in the outside world. Outside world, a cover is making a new recording of somebody else's work previously recorded. In suno, cover is making a new "cover" version of a song that you have already generated through suno. I think "cover" is still in its beta version, but I have founded it very useful at times to clean up some "errors" in some of my original songs.

[shy spear]

Its also a new generation

[leaden yarrow]

Hey friends! dunno who the admins are, but would love to see some greater capacity to direct structure with instrumental pieces as one can with lyrical input. IE I've noticed I can use division tags in lyrics to guide the overall structure of the output. (8 bar Intro, 16 bar Verse, PreCh, Ch, V, Ch, Middle 8/breakdown, Ch, Ch, Outro...) ...but you don't really have that input source if working in an instrumental capacity. It would be great to do something akin to toggling between lyrics and optional structure tags in that input string, depending on which mode you're interfacing with. Is this a feature request yet? PS, great tool!

[fierce phoenix]

"replace section" doesn't add lyrics, and I just bought premium. do I just have to do extensions?

[west yew]

Replace is very buggy right now. It often can do changed lyrics, but it has to be in the right mood.

[shy spear]

Please, does ANYONE have an answer as to why Suno is doing this now? I've never come across this issue before, and it's eating up my credits trying to get past it

[polar spade]

Suno doesn't have an "Add Beta" tool. Just "Replace Beta". You can Replace 10 words with 20 words if that's the goal, but suno decides if it wants to extend it out that much and if so, be prepared to accept whatever it give you, then start at that point and extend off those new words. IOW, likely can't fit 20 words within a 10 word segment and have it smoothly blend back into the rest of the song. Not yet, anyway.

[tidal sequoia]

"Replace" is to replace some lyrics, such as changing certain words in verse or chorus, or swapping one verse for another. That is what it is for. "Extend" is to add something at the end of the song from a certain point of time in the song generation; so it wipes out everything from the point of time to the end and replaces it with something else or, if just extending from the end of the song (rare) then just tacks on what you indicate. However, "replace" can either be keyed to replacing for the same period of time in the song or letting it take the time it needs (there's a toggle for that). So you might try "replace" something (has to be 10 second to 30 seconds in original) while also adding an additional verse at that point and see how it works out. In theory, it would "replace" at the indicated time with whatever you indicate and then pick up with the remainder of the song from the end of your indicated replacement . But I have never tried it that way, so who knows.

[fierce phoenix]

oh shit, I can't add lyrics to instrumentals, because there is no vocal track?

I need some coffee

[shy spear]

Thats what instrumental means lol

[tidal sequoia]

The only answer I can give you is to do multiple extensions, which I have had to do in the past. In other words, extend for the allowed 53 seconds. Then use that to create the "whole song" and then extend that "whole song" by another 53 seconds or whatever, until you end up with the complete song that you want. Yes, that uses up more credits, but not as many as telling it to do the same thing over and over again and just hoping this time it will do something different. You can't argue with AI when it gets in one of these fits. You just have to try to find a workaround or give up, come back another day, and hope that something has changed.

[gentle verge]

Hi, I just bought 1 year of the SUNO version, but I had 415 credits still available for 1 month that I paid, but when I bought the year they took away those 415 credits that I had to give me the 2500 for the month
It had to give me the credits for the month + the credits I already had, how can I get them?

@untold briar

[hollow hare]

Please contact [e-mail]- they should be able to credit you for the missing amount if this is the first time this happened to you.

Be aware that credits refreshes/renews monthly for subscribers - the credits do not carry over

Well, there are some exceptions:

[flint haven]

🎵 Credit System Overview 🎵

🔄 Free Plan Users:

• Daily 50 credits refill at 00:00 UTC+0 (<t:1713222000:t> standard local time).

💳 Subscribers:

• Credits refresh monthly at the time of subscription (2,500/10,000, depending on the plan).
• Once subscription credits are exhausted, subscribers switch to daily 50 free credits at the same refresh time as free plan users.

📈 Bonus Credits:

• +5 credits for generations under 8 seconds or failed generations.
• Daily credit refresh will roll your balance down to 50 credits if your total exceeds 55 credits.

⏳ Carry-Over Credits (Non-Refreshing):

• Gifted SUNO credits 🪙
• Top-up credits (require an active subscription to use).

🛠 Credit Usage Priority:

• Subscribed/free credits are used first, followed by carry-over credits.

[gentle verge]

So if I don't spend the credits I have per month, they are renewed and they don't give me the ones I didn't spend? I mean, if the year I pay ends and I don't spend all my credits, will they be lost? Or if I don't make a song in a month, have I already lost all the credits for that month?

[hollow hare]

You will lose all credits for the month when the monthly renewal happens. (Excluding gifted and topped-up credits)

The yearly plan is a 20%-off monthly plan

[gentle verge]

Omg, so should I use yes or yes all the credits per month?

[hollow hare]

It is highly recommend to use up all credits a few days before your renewal.

One method some employ is generating a bunch of instrumental intro for various genres. If you hear something you love, you can upvote and/or move it to a new workspace to use for future extensions

Basically it is a use it or lose it situation with subscription based credits.

[gentle verge]

aff, I didn't quite understand the suggestion of moving the instrumentals, I'm new to this, but I will take into account the fact that I must consume all the credits per month, I thank you very much, I had asked in the chat in Spanish and they told me that They accumulated, that's why I bought the full year, I wish there was this type of service on the Spanish channels 🙂 Thank you very much, I already sent an email to SUNO for my credits.

[hollow hare]

The support staff are understanding the first time it happens and will gift the missing credits.

I am not sure about subsequent losses since you would be aware about it at that point

[hollow hare]

Oh, that is not good. If you can link me to thet post (right click the message > copy message link), I can talk to the user who said that and correct them

[gentle verge]

Ok, but please don't scold him, it wasn't his fault, he was just trying to help. #1154779774858833940 message

Although yes, that answer was the reason why I decided to buy the full year xD

[gray summit]

To the Suno devs: You are awesome. Just read about the workspace feature. YEEEEAAAAH. NICEEE

[hollow hare]

Hmm, alright, I will address it later in a kinder way

[gentle verge]

Taking advantage of your kindness, where can I find any guidance on how exactly to use the lyric area of ​​my song? That is, I have seen that SUNO accepts certain "commands" within the lyrics of the song, such as if you want a certain part to be sung faster, or things like that, is there a guide to all the "commands" that SUNO accepts? ?

Oh and sorry if I say something wrong, I use a translator 🙂

[plain berry]

is anyone finding that v3.5 is a lot better for drum and bass than v4? it's worse at following the lyrics and it has worse quality, but it's better for the tags and general "vibe capturing"

[untold briar]

Try making a Persona out of a good drum and bass clip, and using it in V4.

[shy spear]

I can answer that, sort of. I use it frequently in my songs.

Generally, you describe what you want, and put it in brackets.
i.e.:

[Intro: Lively, playful jazz piano with a bold, cheeky brass section underneath, giving a feeling of grandeur with a mischievous wink]

[mischievous, playful, charming, smooth male Disney villain, confident, sultry yet menacing, full of theatrical flair and charisma, with just a hint of over-the-top melodrama, slight accent that makes it sound inviting yet dangerous, almost smug, villainous with a hint of fun menace, a villain who loves to ham it up, yet never quite intends real harm]

It really helps with bigger projects. For example, I'm going between these distractions, and creating a full grand opera, within Suno. So it helps if you want to do a duet, for example

[gentle verge]

Uff, I've tried, but if I put a text that long it takes it as lyrics of the song and sings the instruction xD

[shy spear]

well it doesnt HAVE to be that long lol

Its how I got things like this:
https://suno.com/song/440c47d6-e7db-4c0b-9a02-42022e775662
and
https://suno.com/song/a8fa17f5-4a09-43c0-bb3b-a09014d9488d

oh, didnt think itd post the playable in this channel lol

I'd say use my personas as an example, I have a BUNCH. But my duets so far are male/male lol

[celest ether]

The workspace thing is nice, but it's not nice being taken to the workspace automatically when you're for instance remastering old songs and maybe going through a hundred of them. Having pressed remastered, I'm taken out of the library to the workspace and now I need to reload the library and go to the last page of what I was looking at. Maybe page 36 or something. Workspace would be nice if it didn't take over whenever I rightclicked and created something based on another song.

P.S. Oh and it's time v4 became the default. Don't know how many credits I've wasted accidentally using 3.5 because it was reverted back to that automatically.

[gentle verge]

Thank you very much

[wet fox]

does anyone here know how to mix a song at all? i recorded vocals for my song, and i cant get them to sound like theyre not on top of the instruments

not ai

[main cairn]

Hi everyone, so when I click an option in the create menu like cover or extend, the menu is disappearing on me for some reason, without pulling up the extend or cover dialogs. Has anyone else experienced this? I can reuse prompt, but anything that has its own dialog just vanishes.

Also, do workspaces work in the suno app?

[tiny leaf]

My account was banned(unable to generate songs, etc.) from the last wave for mistaken my account as a bot.
I sent an email to suno support, and I am able to generate songs again, but the "get whole song" button is still unable to use.
It's been 3 months since I contact the support for this issue, but the "get whole song" button is still not working for my account, I tried switch to an alter account, the get whole song button is functional.
But I can't just switch to my alter account because I can't transfer my works from the old account to the new one(limitation that you can only upload up to 1:00 long song)

[elfin isle]

It’s really hard to get vocals sounding cohesive to an ai generated instrumental. You could try Suno’s remaster tool but that also has limited success. Depending on the style or genre of the instrumental you could try sending both the instrumental and your vocal to a room reverb and mix to taste (this can sometimes help the vocal not sound so on top or infront of the mix)

[wet fox]

sorry. i actually paid a guy to do the guitar part and paid another guy to do the percussion

[elfin isle]

Lots of trial and error. I’m really looking forward to when Suno allows users to make personas of their own voice. I understand certain steps will need to be taken so the feature doesn’t get abused. However I think that will garner much more cohesive results than singing over the top of the instrumentals

[wet fox]

so none of the instruments are ai

[elfin isle]

Oh, well in that case it might be worth hiring a mix engineer?

[wet fox]

yeah, thats why i was hoping to run into one on here lol

[elfin isle]

I mean, I guess that’s my field of expertise but I can’t take on any jobs rn

[wet fox]

but i need to re-record my vocals in a studio

[elfin isle]

If you want the vocal to sound more upfront have it drier

[wet fox]

i did a rough one on my phone

[elfin isle]

If you want to place it back use more verb

Or delay

If you want to glue everything together, try some bus compression on the mix bus

API 2500 is a great bus compressor

Waves version is pretty affordable

[wet fox]

idk what any of those words mean lol

[elfin isle]

That’s what the internet is for 🙂

Plenty of tutorials on YouTube

[wet fox]

true, but tbh i've got no interest in learning the mixing. no offense.

i think i'll pay someone who knows all that stuff for it

i know mixing is very important in music though

[elfin isle]

Fair, the problem sometimes is you can’t explain exactly what the issue is you may never be satisfied w the results other people get

if*

However, I’m sure there’s going to be some crazy ai mixing tools in the near future

Already is a bunch

G2g but good luck!

[wet fox]

yeah, that's something i was worried about before i hired the guys to do the instruments. was very picky with them

thank you for your help!

[restive dock]

why all my song keep tag with "fake" description? The music also not following my description. It come out randomly.

as u can see.. got 8 seconds music..

are suno has been hacked?

[cunning mulch]

that looks like you got busted using a fake app

[restive dock]

lol.. im using chome browser.. not using any app

chrome*

as usual

suno already change domain?

got busted but release 8 seconds music? then take my credit?

or suno just making money out of it?

[cunning mulch]

speaking of mixing tools... I have a song I split in Suno and revamped instrumentals on. Now I need to mix it properly, but this is new territory for me

[restive dock]

hurmm

where is the fake app? just 5 minutes ago, can generate

[violet raven]

seems to be broken.. my extend deletes credits.. yours just accuse you of being fake lol

[restive dock]

yeah.. easy for dev to make excuse but not come troubleshooting..

we are pro user.. using money investing in your business..

[violet raven]

yeah i dont think they will fix anything anytime soon..

[restive dock]

i think that dev is other company, suno is just a promoter that run hundreds of AI music app etc. That they didnt want anyone to email technicals with them, because they dont know how to troubleshoot, otherwise make easy money from time to time..

[violet raven]

more problems, more bad gens makes people invest more..

[unique stone]

You'll need to fill this out if Suno thinks you are using a fake app. The songs below being 2 minutes or less (without being extensions) are another symptom of this.

https://sunomusic.typeform.com/to/zGsJSzsf

There are lots of apps that basically piggyback off Suno. This is meant to be a way of preventing them from doing that. But sometimes genuine users of Suno get affected, can be due to VPN usage or problems with billing details for example.

[restive dock]

no VPN, no other app.. just pure clean CHROME Browser

[slender lantern]

I'm a pro member, but notice when using persona's, songs are almost aways limited to 3:20. Asked around, seems a known issue. Is this something that's being worked on?

[restive dock]

still fake status.. still consume my credit tho..

[celest heron]

Hey guys, quick question. How much extra time do you add on both sides for a replace?

[cunning mulch]

Not really. There's only so much data the model can use, and storing information about the older song takes up some of that room

[acoustic vigil]

Hello. Please help me, how can i generate a vocal ai with E4 maximum high musical note? Because I'm bass baritone and musical notes of AI are too high for me for chest voice.

Thank u.

[celest heron]

Why on earth is replace giving me the full song lengths when I'm trying to replace about 15 seconds? I even have the "Make same length as selection" toggle enabled. Wtf

[tired ore]

im having a error saying i need at least 10 credits to create but i got 10000 credits and i am subscribed

^ im on the premier plan btw

[polar spade]

Though we might indicate a range for suno to randomly sing in, based on prompting certain styles, any means to generate, on command, a specific note at a specific moment in time on an output is likely outside its capabilities or parameters.

[acoustic vigil]

So, for the moment it's impossible

[acoustic vigil]

So, for the moment it's impossible

[wanton nimbus]

Is there a chance to recover a song you just sent to trach by accident? The song was amazing, and I miss clicked...

[tired ore]

go to libary and there is a trash icon

[polar spade]

Usually, it, your plan, has just renewed? And often, your payment is lagging behind from being received and posted to your Suno account.

Reasons are often: issues with your payment in some form.

Solution: 1) Check your bank or card services to check that payment has been debited.
2) check your email for receipt that Suno has posted your payment for this current renewal date.

If both have not happened, you are in limbo and Suno is waiting for payment. Electronic payments go through Stripe payment company, so it can likely take up to 48 hours before payments post, I'll guess.

[wanton nimbus]

And you just saved my day! Thank you very much, God bless!

[pseudo grail]

Hello everyone I hope you can help me I bought the annual subscription last month and today I only have 50 credits and my songs have disappeared as if I had never no subscription do you know why?

[tired ore]

hmm seems to be in limbo since the 13th anyway to fix it like unsub then resub?

[tired ore]

all g my guy

[pseudo grail]

I paid the annual subscription but now it's as if I hadn't paid, it's barely been a month since I bought it

[polar spade]

It's a generative AI, so it deals in concepts, nuances and ranges. You might designate "tenor" and it might reach your range , but again, likely never sing a specific melody or note on command.

So, right. Not at the moment and likely never.

[acoustic vigil]

Thank u for youir response. Can u recommand me a good prompt for my needs, please? I want a waltz modern song to generate for my voice to sing... and I need a rang for E4 maximum... on chest but on head voice, i'm high... C6

[polar spade]

Start by follow step 1, then step 2, would be my suggestion as one member to another.

".....Solution: 1) Check your bank or card services to check that payment has been debited.
2) check your email for receipt that Suno has posted your payment for this current renewal date. ....."

If you need support or have billing issues: contact [e-mail]

I read in here of many people having a delayed payment issue, or simply the payment being rejected for some reason, as that is common with bank systems outside of the USA. If this is your first month paid subscription, again, then it's good to check to make sure your payments are going through properly.

[celest heron]

Is replace working for anyone? I'm trying to replace 10 seconds of my song, and it keeps giving me a segment that is 2:01 in length. And I know how to use replace, I've used it many times successfully prior to this issue that has seemingly lasted about a week or so. Extremely frustrating, especially being a paid member. I am trying to replace the first 10 seconds of the song with something a bit longer to get some instrumentals in the beginning, since currently, the vocals start immediately and that's not what I want.

[polar spade]

I've have found it is never "one" prompt that can get a specific action.

You can start with your prompts you wish for a "modern pop waltz" and then add other prompts of genres which commonly have higher range vocals. But honestly, most genres from opera to death metal to Broadway to country to pop to screamo, ALL have vocalist singing in higher registers. I wouldn't know how to suggest something to fit your need.

It's possible to even create a song, then download to a DAW and adjust the pitch to fit your needs as well.

[polar spade]

It does have the toggle switch to limit output to the same length as the selected past. Have you clicked that toggle to on? With that, you might get a random one that goes over, but for me, that keeps the outputs limited to the same length as what I've selected.

Oh, and you are replacing the whole introduction? So from "0" to 10 seconds in? I don't recall if that can change anything as to it's response. But I've changed words in intros without issues before.

[celest heron]

Thanks for your reply. Yeah, I have indeed tried that. Doesn't seem to matter. I've tried different songs, different size replaces, different places in the song to replace...I feel like I've tried everything to just get one good segment, and I have gotten zero good ones out of probably around 50 segments I've generated over the last week. I don't understand.

I can't be the only one having this issue. Was working fine prior to these issues that started about a week ago.

[polar spade]

Wait. You ARE replacing certain lyrics with other lyrics, or certain words with other words?

Or are you choosing a segment from 0 - 10 seconds in the intro and leaving blank or typing in bracket prompts for changing the instrumental aspects?

[celest heron]

Correct. So for this example that I'm working on now, I am trying to replace the first 0-10 seconds (that currently has words starting at like 0.2 seconds in) with the same words. But I have that toggle "Make same length as selection" off, hoping it would give me some variations with some instrumentals in the beginning.

[polar spade]

I know it doesn't help you, but just saying, I replaced my intro with some new words just last night with no problems. It may be on a case by case issue.

[celest heron]

I'm glad it's working for you! 🙂 It's just strange, I feel like a pro with replace because I've used it so many times before, and now replace is essentially broken for me. I don't get it. I emailed support.

[untold canyon]

I am also unable to replace a section of my song - but the issue is slightly different. For me, it keeps saying "prompt flagged for moderation" even though I am fairly certain I'm not trying to do anything too outrageous 🙄

[celest heron]

Interesting

Last night I did see a couple people also complaining about replace and how it wasn't working for them. Was just curious if more people than just the 3 of us were having issues with it. Maybe just some accounts are affected by this bug

[polar spade]

With the toggle off, I can't predict. Theoretically, from day one, Replace should create something, then perfectly splice the rest of the song on the back end of your segment, whether 10 seconds or 3 minutes.

At any time it's been over the designated segment length, it tends to bleed or "overlap" the remaining original bits.

As for your specific issue though, turning the toggle off will more often give 1 to 2 minutes pieces, yea. And then if they overlap as well, you're basically just creating an extension from point "0".

Not sure of a comparison to what it used to do?

What did it do before that worked for you? It would create, with toggle off, a bit of a longer intro, change the lyrics, and splice nicely back into the song? I think that was the concept, but I've never had it do that on the intro side of things for me.

I've always had to "toggle on" and just choose longer/wider segments and let it just change whatever it wants, changing also the lyrics i chose, and then nicely blend back in , all within the designated segment time.

[untold canyon]

This is how it's going when I try to use replace 🤔

[polar spade]

My time of use was about 3pm EST. So by "last night" , things could have changed. 🤷‍♂️

[celest heron]

I have had a little bleeding in some instances, but if I generated a handful of segments, one would blend perfectly back into the song. I've had many instances where I'd have the toggle on and off and have always eventually gotten what I wanted.

[celest heron]

That's really interesting. Maybe it doesn't like the word, 'hell' lol. That can't be it, because I've used worse words, and the replace has worked fine (prior to my current replace issues). There has to be some bugs going on with this feature.

[untold canyon]

Yeah, same 😅 As I said, I don't think the language here should be the issue. And in the second ss you can see it tries to pull in more lyrics, going over the character limit, for a second before saying "nah" 🤔

[celest heron]

Yup, I am also seeing that issue in your second screenshot

[polar spade]

After it's done, using Remaster can sometimes reblend the glitch spliced points. But that's about it, if it doesn't work. Then I'd need to change direction and just start after the replaced part and put in new extensions. Meh... always a work around.

[celest heron]

Yeah, remaster has fixed some of those glitches for me in the past. Sometimes it doesn't though, just kinda random.

[polar spade]

For what it's worth.... "pop the cap" also means to shoot you in the knees. Who knows if that is getting flagged. 🤔

[celest heron]

That would be ridiculous haha. But hey, who knows.

[untold canyon]

And yet it had no issues creating the song in the first place 😅 Sigh. Guess I will just move it to my "unfinished" workspace and try fixing it later 🙈

[celest heron]

Haha good point

[polar spade]

It's an 80s gangsta type term. Is that still used today?

[untold canyon]

Oh god. Are there any kids present? Someone who speaks the lingo?

[celest heron]

Yeah, still being used today 🙂

[untold canyon]

"Then he popped the cap"
"Ohh for real?"
"No cap"

This convo works on many levels 😂

[celest heron]

Haha!

[naive quail]

Well... my dad used to say that, but he was definitely NOT gangsta.

[polar spade]

"This darned rain. My arthritis in my knees, It always pops my cap when it rains and my knees hurt. " yea... there's that too.

[celest heron]

Lol! I'm tellin' ya, the possibilities are endless.

[desert latch]

ola, Is it possible to change the starus to private for a large amount of songs or request the download of the entire library? I would like to move 13 thousand of my songs, and making them private one by one is a bit tiring,

[naive quail]

Nope. 🙂

[wild spade]

With the workspace update, they fixed also the shimmer? 😅

[novel dove]

Uhm... no. 😅

[polar spade]

I'm too lazy to go to a laptop and check, but can't one highlight a page of public songs and click menu and click off Public. Would that work? 🤔 one day I'll go check.

[naive quail]

Just tried it, it deselects all of the track except the one that you try to toggle - and that's the only one that toggles.

[polar spade]

Meh. Oh well. Thanks for testing. 🩵

.

[polar spade]

I knew others were having problems adding songs to playlists. I've clicked a half-dozen times. Nothing.

Either: 1) it's not working now. 2) every click negates the previous click or 3) I just wait and it will eventually show.

This is the way. 🥷

[fair falcon]

Hey everyone! I have a quick question.

I was looking through the Suno FAQ page regarding song ownership. I understand that I own the rights to my song with the Pro or Premier plans, but does this also apply to the song image? I couldn't find anything regarding this.

[naive quail]

I'd probably start with the ones that are blatant copyright violations, like your Tim McGraw cover... 🙂

[stable talon]

Are the notifications broken again? I noticed some of my tracks have more likes now than a few hours ago, but nothing in the notifications...

[unique stone]

Seems if someone likes multiple songs you only get one notification about that person. Saw someone mention this earlier.

[stable talon]

Might actually be an improvement then, lol. So tired of folks just spamming Like on every friggin' song 🙂

[viral gazelle]

Hi I have a problem I sent email 10 days ago but didn’t get any response
The songs I made ended brutally even with extended still stop without any ending of the song
Somebody knows what to do about it??????

[shy hound]

get stem (separate vocal from instrumental), assemble it in DAW, pitch vocal, and send it back to suno to remaster

[lunar anvil]

Sometimes the songs just end. Are you using an [Outro] to signal to the AI to end the song? Also using something like [Fade Out] at the end helps sometimes if the AI is being stubborn.

[signal olive]

Hi! I have the same issue. I have an outro but even the bridge part was incomplete. I tried to extend it but it is still incomplete. Can somebody help, I want to finish my song please 😭

[neat nexus]

?warn @desert latch. Please do not repost them. Thanks.

[desert latch]

👍

[warped furnace]

What seems to be the problem? You have a song that you need to extend to make it complete because you still have lyrics?

[signal olive]

yes, my song is incomplete it is only until halfway bridge, i have an outro, i tried to extend but even the extension from the bridge to outro was incomplete, how do i go about it please, do i just keep on extending or regenerate the song? im super new in suno sorry

[warped furnace]

No problem. Just pick a good starting point in the incomplete generation. It works best to pick the beginning of a line (beware the time stamp). Then you click extend and fill the lyrics from that point. After that (if the extend works for you), you click on "Get Whole Song" to put the pieces together. If there are still lyrics left ... rinse and repeat. 🙂 (You can send a PM if you have more questions, if you like).

[signal olive]

do i just keep on clicking extend from the next part to the next part, thank you for the help

[warped furnace]

First, put the 1st generation and the extend together with Get Whole Song. Listen to it to check if they blended well together. After that you extend the Whole Song if needed.

[signal olive]

i will try this now thank you again

[rugged wyvern]

I've created a lot of songs lately have have gibberish intros/outros, music that just suddenly ends, etc. I use the report but there is no way to say what the issue was in any detail, or get any feedback. Any other paths to getting support as a paid use besides flagging the song and hoping for the best?

[finite grail]

What's with songs taking a long time to upload if 1min+? I'm trying to upload instrumental ones.

[acoustic vigil]

Hello., Thank u for reply. Just vocal to repitch and send it back to suno to remaster?

[polar spade]

These are idiosyncracies which have been common throughout. They flare up randomly, or at times have a day or few where many are having the same thing occur.

Older members seem to realize the updates and modifications are ever flowing and what follows are the glitches and odd bits until something really shines. 3 steps forward, 1 step back.

As I understand, the thumbsdown is still a request. The more dislikes given will eventually show a specific pattern to be reviewed.

The forms are great too, but I assumed looked at individually and handled globally rather than 1 by 1 all the time. It all depends.

Your issues are what I would call "normal" and with workarounds. Not to be ignored though and down vote as much as you can.

[shy hound]

i would send assembled pitched vocals + instrumental. and upload to suno to remaster it yes.

[acoustic vigil]

I will try. Thanks.

[acoustic vigil]

To put in 2 tracks separately, 1 instrumental and one pitch vocals

[acoustic vigil]

And after to remaster the suno

[shy hound]

No, get 2 separated tracks from suno, and send it a Single assembled file ^^

im sorry for my bad english 😄
You must, from Suno, export your song to two tracks: vocals + Instrumental (stem function).
You assemble it in DAW but Pitch the vocals. Then you export the whole song as a single file, and upload it to suno for remaster

[acoustic vigil]

yes. 2 separated tracks. Vocals and instrumental, and repitch (Change the key to the same of vocals and instrumental) and after send it to Suno and remaster the song repitched

[acoustic vigil]

Thank u, buddy

[ruby flicker]

Hi , a can’t send any commands in a chirp bot ? Ones that worked, now it doesn’t work for a long time

[shy hound]

good luck

[ruby flicker]

Please someone help me 🙏🏻

[naive quail]

/help

[flint haven]

How to use Suno

Hi there!
If you’re ready to dive into creating music, head over to Suno.com.

For more detailed information, check out these resources:

• Making Music
• Billing
• Rights and Ownership
• Accounts
• Troubleshooting
• Mobile App
• Suno Server

If you have any more questions, just ask in #general-suno, where you can always find our community, moderators, and guides to help you with most questions. 😊

[hollow hare]

Chirp bot retired like... more than half a year ago? You can make your songs via the website as linked above (https://suno.com)

[ruby flicker]

Thanks a lot 🙏🏻

[rugged wyvern]

Just feels like there is a lack of closing the loop on reporting things.. A place to give specific feedback on what was wrong would help, but an actual path to back/forth would be nice as a paid user. shrug

[fresh girder]

is there something wrong with personas? i pay for the $10 plan but it says i need to subscribe to use them, def have made many in the past

[polar spade]

I have a bit of empathy. If people were only allowed to email once, things may move faster. 🙂

All said, silence is frustrating if people want an app that works perfectly and don't get it.

Taking another view, perfect apps don't get many updates and become predictable.

60% effective with positive updates when they occur? Could still be a plus.

[rugged wyvern]

I'm still a happy user... just feels like they'd want a note to go with the flag

lol

"other" is 100% of my flags

[polar spade]

There's the case of 1) so many varied type of issues - that they just use a general flag. The TOS list of offensive things are long and broad. And 2) the more specific, the faster i can find loopholes to get around a specific flag.

[fluid estuary]

When uploading a song and creating a cover it automatically goes to 3.5 model and usually I miss this. Could this be updated so it automatically is set to latest version instead?

[austere minnow]

I’m having trouble creating a Beta Persona on Suno. I have a monthly subscription plan, and when I try to use the "Make Persona" feature with a selected song, I get the message: "You cannot make Personas from audio uploads at this time." Additionally, it doesn’t work using the other method either.

Somebody know how I can create a Persona or when this feature will be available?

[opaque canopy]

do we have a metal dedicated chat room i know about the metal showcase but i want to talk technical stuff with Suno metal creator

the persona thing is not good.... what we need is when you like a singer that you want to re-use the voice you go in the Band section and have a selection of your favorite voices .... same fore guitar tone or drum style or ...

[naive quail]

Does the upload have vocals? Or something that might be misinterpreted as vocals? If Suno detects vocals in the upload, it won't allow you to make a persona with it.

[opaque canopy]

deleting stuff in the trash folders dont really seems to work

we need something like delete permently trash item every week or every day not unlimited

[hasty silo]

voice is not stable and changing when I extend the song in v4. anybody have this problem? it's better with v3.5

[opaque canopy]

have 7300 credits left but un useable due to the persistant 'taylor swift girly' CHORUS in every song and style

2700 credit with taylor swift chorus even when there is no chorus in the text or prompt or instruction event if all are verse even try with all [break].... ok sorry it was my daylie rent 😦

[polar spade]

Which genre or sound are you going for? Has a good voice popped up in your other songs which you could make a persona?

You may wish to explore other people's publicly available personas* and find one you like. Use in a new output to sing against whatever genre you choose , then make your own person to use thereafter.

*Try searching on Home Page in the genre you like, find songs in the home page list made in that genre with a persona.

[steep willow]

every sing song I've tried today has had verse repetition issues, skipped verses. Tried different bpm, genres, rewriting verses and lyric length. Nothing helps

[opaque canopy]

a voice should be good for any kind of style but consistant now in death metal song there always a pop chorus that you dont want

persona is not good

if i want a death metal voice in a pop style song there should be no probleme or if i want the most brutal death metal song with a taylor swift chorus if its consistant maybe it could work

[opaque canopy]

BPM never work for me

[steep willow]

i've had this issue before but not 6 tries in a row

[opaque canopy]

im not creating stuff until they fix the girly chorus lol maybe my 7600 token will be flush down the drain

since update takes eternity

[shy spear]

So is there a way to fix an incorrect auto-moderation error for song detail image? It is refusing to allow me to upload an image, saying it didn't pass moderation. There is absolutely nothing adult or inappropriate about it, so I am confused.

[shy spear]

I don't seem to have that issue with my death metal lol
https://suno.com/song/2929c79b-807d-4e6e-b282-d136778fd02e

https://suno.com/song/baec7db8-f689-417d-a230-a79ce805d66d

If you want, I made a Persona with a pretty good DM vocal style.

https://suno.com/persona/df02c91e-ca07-46ce-91fb-750e2c9e8518

[hollow hare]

Pretty self-explanatory: any derivatives from an audio upload cannot be used to create a persona. This is due to avoiding impersonation issues

If you had a pure generative track with a voice you love, you should be able to use it on that. I have three made so far with: a deep male voice, ASMR whisper, and Hatsune Miku v1-ish like voice

[fossil bridge]

it's probably your description or styles. I noticed it tends to default to pop or electronica if it doesn't know what to do

[hollow hare]

^ This (above message). People tend to use their own terms for a metal category which the AI will not understand and go towards the more popular genre(s) it was trained with (which is typically pop).

Look up a music genre hierarchy site or even ask an LLM what subgenre a certain music would fall under (but verify if it exists. LLMs can lie to you with confidence)

I like using RYM for subgenre names and category:

• https://rateyourmusic.com/search?searchterm=metal&searchtype=h
• https://rateyourmusic.com/genres/

[polar spade]

[Note: if you don't click "reply" , I'll never know you responded]

.... keep in mind if you don't know: previous month's credits are forfeited at date of renewal, unless you have extra credits.

[icy badge]

@steady kelp Hi, I’ve sent multiple emails about a billing issue where my subscription was canceled a day after being charged, but I haven’t received any response yet. Could you please help look into this? Thanks!

[steady kelp]

DMing!

[inner moth]

Okay I'm trying to use persona for the first time, but why does the sound quality decrease so much when I extend my song? Both parts of the song are apparently V4...

For example, this is the first part: https://suno.com/song/046ce834-8528-4714-8f00-92b1d1b70649

and this is the whole song: https://suno.com/song/91605a07-5696-4653-b146-b84dbb7ef36a

there's clearly a MAJOR difference in sound quality I'm not even an audiophile and I heard it immediately.

[shy spear]

Sort of figured out the issue. I think. Some of the lines on the image may have been incorrectly read by the ai as 'inappropriate' body parts. I used a different image, same character, no problem.

[spice birch]

How can I delete my trash all songs at one time? Or can I?

[shy spear]

shift+ left click
The the normal way to trash, it'll delete all you've selected

[spice birch]

Sounds great, but it want let me select more then one. I may be doing something wrong

There it goes

Got it.. 🤦🏻‍♂️

[opaque canopy]

its impossible i deleted item in the trash but they re appears everytime

i have 138 pages in my trash i gave up

we need something like auto delete after 1 day or 1 week

[spice birch]

Right

I can't even delete in groups because it always says couldn't delete because one of the is associated with a Personas. So you have click one at a time

[last pewter]

Wait, what? You can make persona from instrumental only?!

[last pewter]

Omfg i never heard so much shimmer in one song 🥲

[hollow hare]

No, it is that you can not make persona from uploaded audios which had vocals detected (due to potential impersonation issues).

Vocals generated from Suno can be Persona'd

[shy spear]

[severe dew]

Yup, that is pretty noticeable unfortunately. Almost as overlapping at points and general just volume degradation. Now, I don't know how you extended exactly, but there is no way, for now, to extend with a chosen persona and you simply have to trust that the extension will match the persona. This example is, however, pretty bad - no offense of course. How did the actual extension sound like before getting the full version? no problem?
Enjoyed the first part by the way 🎉

[brittle mural]

hi, i have a technical question...
is just me or the extensions form songs since the "workspace" update is acting wonky?
i mean. it gives me an 1 minute extension of some stuff that it even repeats verses...is imposible to work like that

[high pulsar]

Hey, I’m having an issue with Suno. Since yesterday, when I’ve been trying to create songs, it keeps saying, “This is not the original app.” So, I tried clearing cookies, but now I can’t log in at all. Has anyone else run into this? Any idea how I can fix it? It’s happening across different browsers too. I’m using Chrome (with AdBlock and uBlock extensions) on a MacBook Pro M1 with the latest macOS. Appreciate any help!

[sage quartz]

hey I'm having an issue with 2 of my v4 songs. they were there the day before yesterday and have suddenly disappeared. I didn't have a chance to download one of them. what happened? what can I do?

[fossil bridge]

Here's a helpful list: https://www.reddit.com/r/SunoAI/comments/1d3vmy1/metal_band_style_prompts_for_suno/

[leaden pawn]

Video regens are still broken? Seems to be ok when I download the video, but if I post to discord the video shows old lyrics.

[half dragon]

I noticed the same when I updated my lyrics and I had to retry a few times to get it re-generated.

[dusky snow]

Wait, are you saying you do a cover, then extend the cover with the original structure/lyrics, and it gives you essentially the same content of the original? Wouldn't it just use the cover as context and give you all-new stuff? I'm trying to maintain differences in the final chorus, ending, etc.

[runic wadi]

When you extend a song and it makes a separate thing that I have to like quickly click on the next part is there a way to combine them so I don't have to do that?

[fossil bridge]

there's no way to merge them that way. You have to do that in a third party app. if it's under 2 minutes.. You can reupload the song afterwards and make a cover it.

[inner moth]

It’s literally driving me nuts.

[severe dew]

I get it! Have you tried making a new extension for the original? And is the output the same? From where on the timeline did the extension begin? It can take a while getting the outcome you wish for.

[naive quail]

In the track menu of the extension, click on Create->Get Whole Song. That will merge the parts together into a new track. @fossil bridge

[quasi edge]

DISABLE ALL WITH SUNO.COM AND DISCORD.COM Since you trust this services and you want to create MUSIC. Everytime you erase cookies this will reset so you manually have to do this which is a security just to make sure you are in the correct websites per se. Hope that helps.

[verbal mortar]

This could have been an amazing song but that damn SHIMMER!!! and the stupid issue with Suno just spazzing out and making noise.
https://suno.com/song/9eac713c-9a5c-46a5-a50c-7d414cccf80c

[quasi edge]

I would like to know what BPM are you using and What are your commands being used to tweak what you want. This has to do a bit about learning how to prompt correctly what you want. Have you done a Mastering in V.4? What is your level of Music Production?

[fossil bridge]

Ah cool. Must be new.

[verbal mortar]

i've been making music for 20yrs off and on. released 15 albums.
I've been using Suno since Oct and it's been working mostly well. last few months it's been all over the place.
Prompt for style is the one i've been using with amazing results since Oct
Technical Death metal, extreme metal, industrial Death Metal, French Death Metal, Jazz

and yeah @quasi edge I did a remaster in V4

sound worse remastered LOL

Im wondering if its too short? that's why it spazzes out.

[quasi edge]

You changed your user name ? Ok. As you well know. SUNO saves you thousands of hours including recording the actual vocals and all that magic steps before you actually start making music. I just began less than a month. What i noticed is that you can record the audio instrumental separate what would be awesome is you could do the vocal also then in editing combining both as normally you would do Legacy workflow. Tweaking the frequencies would a nice option here because it does then to go ROBOTIC once every while. Also, theres the plugins + or - where you want them and i have yet to see this option in here. What i noticed is that you can have a PERSONA which is what i am talking about a TEMPLATE where everything is set up just like you want every setting for a specific genre and so on with every other Genre. If you want to compare my music to what i mean and i just started in Synthwave New Retro Wave. So SUNO for me is AWESOME period. I have basic account and once i master this level will be moving to upgrade. Does anyone know where i can read about OWNING ALL THE MUSIC ( BEAT, VOICE, and EVERYTHING ELSE) you create with SUNO at which MEMBERSHIP?

[verbal mortar]

I'm sorry, I dont' understand anything you are saying to me here. I have a persona but its not used for this at all. I don't use it for the metal stuff as it is so terrible that its a waste of credits.
These are render issues with the AI engine. nothing more. the prompts are fine, the lyric count is fine. the issue is all backend. something is broken or just has a bug.
There should not be that much noise or shimmer on a audio render. it's insane. it's in audible at some points with the noise being so bad that it just takes over the track.

and the shimmer...omfg. It's sooo bad again.

I think its got something to do with the complexity of the music you want. Drum and Bass and Electronic stuff (hip hop, DnB, trance, etc) seem to come out mostly clean and usable on first pass. Same with rock and folk and country. even classical. but metal...no, metal gets dirty and all messed up

I've remastered that track 5x already and it gets worse and worse with each re-master

https://suno.com/song/40592623-0847-4e80-be30-df7f033df975

This is terrible. it's unusable.

[quasi edge]

Have you tried editing the file in a DAW ? This is where Timbaland normally uses that approach. I mean everyone has their workflow. The final product Mastered for distribution.

[verbal mortar]

Yeah I have Ableton and its kitted with izotope and wave plug ins. as well as Focusrite and Klevgrand plug ins.

Obivously i'm not nearly as good as he is with mastering

[vital carbon]

is audio upload broken right now for everyone? it was working fine 2 days ago but now its just stuck in uploading.

[quasi edge]

This is where those skills and experience come into play. It's all getting the frequencies balanced that's all without the effects. SUNO works for me as a basic level. I can see it's potential.

[quasi edge]

check network status. You can research networks status sometimes they do go down or bottlenecks occur. By the way, you do know ocean fiber cables are being sabotaged? Well now you know where to research. Nothing is 100% maybe 99.9999% That what you are always told.

[polar spade]

The FAQs are a helpful start. And where most Terms of Services are just about we agreeing to be nice and all, these Terms of Service spell out what you agree to, the contingencies of what using the service means and Sunos position as to your use and rights/ownership. Take a slow read to it.

Much of it, well most all of it is just based on complying with the present laws of ownership and copyrights of the present music industry and artistry all together. So, it really behooves every member to find some means of giving yourself a crash course in royalties, publication, distribution, doing covers, creating art, documenting. Filing and the works.

https://help.suno.com/en/
https://suno.com/terms

[quasi edge]

This is all good and all and my next question is " Does SUNO provide counseling services to get the Artist up and Running Commercializing their art/ songs ? " If NOT. Make it happen agenda will be a good bet. The quicker the better we get " The World Hidden Talents Exposed and For a Good Cause". I have studied Music Business Law and as an Artist / Music Producer / Audio Engineer can be a challenge to find a " MORAL COMPANY" which actually does not prey on its citizenry. I am sure many would need the service including myself. Many ill characters have damaged the industry just the facts here. Moving forward 2025, lets make that change. Also, Crypto is the " NEW GOLD". A great incentive to think about. I got many ideas as i see things changing faster and faster. Any information is greatly appreciated. Let's Gooooo.

[polar spade]

Good to have dreams and enthusiasm. For what it's worth, the idea of generative AI is even far more reaching than music as it's quite a therapeutic device as well.

So hang out and meet all the different "personas" we have here. Welcome to the channel.

[blissful turtle]

Is there anyway to record your own vocals instead of the AI voice?

[quasi edge]

I believe you have to pre record the vox/voice / lyrics and upload it. Any one can correct me on this.

[polar spade]

In short, using upload (record audio) can seed your voice for Suno to randomly generate either something a bit off, or may even get close to your voice. Then you create a song from the upload, or later you create a persona of yourself, if it works, to be used on new randomly created songs. Edit: I got ahead of myself. That's a future tool from the year 2027. 😅 Seriously, I forgot at this time you can't create personas from audio uploads.

But using an exact copy for covering any song. No, it's likely designed to steer away from that.

[balmy monolith]

Are image uploads not working for anyone else right now?

[balmy monolith]

Hmm seems the trick is to resize the image in the preview pane once uploaded...weird

[dry dove]

How to use the option cover? i heard it came out but I can't find it.

[raven turtle]

Suno seems to be having an off day for me. Not following prompts at all. And uploaded vocals I create aren't followed or interpreted at all. V3.5 doing a slightly better job at it ATM, but not how it was a week ago.

[vernal epoch]

Please add support for the Vietnamese font when exporting videos on Suno.ai

[stable belfry]

has anyone successfully used their own voice as a persona?? just curious. i haven’t had time to try as im juggling making app updates to some of my projects

[unique stone]

This isn't supported.
If an upload contains a vocal, it can't be used for making a persona.

[stable belfry]

hmm, no? what if i Cover or extend, then the generated voice sounds similar tto me? i could make a persona of that yeah?

[unique stone]

If you manage to generate a voice similar to yours, it's possible.
But if it originates in any way from an audio upload containing a voice, it can't be made public, nothing made from it can be made public and it can't be turned into a persona.

[stable belfry]

what? even my own voice?

[unique stone]

Any voice. Anything that sounds like a voice (certain synth patches or other instruments get detected as a voice sometimes.)

[celest heron]

That would be awesome if we could 🙂 I wonder if anyone has been able to use outside tools to create a song using their own voice (even if they can't sing like me lol). So train an AI model with their voice (ElevenLabs or something similar), and use other tools to create a song with that trained voice. If that is possible, I need to learn how to do that.

Apparently it is possible according to ChatGPT...awesome. I'm going to learn how to do that!

[balmy raven]

Sometimes my Personas aren't being used. Is anyone else having this issue?

[naive quail]

Not from an upload with vocals; Suno will remember the connection. As Furrybeans said, uploads with detected vocals get flagged so that they cannot be made public and personas cannot be made from them. While you CAN cover and extend an upload with vocals (which CAN result is vocals that sound similar to the voice in the upload), the link back to the original upload is remembered internally by Suno. You won't be able to set those extensions/covers public and you won't be able to make Personas from them.

(I initially typo'd that as Funnybeans)

[polar spade]

Edit: I got ahead of myself. That's a future tool from the year 2027. 😅 Seriously, I forgot at this time you CANNOT create personas from audio uploads ....of voice.... at minimum. 🤣 Sheesh. I'm still asleep

[naive quail]

The limitation is uploads with detected vocals. I can still make personas from my guitar uploads (at least the one where Suno doesn't think it's a person screaming).

(at least as recently as a week or so ago)

Ohhh... but I can't do it with that same upload anymore, so they may have cracked down completely on that.

not a complete crackdown, tho. I just made a persona from it.

[stable belfry]

sir you let me know if you figure this out 👀 my goal is not to do TOO much morphing of my own voice, i still want plausibility that it COULD be me rapping it lol

[outer sinew]

Good day. I'm trying to generate a song and it gives me an error like this: Error An error occurred while generating the song
What should I do?

[rain dome]

How do I make a song?

[spice birch]

Refresh your page

[outer sinew]

I tried: refresh the page, clear cache and cookies, history, re-login, doesn't work

[spice birch]

Are you using real artist names? Sometimes it won't let you do that, or if you're using license material lyrics

[toxic ridge]

Is it possible to get a key change within a song?

[spice birch]

At the top left corner, where it says custom move that button to the right for custom. Then write your lyrics

[spice birch]

I'm sure someone on here knows the answer

[mental echo]

Please extend the limit of duration to create songs from 4 minutes to 10

[spice birch]

Huh? 10min now?

Oh your asking lol would be nice

[mental echo]

Thank you so much ☺️

[naive quail]

That'd be great, but it'd also require 2.5x the computing power, which Is not free. They'd either have to take a big hit to their wallet, or pass the cost on to us.

[unique stone]

You can use the "Extend" button to generate another 2 minutes. Several times. Then "Get whole song" on the final part to join it all together.

That's how I approach it if I want e.g. a 13 minute long track.

[mental echo]

Is this possible with our own lyrics

[unique stone]

Yeah you can add more lyrics in the extensions. It gives you more control over the direction as you can choose the best part to continue with and not just get a completely different song each time.

[bronze pecan]

I find it bad that I get now only 1 notification when someone likes more than 1 song :/
How long is the timespan until I see a like from same person again?
But I find it bad. If you want to prevent like for like problem, I suggest you have to listen at least 30 seconds of the song until you can like it

I myself always only gave songs a like which I indeed liked. I gave yesterday a person quite a bunch of likes because I liked the songs and when he did see only 1 like now its a bit of a shame :/

[woeful grove]

has 3.5 degraded? my generations have some sort of grainy - radio audio quality, voices sound really poor.

[bronze pecan]

I doubt that 3.5 is touched again, they work on the beta things I would assume

but it also depends on the music you create. harder metal has less quality than softer songs

[woeful grove]

I would have thought the same but dunno,

since 3.5 is the only available in free, may be to push ppl towards paid v4

the "shimmer" issue made me stop the subscription in V4 it was destroying half of the songs

[bronze pecan]

v4 has also variable quality

[woeful grove]

has somewhat been adressed? I don´t mind to pay, but in return I need something that works a bit bettern than release V4 did

[bronze pecan]

deutsch rock works good on 3.5 and 4 I noticed^^

[woeful grove]

in trance music, the shimmer thing, some sort of eccoing loop is insane

it is like a whatermark

sometimes does not show up but when it does it breaks everything

[rain hazel]

Has anyone found a way to download multiple files at once instead of clicking on each individual file to download?

[bronze pecan]

sometimes remastering it helps

[woeful grove]

and sometimes just shows up a couple of secs but it is like a watermark

I'll prepare some lyrics and revisit in a month to see how V4 is delivering... while voices where so much cleaner It was a bit of mixed feelings...

[unique stone]

Pretty much just working on v4 and other beta or new features as far as I'm aware. They aren't developing v3.5 or altering it. I checked when this came up before, and one of the devs confirmed it's still the same v3.5 as before v4 was launched.

[hoary rapids]

hey
can anyonetell
i have seen many people adding arrows in this what does that mean

[low sedge]

You can use this format:
https://cdn1.suno.ai/<song_id>.mp3

I've used that when downloading multiple songs using the javascript console and a list of ids...

[keen shale]

I noticed the wiki has been retired. Where can we find the song structures at now?

[cold zenith]

Hello, I'd like to know what the problem is that so far (V4), the degradation problem is still present. This problem has existed since the beginning and still hasn't been solved. Could you please tell us what the problem is, which means that up to now, this problem is still present? Because I'd really like to understand what the problem is. Because this problem affects 90% of my music and it's very disabling.

[dark spire]

We all would like ANY communication & transparency.

[cold zenith]

I'd really like to understand. Because when you have a subscription, you're obliged to use all your credits. You can't even save some for the next month, just in case things get resolved. You're forced to use up all your credits, but the biggest problem is still there. And this will generate double discontent. Because you're forced to use all your credits, and the deterioration is going to get on your nerves as you watch your credits melt away with no conclusive result.

[hybrid jolt]

Hi all,
My French lyrics are sometimes pronounced with an Anglo-Saxon intonation. How can I fix this?

[stable belfry]

are there any community servers for AI music from a technical/coding perspective??

[fossil bridge]

You tried defining that a french accent should be used?

You can use something like this in the header instructions
[VOCALS: A male singer with a perfect French accent, smooth and rich tone, resonant delivery, and precise enunciation. Singing in native French with natural phrasing, capturing the romantic and melodic nuances of the language. Add subtle vibrato for an emotive, authentic performance.]

[fossil bridge]

stem > remaster > stem is what works for me

You should remaster a couple of times until it gets the instruments right. I've noticed it will change things up a bit.

[last pewter]

Stem of stem? 🤔

[fossil bridge]

When you generate a song.. say you like it but it contains the shimmer.. You can stem it to split the song into instrumental and vocal. Then you can remaster on top of that to clean up say if the vocal is garbaled or not clear enough. The instrumental may have some odd quirks in it. This will correct them. Then once you have the remaster tracks you like.. stem it again to clean it, cause the shimmer will show up again.

[last pewter]

Interesting

[hybrid jolt]

Hi Crymic,
Thanks for your reply. I will try your suggestion.
So far I use this kind of tag: [LANGUAGE: French]

[last pewter]

If shimmer will show up again , than you did nothing....

[fossil bridge]

shimmer goes away on the stem, but the remaster will readd it in

I suggest remastering the stem track like 2-4 times to get a better sounding version.

as I said I found it alters instruments

[last pewter]

Got it ! Thanx , i will try

[hybrid jolt]

FYI: I tried but it didn't change anything. The problem of mispronunciation is apparently a recurring problem. So I will wait until it is corrected in a next update.

[random oracle]

"Make same length as selection" isn't working at all for me right now. The section I'm replacing is 0:24.4 but every single generation is longer... usually 0:33 or 0:38 but I've even had one that was 2:02. The setting is definitely enabled and I've tried disabling/re-enabling numerous times but it makes no difference.

[fossil bridge]

You might need to mess with the way things are spelled in the lyrics

or add details on the lyrics itself.. I haven't really messed with other languages

[cold zenith]

I wasn't talking about the shimmering but the degradation of the music

[naive quail]

A common work-around that many of us use is to spell out those mispronounced words phonetically.

[random oracle]

phonetic spelling can help (e.g. "Philip" would always sound like "Fill lip" in Suno, but I pronounce my name "Fill up" so I used "Philup" in the lyrics and it pronounces it correctly every time)

[naive quail]

Much of the "natural language" prompts like this is just prompt noise - words that Suno will either ignore or interpret in unexpected ways. That misinterpretation can often result in some interesting things, but it can also trigger some undesirable hallucinations.

[hybrid jolt]

For the word "question", I replaced the "qu" with a "k". It works well, except once when the AI ​​tried to sing all the French lyrics in English. Interesting to listen to but unexeploitable.
So I added the tag [Language: French] and now I have no more problems, at least on this word.
I will try other changes on other words and will test a phonetic writing to see where it leads me.
Thank you for your advice and suggestions 😉

[naive quail]

I created an accent by spelling out the bulk of the words phonetically for this track. Admittedly, not a very good accent, but it was just a test: https://suno.com/song/5254d589-30a7-4851-b396-882d3902a296

So:

Day and night, in the dark.
Oh, I can’t contain,
Your spark...```

Became:

```Oh, I kohl yohr naym,
Deh en nite, in deh dahk.
Oh, I cahn’t konteyn,
Yohr spahk... Et lit my haart!!```

[cold zenith]

Wasn't talking about shimmer but the degradation, the thing that is there from the v2 to now

[hybrid jolt]

Does anyone use the '<SONG_DETAILS>' tag to give general instructions like key, measures, tempo etc.?
I personally do it but I also tried to put the instructions without putting '<SONG_DETAILS>', for example: [TEMPO: 100 BPM][KEY: G Major][TIME SIGNATURE: 4/4]
Do you have an opinion on the question?

[fossil bridge]

ya I do

[hybrid jolt]

I tested the tag [SOUNDS LIKE: Pink Floyd] and the result was surprising: the bass was very close to that of "Another Brick In The Wall". The rest of the instrumentation was different, but it was quite impressive.

[fossil bridge]

Try [SOUNDS LIKE: The Band *Pink Flyod*]

you may want to include in their typical arrangements and mood to sound closer to them

[hybrid jolt]

The test I told you about
https://suno.com/song/4312985b-b25b-4b90-9e0e-7544b0ee3019

[fossil bridge]

You can also define particular songs as well.. [SOUNDS LIKE: The Band *Pink Flyod - "Song Title"*]

Can also try
[SOUNDS LIKE: *Pink Floyd* meets *Porcupine Tree* and *The Mahavishnu Orchestra*]

or try infleunced by

[hybrid jolt]

great! thank you

[cunning mulch]

just a tip... avoid using the ':' character in tags. Use '-' instead

because if you ever use [Vocals: <blah>], you will have that sung

[fossil bridge]

I haven't had that issue. I'm making instrumental TTRPG music.

Now if you use it within [Verse] Should be [Verse: *Instructions*] it won't read it

[tranquil estuary]

Does anyone know if there is anyway to guide the vocals in the lyrics window? Like making it emphasize certain words or syllables? Or sing a specific word louder?

[neat nexus]

You can try emphasizing by capitalizing the word. Sometimes it helps. I've also sometimes gotten good results with formatting the text.

Like I'd say
THIS
and hope that
it'd get the
IDEA.

The issue is the AI thinks for itself. It's not a wordprocessor for sound. So you have to hint and cajole it into doing what you want.

[eternal vine]

Hi guys my video won't download the updated title and image any ideas?

[austere minnow]

WHY dont permit the vocals ?

[hollow hare]

The video does not get updated until you use Regenerate Video (subscription locked) as it would be too costly to regen every time someone changes a title or lyrics.

For more details, see #💬┃general-chat message

[hollow hare]

Because people will commit impersonation

You can upload audio with vocals but it cannot be made public on Suno - the only way to share the derivatives would be through link sharing

[mint nexus]

This is what I've been doing, with mispronunciations too.

[mild rain]

where do you see the Workspaces once you have created one and moved some creations in to it?

[hollow hare]

Via Create tab. Just close the creation popup

[mild rain]

Ahh perfect thank you very much 🙂

[muted jay]

How can you generate music sheets from your songs?

[naive quail]

Bah. The day after I add a few extra menu items, Suno goes and adds icons to the original list. Now I need to add 'em to mine.

Tho I did already have one for the trash

[spice birch]

Why they changing things up with the set-up is beyond me. Cause ain't notin wrong with how it was.

I liked it How I can see the listing. now I got a search for what I was doing

[near badger]

Hello Community, I'm curious regarding certain genre prompts for the Suno program. I'm trying to create songs that have the Tone and field of five finger Death Punch, along with lyrics that while aggressive are easily to understand. In previous generations I've had the best luck using the genre "groove rock" rather than "groove metal" which is five finger Death punches genre would agree. With that said however anytime soon detect "Metal" in the genre prompt it always default to over aggressive thrash metal sounding vibes with growling lyrics that are hard to understand at best. Can anyone suggest a better prompt to be more consistent with the five finger Death Punch sound rather than burn a bunch of generations using "groove rock" until I get lucky. Thanks in advance for anyone goose bumped into the same problem I look forward to your responses.

[quick wedge]

You can’t generate music sheets yet. Suno doesn’t have that feature.

[spice birch]

OMG! Every time I use a Personas it makes beginning of the song ( Verse1 ) it's talking instead of singing. I mean every Personas on every song!!! 😡

[barren gale]

How do you get the Cursive singing style?

[spice birch]

OMG I'm about blow!! They need fix this shxx!

[blissful turtle]

how do you change the lyrics to the song and it doesn't play back what 's been changed the correct version?

[spice birch]

You can try to change your lyrics on that same song. By doing a cover of it, but it don't always work.

[blissful turtle]

This is why I want the instrumental stem but when I downloaded I can still hear the AI voice so I could record over it

[cold zenith]

It's really impossible to have a reply from my post from them....I'm disappointed 😞

[rocky storm]

New here and playing around with this and searching I see that you can't create a persona with your own vocals so let me ask this. I'm a baritone and mostly played acoustic gigs, not a ton of covers but was always able to do Goo Goo Dolls and Howie Day song well as my voice is in that "breathy" sounding area...

I've been prompts like breathy male vocals, baritone but wanted to jump in here and see if anyone had other ideas or tips on how you got it to sound similar to you?

[rancid notch]

does anyone know how to fix it not creating my songs at all

like it just says a system error occurred

[hollow hare]

Are you using v2 or 3.0? They have a 1250 char limit (3.0 max char limit still has not been corrected yet)

[rancid notch]

How do i see what version im using i just started using it

[hollow hare]

Changing the lyrics will not change the music, it is only for visual appearance.

If you are referring to the video, it does not get updated until you use Regenerate Video (subscription locked) as it would be too costly to regen every time someone changes a title or lyrics.

For more details, see #💬┃general-chat message

[hollow hare]

When you create, the option is to the top right

If you have not done so yet, check out #1294418290348921007

[flint haven]

How to use Suno

Hi there!
If you’re ready to dive into creating music, head over to Suno.com.

For more detailed information, check out these resources:

• Making Music
• Billing
• Rights and Ownership
• Accounts
• Troubleshooting
• Mobile App
• Suno Server

If you have any more questions, just ask in #general-suno, where you can always find our community, moderators, and guides to help you with most questions. 😊

[rancid notch]

This is all i see when i try to look at the version

[hollow hare]

You are using https://suno.com right?
If so, there may be an extension blocking that element from appearing

Try using incognito/private mode and/or another browser to see if it is a potential extension issue or a browser issue

[rancid notch]

Yes

[rancid notch]

Ill try that

Its still not working

[dusk shadow]

Hello. Why are the uploads of original songs that I sing only 1:59?

[rancid notch]

I got it fixed

Thx for the help

[blissful turtle]

This is why I want to record my own vocals in Logic Pro when I tried to download the stem instrumentals but I still here the AI voice in the background

[olive urchin]

I have a song that is about 4:00 minutes long. I have my own lyrics. After entering the lyrics and clicking "create" the song is not generating properly. It stops about 80% to 90% through, and a couple times is simply starts over about 30 seconds into the song. I've lost about 50 credits trying to "create" more versions but it is not functioning correctly. Ideas? Help suggestions?

[polar spade]

You may upload up to 120 seconds on a paid subscription. Are you asking why it stopped at 119 seconds?

[untold briar]

Suno infinite credits glitch. Edit->Song Details->Add New Image->"piano sheet music, close-up sheet music, sheet music" https://suno.com/song/b43d7367-3f73-401d-984e-06e77cec899e

[covert marsh]

Suno cannot create sheet music, however there are a few options you could use to create sheet music after you have created and downloaded a song:

MuseScore
AnthemScore
ScoreCloud
Melody Scanner

are some. I imagine there are others. I haven't used any of these, but it's worth a try.

[covert marsh]

Keep asking it to make it. Sorry, that's the nature of AI creation at this time. Usually takes me about 20 to 50 tries to get what I'm wanting. I'd be using other programs for music creation if I could sing. But alas, those vocal chords have gone. But Suno is fun. It's also great for inspiration as well.

[solemn echo]

Hello everyone. I wrote a song and it turned out well, but I have 1 problem, there is 1 spelling mistake, how can I fix it?

[sleek canopy]

So, just renewed my suno pro subscription today and when I went to go make a song, it says I don't have enough credits despite my having 2500, as well as when I go to generate a v4 song, it says I have 10 free v4 songs remaining, which I should have unlimited v4 songs for the time I have my subscription, but no.

I checked my bank account and the money was indeed taken, and I also have a receipt from suno saying I bought it

Please help??

Idk if this is even the right channel to use for this type of issue

[solid depot]

I really need some help with this. I wasted now over 500 points on a single damn song which REFUSES to continue at the Bridge and ALWAYS repeats the 2nd verse and chorus.

I mean if it happens a FEW times okay, but THIS MUCH without a single time difference is getting ridicolous!

[untold briar]

Things to try:

1. Change the Extend From time, change it as much as you can while keeping the part of the song you absolute need, but if you can't, backing up even a few second could help. (didn't test)
2. Try Replacing over the failed Extension instead (crop off the end of the song if you need to). (didn't test)
3. Cover the song, using the same style prompt and lyrics, but adding a new verse after the Bridge (works)
4. Change [bridge] to [post chorus] or [hook] and Extend as normal. (works)

[solid depot]

cant do the 2 things since im just a dirty free user -.-

which makes it more infuriating wasting now 10 days on this

[untold briar]

If the song is less than 2:00 you could try Uploading and Extending, as a free user.

Oh it's 1:00 for free users, actually, probably won't work.

[solid depot]

Its "part 3" so i already have an intro i made a lot earlier and the first part extendet

[untold briar]

Oh, if you are extending Part 3, try extending from Get Whole Song instead.

Or the opposite.

[solid depot]

and yes, i even rewound BACK until verse 2 and chorus and right after the chorus it repeats verse 2 AGAIN

every single time without fail

[untold briar]

If you link me the song, I'll Extend it, then Cover it, then link you back that version (assuming that works).

[solid depot]

ill try the extending from whole song now i guess...

[untold briar]

A cover will change the audio of course, somewhat

[solid depot]

and another 4 attempts, STILL repeating the damn 2nd verse. Holy shitterony this is infuriating. Never was it THIS bad when i made a song

I sent it to ya as PM, as to not spamm here any more then nessesary

[celest heron]

Holy crap, I cannot get Suno to pronounce 'horizon' correctly no matter what I try. I've tried her-eye-zen, her-eye-zon, her-i-zen, her-i-zon, her-eyez-en, her-eyez-on...ugh!

[hollow hare]

Hurizan? Huryezan? Hurreyezen? (Adding dashes would give it a small pause)

[celest heron]

Thanks for the other suggestions, I'll give them a try

[unkempt sinew]

I had some suggestions but it would probably ban me from the channel. heh.

[spark cargo]

Try putting this at the top of the lyric box. It may help.

[Reset Memory]
[Clear Memory]

It's helped me in the past.

[celest heron]

I know exactly what word you are talking about lol.

[unkempt sinew]

Phew, great! Try that one 🙂

[celest heron]

Damn, still none of those suggestions work. I've made like 30 replace segments now, frustrating!

[autumn jolt]

Any suggestions for extending V4? I generated a song I love last night but after 2 extensions the 3rd just would not naturally flow on music or vocals at all. I picked something that sort of worked because on its own it was great. But no matter where I extend from there were issues. Do some people revert to 3.5 or 3 to extend and then revert back to 4? Any advice appreciated as I haven't used V4 that much since it was introduced.

[lethal lantern]

Hello everyone. I have a quick question: would you know how to re-use the same voice on different tracks? I have the impression that the same voice can be reused if you keep the same musical style, but maybe there's a system for locating the signature code of the voice concerned to include it in the prompt like [voice ref :.....]
If you have the answer, don't hesitate..

[celest heron]

I don't think it's possible

Can maybe try making a persona, and then keep making generations...might get really close. But it's all trial and error.

[humble shale]

Why can't I extend my songs? I have credits and I'm subscribed to Suno. Is the website slow or something?

[lethal lantern]

i was thinking of this solution but i think i also will sadly have to spend a lot of credits on it 😦

[celest heron]

Every single one of my replace segments is coming out super glitchy right now...I think something is going on with the site. Just the last 5-10 minutes.

[celest heron]

Yeah...exactly. It will cost credits for sure

[humble shale]

Dang, hopefully this'll be resolved soon

It's not even taking credits to attempt an extention

[celest heron]

I got lucky and have two songs that sound like the exact same singer, but I spent a lot of credits to hit that jackpot lol.

[lethal lantern]

I was wondering if there wasn't a “signature code” to identify the voices, which I'd like to use.

[celest heron]

Interesting

[celest heron]

I'd love that to be a feature. I'm sure it will down the road.

[lethal lantern]

If the idea hasn't been submitted to Suno yet, maybe I'll ask, I think a lot of people would be interested.

[humble shale]

There is personas which contain unique voices and can be reused for new songs, but it's a paid feature

[celest heron]

Oh, I'm sure it's been submitted. Everyone wants that feature.

[celest heron]

It's more of a style, not voice...as far as I know. At least it hasn't worked for me the dozens of times I tried that.

[lavish holly]

can someone tell me if there a fix for upload audio that switch from 4v to 3.5? i need to change it manualy everytime i upload a audio to 4v

[humble shale]

Oh, I thought you could make a vocal persona. I guess not?

[celest heron]

Unfortunately not. Like I mentioned to the other person above, you might get lucky and get the same voice if you keep making generations, but it's certainly not guaranteed.

[humble shale]

Alright

[celest heron]

Can't help with this one, buddy. I've never uploaded audio yet.

[frail gazelle]

my songs sometimes are over 4 minutes. how can i fade this song to fit the time allowed

[celest heron]

You can use the extend feature to make songs over 4 minutes.

If it gets cut off, crop it where you want. Then extend using the lyrics from that crop point, if that makes sense.