📚┃suno-school

[half dragon]

I'm afriad that you'd have to live with it if you're using Suno 🙂

[lusty juniper]

yeah the moment i saw this i thought i finsihed my last song with suno today

i hate when a decent interfaces becomes like some Apple style "hover" your mouse before it appears trying to be all futuristic

but yeah lets see maybe its not yet complete

[zenith smelt]

Hi maybe this is a better please to ask my technical questions being new to suno

Can you use suno to upload a voice sample so you can use your voice on your music ?

Or am I still forced to do that in other programs like kit.ai which im having bad aberration in the way ot uses.my voice

[half dragon]

Yes you can - but keep in mind Suno doesn't clone your voice so it'd be as close but never the exact same.

[lusty juniper]

you can sing a song and suno will cover your song, if you are lucky it sounds like you

[candid moat]

Suno has changed. I can't get what I type in, and I can't find the Lyric model anymore either.

[lusty juniper]

yeah i was working on a song for my album and i was in the zone knowing what i wanted and then it changed and i lost what im working on.

i don't know about others but i hate it when someone touches my workplace. i gotta readjust myself

[lusty juniper]

right now this is how it looks like....

why not this

[rare moth]

Hi Suno!
I’d love to have my Workspaces as a top menu item on the Library page.
Thanks for all your hard work!

[polar spade]

Gnarly, dude. 🤘 i guess Ryan didnt like the suggestion though. 🫣

[polar spade]

I like when its pretty close that I can say its me, but better so I dont sound like I stink at singing.

[queen linden]

anyway to get Suno to understand I want the word Wind ( as in winding road) instead of wind ( as in blowing wind) ? I really dont want to rewrite a whole new chorus....

[lusty juniper]

P.S. the current UI at the style section where i type in the style for me it is super super delayed. if i type for example acoustic, its going to be "aotc"
it can't keep up. so i have to type it else where on my computer and copy paste

not sure if this is happening with anyone else

[viral pond]

i lost the prompt templates

where are they hiddeb

[rapid summit]

They dont work correctly anymore either

they are in prompt

the bookshelf like emoji

[viral pond]

that was invisible 😄

i dont really like the prompt template, it's really non interactive, I might have to make my own

[velvet granite]

Hi! I'm still trying to get a solution for this. It's pointless for me to continue using Suno until I figure out why it's returning these kind of results. I hope someone here can help me. I was asked if I used a persona and the answer is yes about half the time. How does this affect a 58- second return?

[topaz summit]

Yeah I'm experiencing the same, also the music quality is significantly worse than it was before

[polar spade]

Spell it "whined" or even "wined" as in wined and dined?

[ionic turret]

Orrrr the change is because something new is coming and the layout needs to change for that

[hard venture]

I thought I saw an api link on the Suno sight a few months ago. Also, where do we find info on training affiliations and partnerships in general?

[half dragon]

There was never such a thing. Do not trust any 3rd party site that says otherwise.

[solar badger]

Thank you. Appreciated

Didn’t read all the messages but i do have similar issue last week. Personally i think your meta tag made the model confused. I’m not an expert though been spending 2000 credits to make one track but I couldn’t get it right either 😅. Try duet instead of male and female singing at the same time, i think that tag is too long. And bracket (…) didn’t work well, i tried before as well. Your prompt style couple singing might not works -> duet maybe. About the scream metalcore trigger scream for sure.

[graceful condor]

I asked this for the 3rd time. Would be nice not being ignored all the time. Suno is a mess lately and I noticed that some stuff is highly interfering with each other. SO PLEASE EXPLAIN
If I have as follow:

Persona
AUDIO Track
Detailed complex Lyric Prompt (includes description of vocals too)
Detailed complex Style Prompt (includes description of vocals too)
Exclude
Sliders used mostly around this Weird: 20 / Style: 85 / Audio between 90-100 (especially if I use a Persona)

So if I use all of that at once. What does Suno prioritize the most and what interferes with each other and what kinda cancels the effect of the other one?
This is what I really would like to know.
Thank you

[smoky dome]

Why has the custom create lyrics auto button disappeared?
Currently on annual payment pro plan

[jade mist]

Anyone who uses accessibility tools like a screen reader on Windows: Do you know how to adjust the new sliders in the create page? For example weirdness etc. They used to be standard sliders, but since the create page updates, it got inaccessible

[true dock]

I need the exact release date for Suno Studio – is it out yet?

[flint kraken]

Can I get some help with a song? It's just cutting out unexpectedly no matter what i do.

[pure ice]

Yeah wth, it was so great to be able to just put a few things there and have different styles and lyrics every try. Why would they get rid of it??

[random surge]

Is anyone else having issues with every damn song cutting off early? way before 8 mins? it's absolutely pathetic what the hell Suno?????

It has been happening non stop for days if it continues i will go elsewhere and let everyone know on social media. It wastes so much time and ruins amazing mixes. Bloody fix it immediately

Not to mention it doesnt follow lyrics properly and the changes to the description box are a major step backwards, so well done to the staff

oh great and another one cut off, this is pathetic

🤬

[ebon sand]

I would love to have a customer who barges into my store, threatens me with ruining my reputation on social media and leaving for good if I don't jump immediately... Haha

[random surge]

wouldnt have to if what is advertised works, thanks for your useless opinions

[ebon sand]

You are very welcome, at least I'm entertained

[random surge]

small mind i take it

[flint kraken]

yeah, I'm a little discouraged too.

[random surge]

it's happening to every song, so perhaps the trolls here dont care but true creators do. 😉

ooops troll incomming

[ebon sand]

Oh I do care about I don't like to see this mindless threatening

[flint kraken]

Yeah, it does indeed look like a bug. I'm sure they're working on it.

[opaque portal]

They took out "Auto Lyrics" ?

[random surge]

the troll actually thinks i care about his comments which is priceless lol

they have made several changes all taking it backwards

[unique stone]

If you are using Persona, Cover or Inspo some of the 8 minute limit is used for accommodating the original audio to use as context/reference.

Use Extend to continue from where the song cut off.

[flint kraken]

hmmm. that could be difficult considering it cut off mid lyric

[unique stone]

The time it starts from can be adjusted so you can set it a few seconds earlier if it did this

[random surge]

it always cuts off mid lyric so you cant even fade it out in other software

[flint kraken]

ahh - that's cool. I might give that a try. Thanks for jumping in!

hmm. i tried that.. it only extended it by 10 seconds or so.

no wait, i dun goof'd got it now!

@unique stone Thank you so much for jumping in on this. I've got exactly what i wanted!

[sweet siren]

@random surge Suno support reacts really well. Might sometimes take a few days but they do.
Often when my song cuts off it has to do with me putting too much info in both the lyrics and style boxes. Not to say sometimes it happens regardless.

[quick sonnet]

I have track which looks like:
insrumental intro
verse 1
chorus
verse 2
chorus
bridge
chorus
vocal outro
instrumental outro with fade out
Suno generated a realy good result, but outros mising. Also not 100% happy with the bridge, since it sounds more like a short verse 3. So think extend could be an option to fix it. So how to tell suno make the extended track exactly the way, once it merged, it sounds like it was one track? Also is it posible to select, from where the extending should start?

[ebon sand]

Yes, when you extend there is the option to select from where the extension is supposed to start
You might have to try a bit, sometimes the extensions tend to get a bit more creative which sometimes doesn't fit the idea you had

For the bridge you could try to edit it and attempt to get a different delivery

If the extension should consistently fail to give you anything reasonable you could also try to do a replacement in the editor instead; the editor isn't very good unfortunately, but it offers the option that you keep the end that you already might have, move it further to the back to create a space and then replace the space you created for the outro

[graceful condor]

That's Suno since I use it. It makes no sense when or if it cuts off songs. And most of the time it does. I experimented a lot it seems it can not stand silent parts which makes it think the song comes to an end. But not even that is really the case because if a song is silent to begin with like a ballad it can go on and finish it as a whole. And then you have a song full of power and action and it cuts it off.

At least I noticed that most rap songs are finished as a whole. Metal is a huge issue. Also instrumentals are nearly never cut off only if there is singing. It has really no reasoning it seems really random at many cases. But then it sticks to it like glue. Like the same time over and over again even if you create a new song with the same prompt it will cut it off for sure in 99,9% at the same time like always at 4:56 for example or 4:30. I really hate it.

Especially if the Extend is ruining your whole song like it was on my case. I had a perfect part 1 clean, no issues, artifacts etc. Then I created a couple of extends and after merging them together the whole song got artifacts even in Part 1 which was already good to go.

[polar spade]

You should still be able to have Suno write your lyrics if you wish.

If you mean in the Lyrics Box, you may now click the Magic wand button (it turns orange if you type something in the lyrics box) and choose "Write Lyrics"

[graceful condor]

Suno reacts to nothing. They do not inform us in any way only say "Thank you for telling us" and that's it and nothing changes. No help at all. Same as here most of the time. I already asked for advice 3 times within the last 2 days and no reaction at all.

[ebon sand]

It's just users helping users here

Maybe they just don't know why you have problems and therefore didn't answer your question

[wispy laurel]

Persona and Audio track are both controlled by the Audio slider. I'm not sure if either one naturally has priority over the other, but ideally your persona should usually be voice-only - in which case the persona's vocals should override any from the audio track, because at high Audio influence it has to use something from the persona and if the voice is the only thing, it'll be forced to use it
From my experience, style prompt will usually override conflicting or differing prompts within the lyrics. But it appears that adherence to lyrical prompts is not controlled by the style slider, while the Style prompt is (evidence for this, I quite often cover songs with 0% style influence, and find that prompts within the Lyrics box are still followed most of the time)
Exclude doesn't seem to be reliable at all, and also doesn't seem to be affected by any of the sliders

[patent widget]

New UI messed things up. Tragic. Lyric selector.

[graceful condor]

First thank you. And I just noticed especially with the new Changes (UI and what we not know might cause stuff in the back) that if I use a Persona and an Audio (pure Instrumental) that the Audio lately hardly has any effect anymore onto the finished end result. I hardly hear any melodies of the OG Audio I used. Which is weird. Also I get the feeling that it can help but also changes the course of the voice if I have detailed describtion of the voice in Style and Lyrics while using a Persona. Also personas are very inconsistent to begin with sometimes they start strong as they should sound like and then loses their tone on the way.

This is why I asked because a lot seems random, bug influenced (or what ever they are doing in the background) but also interfering with each other.

[ebon sand]

Yeah it lacks a bit of functionality, I hope they fix it quickly

[void forum]

Y'all can leave feedback about it at https://suno.com/?open_feedback=true ^^
I just did as well

[ebon sand]

If you use persona and audio it likely cuts off due to the limited context window, maybe you can shorten each, persona and the audio to get longer outputs if they aren't short in the first place

[paper reef]

wtf, i have a pro plan and now it tells me i cant generate anything cause i have no plan. that's professionalism all the way.

[wispy laurel]

Yeah any prompts relating to vocals will have an effect even if the voice is being drawn from a persona. You need to make sure that any voice-related prompts are at least somewhat compatible with what the persona naturally "wants" to do, otherwise it does get weird (if your persona is made from a soft ballad or a shouty metal song, and you want it to rap, you're gonna have a long and frustrating day 😂 )
Losing the melody from the original audio is weird, I don't think I've ever had that problem with a high Audio slider. Not sure why that would happen, unless your prompts or something in the persona are strongly conflicting with the audio track
Try switching models. 4.5 does some things better than 4.5+, and vice versa, you might need to switch back and forth during the course of completing a single song

[graceful condor]

Why would a shorter Audio make a longer song? I tried both long and short made no difference honestly.

[ebon sand]

Because they have a limited context window and they need space for the output, so if you spend a lot of that space for the input you can only have shorter outputs

[quick sonnet]

I know it's a gamble, but want to increase my chances. I am a free user so editor is not a possibility for me.

[graceful condor]

Does this mean longer and more detailed lyric prompts like full 5000 signs is shortening the song too?

[ebon sand]

No, only the provided Audio

And only if you are even over the threshold which I don't know by heart where it starts for the newer models; on v3.5 it shortened starting at around 2:10, for the new models it's later because they have a bigger context window

Likely somewhere at 4 minutes

[sweet siren]

I am sorry to hear that. Not my experience.

[graceful condor]

And weirdness I guess is also changing the Persona. So if I turn up weirdness it might not sound like the persona anymore, correct? At least what I experienced.

[near flume]

Declare only 1 outro, you can combine vocal & instrumental in one. Also use a specific type of Bridge to have it set apart from the verse.

[polar spade]

A longer audio will mostly likely CONTRIBUTE to shortening your new output.

A shorter one, will BENEFIT by having more "space" in that 8 miniute limit, to have a longer song.

Its not short audio = long output and long audio = short output necessarily, but thats how we should hedge our bets.

[near flume]

I get a personal reply within 3 days. But then I rarely email support unless there's a major issue.

[wispy laurel]

Weirdness affects everything
High weirdness will make it more likely to contradict your prompts, and also allows for more dramatic and noticeable changes from any and all source audio (persona, inspiration, and song being covered/extended)

[graceful condor]

But I thought it is good to have enough Persona material to give the AI the best source material to work with? I mean I know that an Audio is only looked at for the first 2 min maybe AI side wise. That's what I have been told. But with Persona it takes the whole file in consideration. At least that is what I have been told over and over again that more is better with Persona.

[polar spade]

Also, Weirdness is not specific to things sounding "a weird" concophany of noise all the time. It could also mean that a classical Quartet of violins might suddenly switch out and have a bassoon or even a guitar instead of a violin if we slide Weirdness up. So it can be a pleasant outcome.

In the past we had a thing called "Variations", that few people used, but I loved it. It provided a more controled form of Weirdness, much as if we set Weirdness to about 60%. So maybe it simple was redundant to have that Variation Slider and left control to us.

I think im getting off track and just providing trivial knowledge. So ill let you get back to your questions.

[wispy laurel]

That depends what you're using the persona to achieve
Mostly, it's to get a particular voice. The best way to achieve that, is to take just the vocal stem from a song, and crop it to just the "best" bits (where the voice most demonstrates the particular sound you want). Keep just the distinctive and recognisable parts.
Say, you want a low, husky, rough female voice with a Southern accent. You'd make a persona using only the part of a song where "she" sounds exactly like you want. That way, when Suno looks at the persona and starts extracting data from it, it's guaranteed that the only data it finds, is going to lead to that particular voice

[polar spade]

Yes. What you say is valid regarding "enough" . When everyone was getting cut-off songs once personas were introduced (last August?) we figured out that about 1 minute of the seed output is best for making a persona have enough info, but also take up less allocated space, leaving enough "seconds" of space to have a present decent length of new song output.

Since then, we now suggest, for power users, to edit an output with the "best of" information snippets down to about a minute, THEN make a persona from that.

[wispy laurel]

You want your persona to contain as little data as possible, and only data that specifically supports the sound that you're trying to create

[graceful condor]

But I really do not get the logic behind it why a Persona is taking away space? I mean it is only reference material like lyric prompt etc. This should not have any impact to the actual length of song especially if we are talking about that those are "Taking away creation time" which is so weird.

[polar spade]

Oh, I dont do logic questions. Anyone ? 😁

[graceful condor]

Great I have nearly 60 Personas. Which means I can cut them all down to 1 min. Man it took me forever to create those.

[wispy laurel]

I'm not a programmer or an AI specialist so I don't know the technical specifics
The way that I visualise it, imagine an artist or a sculptor or something, sitting at a table and making something.
The persona, source audio, whatever, are like objects sitting on that table. The artist can look at them as a reference for what they're creating - but, now they have less room to work in, so their final product can't be as big as it would have been, simply because it's now being made in a smaller space

[polar spade]

The Persona is not a thing of size. .... but only the same code for every Persona, to read the seed output. So the seed output is always referenced each time to use a Persona.... thats why you cant delete the seed output or else ther persona "code" has nothing to read.

So a 8 minute song as a seed output (when making a Persona) will take up more room than a 1 minute seed output if you use them to personafy a new output/song.

[graceful condor]

Man this was work for weeks. Maybe I will change them with time and cut some 1 min together. But this really killed my motivation right now. Because cutting down nearly 60 Personas...no thank you. Not in the mood right now. But thank you now I know. I just wished I had known it sooner.

[ebon sand]

I think they could add a value for the "remaining maximum length" somewhere but maybe then some people think it should output this length when they click create and not something in between 0 and the max

[polar spade]

Yep. We all have done the same. I just started earlier than you.

But truly, .... amd sorry to tell you this..... with each version, the older personas dont work well anyway. So we need to make new personas which incorporate setting from the most recent model. (It just says Create Persona" but any personas is created with the most recent model's traits.

So ive redone many personas, each time we get a new model.

The bottom line is though, if the old personas sounds good still, no worries, keep using it. And the 1 minut rule was made when we had a 4 minute time creation limit. So, technically, you can experiment with 2 minutes of seed information to make a persona (which secretly i use now) , in order to add more weight to a persona over the audio.

[polar spade]

Ill also say, theses are things I worked out with my little pea-brain over the time ive been here. I dont know any secrets or hidden programming info. I have my own conspiracies (i say tongue in cheek) and just have verified by lots of testing.

So, same as any AI, I could be wrong or what I say may not benefit what youre trying to accomplish.. Check your work and confirm with others.

[ebon sand]

Personas do not work well for me on the newer models overall, the last model which has worked well with them was V4

[polar spade]

I think my last personas was mostly 4.0. Only one 4.5 worked well at the time to get my aame fake-artist vocal that I lost back when she sang 3.5 songs. Recreating whole albums or recreating a new Bio for the fake-artist is a bear to do. 😒

[polar spade]

Just make a new Persoas when the old one isnt giving you the same vibe or characteristics to match what you had before. This is usually only important if trying to have the same "artist" sound or band's vibe consistent. So maybe a few personas, at most.

And certainly now try 2 minutes instead of 1 minute on your seed to get a bit more weight to it. Either 1 or 2 minute seeds , this should logically affect where you will need to set your audio slider settings, as well, so ...... a new learning curve for awhile ....if you make and use new personas.

[barren fox]

where do i post suno covers

[polar spade]

Damn we missed the Early Bird Breakfast discount cause we were busy being "no help."

[polar spade]

Suno Covers? Any output you make, if you wish, you may post in Suno Showcase. https://discord.com/channels/1069381916492562582/1083129341870878780

Otherwise, sometimes, if you publish/public your song, it will show up in the New song list on the Home page.

[storm pawn]

I've tried couple approaches but I'm failing and need to hear how you would handle it 🙂 I was playing with creating a song based on uploaded instrumental, it went really good, but vocal kinda sucked. So I played further, got great vocals in verses, but no original melody and poor chorus. Played further, got great chorus, but poor everything else 😄

What would be your approach to somehow create a cover that would use instrumental from option A, verses from option B and chorus from option C? Any way to do it in Suno or only in outside DAW, adjust the pitches so all match, then upload to Suno and cover to fix the "stitching"?

Is it possible to refer in Style section what parts of which songs from "Inspo" playlist it should use?

[quartz adder]

personas were a great way to do most of this, but I don't see v4.5 personas work at all

v4 personas were/are gold

but now the new UI blocks using v4 with personas

[polar spade]

Adding instrumental to vocals would provide a consistent chord progression to whatever melody the vocals have. But if you like a certain chorus, you cant presently patch one in.

So you can fiddle about with your chorus in your best vocal output fost, then stem out the vocals and add an instrumental. That will get you closer.

Otherwise. You can patch in the chorus with a DAW first, adjusting for pitch/tempo of course, upload into Suno and go from there.

[quartz adder]

screams into the void can we just allow v4 model for personas again?

[polar spade]

Are you on the list for using Suno Studio? That may have more DAW like features. Who knows. Still waiting. But dont delete your good stuff quite yet.

[storm pawn]

No, never knew I can get myself somehow on such list

[unique stone]

In what way is it blocking it? I'm able to add a persona with v4.

[pastel plaza]

hi ,how do i get a child to speak one line ? im trying [boy vocal – spoken word, high-pitched,natural boy voice, high tone, innocent]
it worked 2 times with not a good style song but no more.

[polar spade]

And no... we cant" choose" specific Playlist songs, even within one Playlist. Just put the best ones at the top and that should help get inspiration from those.

As @wispy laurel has noted before.... and often.... (search her comments posts here on Inspo and Inspire) . You can put each element as a song in your playlist... vocal stem from one.... instrumentsl from another..... maybe just the chorus you like as another output in the Playlist. Then click INSPIRE. The outcome is unpredictable, but you never know if something nice could come from it.

[polar spade]

There may not be list now. Let me check.

[polar spade]

Check out the announcement... I dont know if the link is still active. Ive seen some Hooks users reference it in their hooks. So it exists i guess for some to try.

#📢┃announcements message.

https://suno.com/studio-waitlist

[quartz adder]

if I try to extend an audio, and use a persona, it forces back to 4.5

[unique stone]

Ah. Support for using persona with extend was only introduced in v4.5

[storm pawn]

thanks! signed up

[polar spade]

Oh. Great! Good luck.

[pastel plaza]

anyone ?

[quartz adder]

no I've been using it for months with v4 to amazing results

the latest update made it impossible to select v4 now

and v4.5 completely ignores personas with extend!

[sullen marsh]

Why can’t you combine inspo and persona anymore?

Why do the personas always sound different?

Until yesterday, it was still possible in v4.5+ to combine a Persona with an Inspo playlist. Since the latest update, however, this no longer works – the system forces me to choose one or the other.

This combination was extremely valuable:

Persona = consistent voice/character

Inspo = stylistic variety through vocal stems and playlists

Please bring this option back – it was one of the biggest advantages for creative work.

[quartz adder]

exactly, there are new "trivial" guard rails around certain features like this

where the only reason it won't work is because it's blocked!

crazy powerful features

[sullen marsh]

Then I don’t really need Suno anymore if good features keep getting removed or blocked. That way, it just doesn’t stay usable in the long run.

[quartz adder]

@stuck junco could the team explore why now if you "extend" a sample and use a persona, it forces you to use v4.5 or better?

v4 is by FAR the best model for this workflow still

[jagged oar]

idk man my 4.0 personas are audibly not as good as my 4.5s

[queen linden]

I wasted a lot of credits.. so Im grumpy..... i really ( i can not express this enough) hate the YERNING its like having your words dipped in trecle .... Please please Suno allow negative prompts no belting, no high notes, no long sustain and above all ... stop the disney princess yerning sound!

[polar spade]

In discussion among some users about persona+inspo weeks back, I think it seemed that though we could choose , I think, inspo first, then add persona and they would show on the panel, test after test seemed to indicate only one was functioning. Usually, if I remember, the last one chosen, inspo or persona.

The other one just didnt "grey out" as it should were as most un-applied or turned-off items would do in those situations.

If you think they both worked together, thats cool. Maybe thats something for them to work on. I like when we can hear the influences of all the combos we apply.

I'll check out the recent settings you are talking about as well. I have NOT looked at that combo in awhile.

[sullen marsh]

It actually worked really well – Inspo + Persona was a huge thing, but now it’s gone and you can only choose one. That makes it unattractive for me.

[polar spade]

Cool. Something i didnt get in tests. Good to know it wasnt such a glitch and maybe they can go back to make it permanent. Keep giving the feedback on their feedback forms. If you miss it, likely 12,000 other users do too.

[sullen marsh]

Yes, absolutely – I really hope they bring it back. Inspo + Persona was a real game changer for many of us. I’ll definitely keep giving feedback, because it’s exactly features like this that make Suno exciting. If so many users miss it, it should be clear how important it is.

[sullen marsh]

You just can’t build anything consistent when features keep disappearing.

[quick sonnet]

How to combine the outros?

[sullen marsh]

What exactly do you mean by that?

[quick sonnet]

Don't wan my track to end after the last vocal of the vocal outro, should be stil some posibly short, but instrumental part after it with fade out.

[sullen marsh]

Try this in Suno for an instrumental outro with fade-out

Try adding a note at the end of your lyrics that you want an instrumental outro with a soft fade-out. In the style prompt, mention something like “short instrumental outro with smooth fade out.”

If it still ends right after the vocals, try using Extend and describe the instrumental outro there.

That way you’ve got a good chance of getting a short instrumental part that fades out nicely.

[near flume]

I use [Extended Instrumental Outro: Fade Out] then place your vocals under the tag, once the vocals stop it will continue with the instrumental part.

[quick sonnet]

I try it, but it could theoreticaly generate instrumental outro first and then the vocal one, or am I mising something?

[sullen marsh]

No, normally the instrumental outro comes after the vocal outro. If you specify both in the right order, Suno should follow that.

[brazen sand]

alright ive been busting my buns trying to get suno to give me a pitched up vocal on my chorus. The idea is like what "Chill Bae" by Lil Uzi Vert and "Yukon" by Justin beibs has. That sped up vocal sound but without it being sped up. Suno just refuses to do it. any expertise and help would be greatly appreciated. its 7am and i havent slept yet so im slumping, but I hope anyone knows how to help. Thamks in advance

oh yeah and my main vocals are male, idk if that matters

gn y'all

[random pine]

Does anyone know if it's possible to cover an exisiting AI track generated earlier in Suno, but then replacing the AI vocals with your own voice? So basically re-amping the vocals with a me-sauce.

I have an instrumental, guide track (instrumental+AI vox) and acapella vocals from a song I recorded in studio.

[quartz adder]

when I use personas in 4.5 to extend it straight up ignores it

I've tested it hundred of times

4.0 it worked

it just picks a different voice

no interest whatsoever in the persona!

[random pine]

I did something similar by putting in the segment in the lyrics field that there needed to be modulation. So [Chorus - Modulation]. That seemed to work. My guide track did have it too, so it might be a false positive. But work trying I guess

[wispy laurel]

You can't do it within Suno, there's literally no way of making it incorporate your recorded vocal into a song.
It can try to copy your voice, but if you want to actually use your own vocals you'd have to use an external DAW to combine them with Suno's instrumental

[random pine]

Hmm meh, I saw your tech help question and was like ''AMAZING IDEA'', sadly it's broken haha. I am not looking to release it. Just looking for guide tracks that have vocals that are very similar to mine so it's easier to hear how the result should sound like. Using Suno purely as an AI producer, giving me ideas what I could do with my already exisiting music 🙂

So copying would be great for my purposes. If that would work, that is

[random pine]

Mind you, I recorded guide vocals over an instrumental and was able to make a persona of an acapella of my own recorded vocals. Still Suno is like: ''Nope, not happening!''

[wispy laurel]

There's an odd workaround for this, unless they removed the function very recently
You can alter the speed of a song, and when doing this you can choose whether or not to also alter the pitch
So what you can do, is split the vocal and instrumental stems, slow the vocal down without pitch shift, then speed it back up to the original tempo with pitch shift. Result being that it ends up at the same speed that it originally was, but at a higher pitch
Then download the modified vocal stem and the instrumental stem(s) and combine them in the DAW of your choice
If you want to do this for just part of your song, just crop out that bit of the vocal stem, do the slow-down-speed-up just on that cropped section, then again use an offline DAW to edit that into the song

[jagged oar]

Mine works fine. Not sure where you are running into issues

[wispy laurel]

Best way to do that, that I can think of at least, would be upload a recording of only your vocal, then cover it with a very clear and detailed style instruction and high Audio influence. It won't exactly match your voice but it should at least be fairly close (assuming you're a reasonably good singer - if you suck, Suno will probably "improve" your vocal whether you want it to or not 😂 )

[random pine]

That would work for new songs I think! However I want to cover songs since I'm uploading my own music in there

[wispy laurel]

That'll work too. If the Audio influence is high, and as long as the lyrics (including any prompting in that section) and the style instruction are accurate to what your uploaded song actually sounds like, it can produce a fairly close imitation of it
The voice is usually the hardest thing to get right, but if you're just using it as a guide for your own performance anyway then it won't matter if it's not exact

[random pine]

exactly, gonna try, Thanks!

and otherwise I need to learn how to fix it in my DAW. Without singing it since I an a singer and not an audio engineer (which gives me little hope to think I can fix this in my DAW lol)

[quartz adder]

that's crazy. when I use a persona on 4.5 it will not even match the GENDER

you are breaking my brain

I don't understand what the disconnect could be lol

[rocky musk]

chat question, is this how you use the lyrics box to put in the styles?

doing it so that i can purely use audio quality tags in the actual styles box

but i've never done it so i dont know what im doing

[thorny cypress]

I would put all that stuff in the style box. And in the lyrics box just put tags on the segment of the song you want them to appear on. For instance things like [epic guitar solo] or [dramatic orchestra strings] or [fast beat]

For instance this is part of one of my songs, which is instrumental but has a few choir notes so that it won't try to throw in a random voice singing gibberish.

[wispy laurel]

I get a bit more detailed than that but not much. And still broken down by each part of the song, rather than all lumped together

[harp, upright bass, drums]
[Dark and unsettling] [minor key]
[50% acoustic blues, 25% folk, 25% acoustic rock]

[Instrumental intro - harp, drums, upright bass]
[Bass - melodic bassline]
[Harp - fingerpicked melody in minor key] [complex melodic solo, improvised blues-style hooks] [dark and unsettling]

[Verse]
[low, gentle, relaxed, soft rap] [low, smooth, sexy voice] [smooth, steady flow] [clean, clear vocal] [low, smooth voice]
Late in the eleventh hour
I clocked a face atop the blessed tower
And I paused amidst the shifting sands
Etc etc

[rocky musk]

ohh so thats how you do it

[umbral osprey]

So i made a song were i really really like the hook. But i want to changes the verses around it. Is there a way to use the exact same hook?

[storm pawn]

How much changes in the verse?

Lyrics or completely new melody?

[umbral osprey]

almost everything, I want the same sort of melody, just all the verses are different except for the hook

[storm pawn]

Did you write the lyrics?

If you use your lyrics, then if you put lyrics in, with new verses, the same hook, and use cover section it is very likely to do the same hook with different lyrics.

If the melody/tempo of the verses should be quite the same, you can go quite high with audio influence slider, otherwise go down. Even if you do 5% audio influence, it will still mostly use the same melodies

Weirdness slider 3%-5% and trying audio with different percentages (I usually try 5%, 25%, 65%, 90%) - this is how I approach changing full verses or full hooks

[umbral osprey]

Aight let me try that.

yea its my own lyrics

[storm pawn]

Another option is to get the stems, remove the vocal part where you don't like it, upload it as an audio track and cover it - it should fill in the empty vocal space with new one - I'm testing something like that right now, getting drums, melody & bass, verses, hook and intro/outro from 3 different tracks

[polar spade]

@unkempt wolf. question about lost files in Library

Good that you added to a Playlist so you have them. ✅

If you wish, go to edit UI make an extension at the end, then recrop to save the song without the extension (one way to make a copy), and save file so you will now have a copy that you can re-publish again.

Sorry that its un-found. But again, if its in showcase AND in your Playlist (which is just a placekeeper) you're looking at the file. It exists. Why its not in the Library OR the workspace list after clearing the filter from all checkboxes, and you say youve scrolled through the trash bin a few pages as well -- i have no idea..

If it pops back up, let us know what you did to find it, if it seems unusual. Thanks.

[umbral osprey]

ahhhhhhh that worked thanks bro!

https://suno.com/s/L5wek6qvwQzRuw78

[whole tiger]

Question? Why would I All of a sudden get dropped down from having 8 min to record a song to 2 min? I have the pro plan.

[storm pawn]

Happy to hear! Also the song is fire!! Suno is so much more creative with American rap than non-English ones 🥲 (probably way more quality input for training). If you’re going to release it definitely lands on my Spotify playlist

[umbral osprey]

yea im thinking about it, but idk how the whole releasing thing works bruh

[bleak hawk]

https://suno.com/s/Zc4oG5TSEVuTAeqW
So I've made this absolute banger of a song however, I can't figure out how to get the final guitar solo to last even longer. It goes for about a minute then returns back to vocals and repeats. Any ideas?

[wispy laurel]

Extend from one or two bars before the vocal comes back in.
For the extension, delete EVERYTHING from the lyrics and style prompts, and write this in the lyrics box:-

[Instrumental outro]
[Long guitar solo]
...
...
[Ending]

Do NOT switch to instrumental mode.

You can add more [tags] right after [long guitar solo] if you want to steer it in a certain direction, if you don't it should theoretically carry on in the same style as it is now.

[cyan light]

Not sure about the replacing the vocals and all but its entirely possible to cover any Suno track 100pct but you need to have permission's from the original artist first.

[short idol]

Need some help. I've been inputting my own lyrics into suno. On the Outro, it keeps changing the lyrics I've typed into something different. It's there any way to fix this

[brazen sand]

im desperate so im thankful and willing to try anything lol easy enough to try too so ill give it a go

[brazen sand]

Now thats some next level stuff right there... so i split the vocal stem and download it from suno and then I gotta do the speed up and pitch shifting outside of suno right?

[wispy laurel]

You should be able to do the speed/pitch change inside Suno
Should be an Adjust Speed function in the Remix/Cover sub menu
Although you could also do it externally if you're more comfortable doing it in a DAW

If you only want it for part of the stem then you could do the cropping and the pitch shift in Suno but you'd have to go elsewhere to edit the vocal track back together with the altered section

[humble verge]

Lyrics box is broken

[spare hazel]

what happened with the new ui update there are a lot of buttons missing some songs do not have the publish button any more

[coral plank]

why can’t i upload my saved prompts anymore

[spare hazel]

what happened with the new ui update there are a lot of buttons missing some songs do not have the publish button any more

[tawdry scarab]

How are you going to call a pure sine wave robotic

Everyone knows robots speak square wave

[slow crag]

https://suno.com/s/wp9RrGYcSGHx7qOQ this is amazing making a song like this made the vocals from scratch suno is amazing

[tidal sequoia]

Need help with a problem. Uploaded lyrics for a song. For some reason, Suno decided it would be good as a duet even though did not request it that way. So some of lyrics are female singer and some are male singer. Though I kind of liked it, I wanted to hear as just a female singer. Have tried multiple times to do cover by specifying female vocals in multiple ways (just saying female vocals, saying female vocals and no male vocals, saying female vocals and excluding male vocals). No matter what I do, Suno insists on making it a duet with female and male vocals on different lines or verses (changing them up at times). It has even given me duet with multiple female voices trading out, but also including a male in there somewhere on some lines as well. Any suggestions?

[storm pawn]

Have you tried clicking the female vocal
in settings, specifying it in style as well as in the lyrics itself?

[cursive tartan]

Try again with a lower level of "weirdness".

[graceful condor]

What I noticed is that the more you add it gets weird, faulty and could cause more issues than any good. I tend to do a lot of really detailed and complex prompts but I get the feeling in many (not all) cases it has more impact on interfering and disturbing the song than doing any good. Like for example if I use a Persona it makes not much sense to describe the vocals too. Because it somehow interferes in the end. So less I guess is sometimes more. At least at this stage and time of Suno

[umbral osprey]

Is it possible to get like one sentence made by a girl

[frail gazelle]

why did they mess with my create page, It wont give me options to Use AI to help with songwritings

[indigo gazelle]

hello guys, im still getting echo, reverb, autotune: https://suno.com/s/ZdQ7zbmhdf4h22rQ , prompts: Latin urban dembow/R&B, 95 BPM, warm synth pads, clean guitars, nocturnal vibe, negative prompts: echo, reverb, autotune

tried with 4,5 and 4,5+

can anyone help?

also tried : dry signal chain

as prompt

suno is always making echo to the vocals

[half dragon]

Click that magic wand icon on Lyric box to see the AI tool to help with writing lyric.

[polar spade]

If its less than a 5 minute song, you might Cover it with just [female vocal] in the style prompt box as well as clicking "female" ..... and see what happens.

Or, Not sure of your genre, but if more than about 5 style prompts, put "storyteller" at the end to give it a "solo vocalist" influence. If something changes, move it further left in the list to increase its weight and test again.

You could also try Covering with the lyrics all together as one paragraph. Ive never tried that, but since I thought it, ill suggest it. 🤣 but I will experiment as well with that idea. Might just get a rap, but I feel maybe not. I feel this would un-signal a change in vocalists to Suno.

I feel the lyric structure and general content of the lyrics are indicating that its a duet piece to Suno then. To get Suno to believe this is just one guy singing in the mirror might take some tweaking the arrangement [spacing, interjections, quotation marks, combining lines] .

[polar spade]

I call a 40 year old woman , "girl". Thats just me. So what do you mean by "girl"? I try not to assume.

[umbral osprey]

just a woman voice hahahhaha

[polar spade]

Using [female vocal] before the line is a good start, along perhaps with putting that one line in quotes , separate from the others.

[Verse]
This is the way
Life has begun

[Female vocal] "I know it has to be this way"

We'll be as one

Because of the 4 lines and rhyming scheme, hopefully, from my experience, it still sings it as one fluid verse.

[storm pawn]

Do you have any ideas on how to create a medium-deep female vocal? I'm trying to have a female singer singing to F4, E4, D#4 chords, I want the voice to be deep and calm, but Suno always gives me high pitch female voice 🙁

using "Contralto" in styles worked in second try 🙂

Worked in new song, but doing a cover of instrumental only still makes high pitch voice, regardless of style instruction 🙁

[lean crest]

where are the saved prompts after the interface update?
I found a similar button in "styles" but it only fills in the style, without putting the song lyrics (it's as if it could only get that style information from the saved prompts)

[smoky flint]

Can you train the vocals to sound like yourself?

[storm pawn]

with 3rd party software, like Applio

[umbral osprey]

It worked thanks bro

[polar spade]

Oh. Cool. Hazzah!

[umbral osprey]

Sending results in a bit

[boreal cloak]

Hi there

[polar spade]

Might not be perfect for every time. All depends on the lyrics. Glad it worked this time.

[boreal cloak]

I have some trouble making song in Bahasa, since v.4 until v.4.5+ the problem still the sama thing. It's about the pronunciation

[polar spade]

"Caberet" as a prompt has helped get me lower voices. Obviously, depends on placement in the prompt list or even the metatag, style of song, and the chance the genre may get too caberet-ish.

[storm pawn]

I'll give it a try, thanks!

[boreal cloak]

The way it speak words with C sound like K. And words with E supposed be read like é. And with this bugs it will change the meaning of the words.

[umbral osprey]

do u think there is a way to make it really high pitched

[polar spade]

Im finding my older saved prompts are showing. Any new prompts are hit and miss whether they get saved or not. So a bug they need working on, I suppose.

In the past we had all three lyrics, prompts, and excludes save which was a godsend for me. We could even mix and match which was great. But maybe a drag on Suno to keep storing lyrics. I get it. So now just prompts and maybe excludes.

Lyrics will just have to be copy/pasted again from older outputs. If so, they need to bring back that lyrics' words are searchable. Thats on hiatus as well, it seems.

When I click on an old prompts save that had lyrics as well, it only pasted in style prompts only. I assume lyrics are now on hiatus or possibly deleted.

I dont know what the three dots to the right do though. So far , nothing for me, on Android, at least.

[polar spade]

Like an old witch? A Manga girl? Anime? Vocaloid?

[umbral osprey]

Uhhhh Just a girl. Its a song about a shooting game and a girl asking a question

https://suno.com/song/f07c334c-e220-4e98-aac6-e97baee88d5b

[polar spade]

So "girl" means a child or kid? Someone who sounds less than about 14. In choruses, a children's choir, for instance sound professional and spot on. A kids song, might sound more like 7 year olds singing.

To get a kids voice, especially a solo, from my trials Ive never landed it on one try, and only get kids singing or rarely talking if the song itself really leans towards an obvious kids song. (I write children songs occassionally)

Once I got that, I collected snippets of all the kids voices and made a persona. So I now use that persona for kids songs.

So far. I have no success to consistently use a tag or prompt to get a kids voice. Mostly trial and error. Other users might need to pitch in on that.

[indigo gazelle]

hello guys, im still getting echo, reverb, autotune: https://suno.com/s/ZdQ7zbmhdf4h22rQ , prompts: Latin urban dembow/R&B, 95 BPM, warm synth pads, clean guitars, nocturnal vibe, negative prompts: echo, reverb, autotune
Suno
suno1 (Cover)
Listen and make your own on Suno.
Bild
tried with 4,5 and 4,5+
can anyone help?
also tried : dry signal chain
as prompt
suno is always making echo to the vocals

[polar spade]

From your lyrics, you might just want to try "anime" as a metatag, to get a high pitched voice. That may work or it could end up having other lines sang that way as well.

[umbral osprey]

i can try

[polar spade]

Just saying that anime is definitely a known vocal quality that users use often.

[umbral osprey]

Its almost perfect haha, thats the only thing i dont like currently

damn im almost out of credits lol. ive wasted around 3k on this song

[tranquil leaf]

Wow... been a while since I used it, but they really arsed Soundcloud. Commercials after every song? 😛

Sorry, wrong channel.

[polar spade]

I hear ya. To better separate out that one line, if youve used the Edit UI and are comfortable, just replace the one line with the SAME one line but add the metatag of

[female vocal] [ anime] (gasp)

....before the line. You may wish to leave out "gasp" thats optional.

[tranquil leaf]

I blew about that much on my favorite song too... and still ended up manually editing an older generation because I could never get it to go back to that. 🙂

[polar spade]

No worries. Have a nice day. 🙂

[umbral osprey]

Yea i dont understand the editor haha

[polar spade]

Try it in the lyrics box then... I think it won't work, honestly. But one go and see, eh.

[umbral osprey]

haha i can try, i just keep generating new songs

So i made one now, That Line is PERFECT. Is it possible to get only that line and place it in the other song?

[polar spade]

Yep. I have a Suno endowment fund to buy all my credits now. Hahaha.

[polar spade]

Not yet, we cant patch one song into another. Having said that, theres this Suno Studio coming soon that someone may have access to as a trial. It is supposed to have DAW like features.

Otherwise, you download both and patch together in an audio workstation on your own, for now.

[polar spade]

Good news is you're making momentum. So thats something.

[polar spade]

Stick a copy of the line at the end of your song. Seen if it does it there again. If so, learn the edit UI a bit if you can, so you can cut and paste the line where you want it.

[umbral osprey]

aight ill see what i can do

Thanks man!

I think this is it

https://suno.com/song/c9775725-f043-48ec-a6d3-134b5dfed3ec

[bleak hawk]

Okay, so there's something bugging me.

How exactly am I supposed to make a 7+ minute song when I add songs for inspiration above? Adding any songs for inspiration seems hardcoded to always make the song less than 5 minutes regardless of anything I do. And yes, I am using the 4.5+ pro module. It's really blocking my ability to further improve my songs.

Anyone know how to fix this?

[tranquil leaf]

I've run into that with anything but a fresh creation. Inspo, Cover, Remix... they always come out <5m no matter how long the original was.

[flint jetty]

New to Suno. My cousin and I have been talking about taking this video clip and turning it to a full length version of the song using his voice from the clip. can I do this with Suno and how would I do that?? Would be eternally grateful if I could as we think his voice is the best of anyone thats sung the song!! https://www.youtube.com/watch?v=01AySy3H39M

[spark pasture]

If I want to contact support, where do I go?

[polar spade]

You can not use previously published or publicly available audio or text. Typically, any attempt, if censored by the auto-mod, will give an error such as "blocked due to copyrighted material"

[agile imp]

Is it temporary issue or done in purpose that no more refresh for genres ?

[polar spade]

I posted a little of my opinion regarding the newer access to saved prompts.
#📚┃suno-school message

Since it is fairly new change to the prompts save feature, it may need some review by the developers to work out what is the best means to provide saved prompts. Right now, its seems a little glitchy to me. So, Im just keeping an eye on it and.will give feedback in the bug channel and appropriate channels.

You can post in the bugs channel or suggestion channel for you ideas as well.

[agile imp]

thank You very much for quick answer
Lagging since last few days is another issue
Works brilliant before update but now...just flustraiting

[polar spade]

Oh... as for your question... as for genres.... it seems you can just scroll them to the left to find more suggestions. Looks like it stops at about 20 suggestions with no refresh.

[spark pasture]

How can I solve this problem?

[polar spade]

Also..... I can scroll through the suggested prompts with my finger on full screen, but it won't scroll if I zoom in.

[agile imp]

yes, that what I was asking for 😉
that limit of 20 is not good at all because many ppl like me (new on this) are struggling with writing prompts for it and multiple and refreshing choices were just very helpful. It's sad that there's no more refresh option available 🙁

[agile imp]

I'm working on PC so it's just limited as You mentioned

[polar spade]

https://discord.com/channels/1069381916492562582/1154587937611190354

If enough mention in one of the suggestion channels, it would help them to consider bringing it back, im sure.

[agile imp]

hope so because mainly I bough PRO because of this where other tolls didn't have that many choices and now...I don't know

[boreal cloak]

Hi, how to close this tab? It's really disturbing I can't read clearly the minute for the last track

[half dragon]

I don't think it's possible to close the play panel.

[boreal cloak]

This afternoon I was giving song title without it. The play panel not there if I didn't click play. Now after click play one time it's there forever🥲 even when I close my Google chrome and started it all over again it's still there

[storm pawn]

What's the most effective way to have both male and female singing a chorus? I've tried [Duet], but the results are far from the random duets I got in the past.

[agile imp]

I'm using [Choir] without marked Male or Female voice and it's works on 80% 😉

[storm pawn]

well, never thought about using this one except for some gospel songs 😄 thanks! I'll give it a try

[coral plank]

where is the auto lyrics option ?? it’s not there anymore

[half dragon]

Did you mean the tool to generate Lyric for you? or like in Simple mode, you describe the song, and it'd generate the song with lyric for you?
Both are still there I believe.

[storm pawn]

geez I just realized that I generated 220 versions of a song to get the choir the way I like and still can't find the sweetspot 🥲

[coral plank]

the section where u can describe the lyrics and it generates it like it ask you to describe a song not the lyrics and it automatically genrates the lyrics about what you want the song about

[half dragon]

If this is what you meant, they moved it under that magic wand button on lyric box.

[coral plank]

this section

[half dragon]

Yeah they changed that with the new look in my SS above.

[wide pond]

Whta have you guys done to my Suno? LOL

[humble verge]

i wonder why SUNO still can't get percussion and drums right

[gilded mortar]

I feel like the last update of suno messed up things

[agile imp]

Agree

[humble verge]

The sound degradation from 4.5 is back. They had it (almost) fixed, but I'm getting a lot of bad quality results today. Instruments sounding like they're inside a cookie tin. Drums sounding like cardboard.

[agile imp]

True ...I've stopped to do anything (saving credits) with hope of fixing so many issues after last update

[humble verge]

it's annoying they don't have direct bug reporting like Midjourney does. Their Discord bug reporting is for platforms, not the AI itself. The other option is some standard feedback thing nobody reads.

A bug I'm experiencing recently is SUNO only generating one song instead of two at a time.

[jade quail]

There used to be a toggle for "instrumental" - why is it gone now?

[half dragon]

Its still there, just looks different.

[jade quail]

I see. Thanks. The unselected state is almost invisible, but that’s dark mode for ya.

[scenic coyote]

what settings do you use if you want to cover a song and change the lyrics. I find putting weirdness to like 10% and audio influence to like 80% works sometimes but it's not really consistent. anyone have tips

[scenic coyote]

I feel like it has to do with server load or something. I swear to God I'll have streaks of brilliance with this thing and then garbage for a while. it seems like it has less to do with the model or parameters you put in and more to do with random times of the day but I could be completely wrong

[jovial yew]

I had some success with changing the displayed lyrics and using remaster... I think

But I only changed one word

[scenic coyote]

yeah that works sometimes for me. thanks for the tip. I ended up doing zero weirdness and 100% audio and it worked pretty good. not exact but meh

[untold briar]

Remember you don't necessarily need to do it all in one step, you can Cover a Cover, and so on.

[scenic coyote]

oh don't I know .... last song was "Follow Thru (Edit) (Edit) (Remastered x4) (Cover) (Cover) (Extend)"

😂😂

[ionic turret]

https://suno.com/s/V9YxOml47g3Cdm9H 🤘 just need to edit the vocals a bit more 🙂

[severe kindle]

why they changed v4 without notifying people?

its bad now...

[raven turtle]

I’ve noticed this too. Getting that washy shimmer sound on synth pads and cymbals

[coral plank]

i hate the new layout too many options i used are now gone i’m def canceling my subscription

[agile imp]

I was thinking this same, there's more other music generators 🤷‍♂️

[gilded mortar]

You are absolutely right. The pervious one was better and easier to use

[ebon sand]

Which options are you missing? Just curious

[bronze helm]

Wait which options are gone?

[jade mist]

Does anyone here have any information if Suno is working on a new model? For example V5 or similar? I mean V4.5 plus is already out of beta so...

[graceful grove]

So I can't get it to give me tubular bells no matter what I do. I have actually tried only putting "tubular bells" as the only things in the style prompt. Here's another example of a style prompt I tried:

Tubular bells ring throughout the track, playing different notes to form their own constantly present melodic riff that drives the music and interacts with the other instruments, adding a chilling, funereal atmosphere. Distorted electric guitars soar with complex high-register riffs, cutting through the mix with intensity and tension. Drums are pounding and martial, with a steady, militaristic pulse that propels the music forward. Bass is thick and forceful, anchoring the low end with weight and menace. Vocals are raw, commanding, and defiant—gritty or shouted—channeling rage, pride, and refusal to break. The arrangement layers intensity through repetition, while the interplay of guitars and bells adds subtle, ever-present complexity. Tone is foreboding, proud, and tragic, focused on defiance in the face of doom.

[bronze helm]

This is nothing to go by, but check #🗣️┃chill-chat message

I have this bizarre... bug? Idk. Others can see it too.

[stuck junco]

Hey there! Thanks for sharing this! Lets see what we can figure out!

So with song guidance using bracketed information helps a ton. Let me point you to a thread that provides some visual guidance

https://discord.com/channels/1069381916492562582/1410014597338894396

What you'll see here is a structured approach from within the lyrical field

For me (and this isn't the only way, lots of room to discover and have fun exploring), I keep the style of music field pretty simple and tailored toward the sound I'm looking for. Your case it may be something like Epic Doom Metal / Post-Metal / Tubular Bell Ambience

[orchid galleon]

How do i find my saved prompts??

[stuck junco]

The lyrical field though I would add a lot of specifc guidance prior to the sections

[stuck junco]

This is a reported bug that's being looked into currently. The bottom left button in this image is where it's supposed to be though. I'd try there to see if the bug's been cleared up

[orchid galleon]

yea so I do click that and it only brings up styles

omg please god dont tell me they lost all the saved lyrics I had the best friggin song written

[agile imp]

I've just tried to do one song (typing my own Styles since is not possible anymore) and sound just awful 🤮
Music jumps every 10 sec separately into different speakers, by the half of the song volume just pump it up like + 20dB, just terrible

[graceful condor]

Had the same experience. Like if someone is messing with the soundcables at a concert. Losing sound, jumping back and forth, hiccups, then dull, then one speaker only or the other. It gets worse every day.

[tepid nebula]

Hi everyone! 👋
I have a question about using Suno AI for songwriting. Suppose I have a melody in my head that I’ve recorded — just humming or singing nonsense words — and then I’ve written some lyrics that fit more or less with that melody. Is there a way to feed both (the melody and the lyrics) into Suno so it can generate a voice based on that? If yes, could someone explain how or share any tips/resources? Thanks a lot!

[fair adder]

hello everyone, i'm a little stressed right now. i just opened suno but it seems the interface has completely changed, and mostly buttons are not labellized anymore. as a blind user, i struggle to find my marks again and i find this situation very frustrating. i don't really know where to put this since it's technically not a bug so... it ends up here.

[tranquil leaf]

May not be a bug, but still well worth a report.

[severe kindle]

Has anyone else noticed the vocal quality changing in older models when a new one gets released? I used to get really good vocals on v4, but once 4.5 came out, I couldn’t achieve the same style anymore. It feels like v4 isn’t running on the original algorithm. For a specific genre I work with, v4 used to give me great vocals if I generated 20 songs, at least 7 would have the style I was looking for. Now, even if I generate 100 songs on v4, none of them come close to that vocal style. I’m convinced something was removed from the vocal dataset on Suno, because those particular vocals just aren’t there anymore.

[tranquil leaf]

Wouldn't surprise me if they deprioritize older versions, so they're simply not getting the processing time... but I have no way of proving any such thing, short of the quality of outcome.

[patent coral]

Hello Suno Support Team,

I recently reinstalled Windows on my computer. After installing the system again and logging in to my Suno account with my subscription, I noticed that all my previously generated tracks (“branches”) are missing from my account.

[lost holly]

You can upload a version of you humming or record it with the + audio option on the top left. Then you can cover the uploaded/recorded version and add the lyrics you are thinking off and hope for the best.

[lost holly]

You can submit this as feedback under the following link: https://suno.com/?open_feedback=true
Also feel free to write here anytime, if you need help finding/locating any options.

[fair adder]

thanks a lot. but i don't want to bother anyone, especially just to find a simple option. my biggest issues are that i can't find the auto option for lyrics anymore and i can't adjust the sliders

however the new add vocals options sounds promising

[fathom cedar]

Hey there Suno team

I was wondering, could you include some of the generation data in a COMM field within the ID3v2 of the mp3? It would be really useful for me to just inspect what the query was for a specific Mp3 I've generated.

Comfy UI for instance does something similar encoding the whole flow within the generated png metadata.

Thanks!

[neat nexus]

#💡┃suggestions-web

[fathom cedar]

Thanks!

[neat nexus]

Not really the best place but we don't have a suggestions-general for something like this. I like the idea though. So please post it in there because I have been assured by the Team more than once that they do read those channels.

[fathom cedar]

I did, sorry for putting it here, I read the introduction texts and they each mentioned some channels for support and stuff and finally, this one. True, could have been more thorough in scrolling through the channels. 😄

[humble verge]

I use SUNO at all hours. I know some say it works better for them at certain hours than others. And I suppose it makes sense that server loads have bearing on quality, though I am no expert on AI. What I can say is that the problems with rapid deterioration of sound quality seemed almost fixed with 4.5+, but now it's back and it's even worse than it was in 4.5.

I contacted support, but their reply was less than satisfactory - claiming it was an issue with the prompting or an issue with browser cache, yadda yadda. It's not. The only thing that changed between when sound quality was fine and when it was no longer fine, is the recent update.

Don't count on this issue being fixed anytime soon, when SUNO won't even admit that it's a problem with the model. I really wish I could just pause my subscription until the next major update, because right now it's a complete waste of credits to do anything.

[desert vector]

Has anyone successfully found a way to keep your song in the EXACT tempo and KEY?

I'll include to have the tempo at 130BPM and A minor for example, but then it isn't.

So I'm wondering if there's a good way to really keep that in check?

[viral pond]

geees.. imagine that. deleted the wrong song.... deleted the version i was supposed to keep, hope suno can restore it even tho i wiped it from the trash as well

i can still open the file from url from other browsers so it seems it still available

[wispy laurel]

Give me the URL, if I find it and comment on it you should get a notification
Then you might hopefully be able to follow that to find the song on your own account

[viral pond]

https://suno.com/song/a034c28c-3283-41b4-945b-bf564977ad57

i do have the original downloaded so in worst case i'll reupload it

but i loose all the plays and likes 🙂

[wispy laurel]

Well that's weird
I can find the song, I can play the song, but when I try to comment on it I get an error saying Not Found

I can click the three dots though and it's offering me "Restore to Library"

[viral pond]

yeah

so it is on the server, they might add an redemption period to it before it is deleted maybe, i sent email to support, but not sure if any of the teams are on this server if that could speed things up

[wispy laurel]

Weirdly I can click Restore to Library, even though it was never in my library
Clicking it doesn't do anything though, not even an error message

[viral pond]

yeah too weird

this is the bad thing about adhd.. not paying attention, 😄

[scenic coyote]

can't you set the tempo in the new editor?

[quick sonnet]

Is there some good way to prevent verse and chorus from overlaping, when sung?

[storm pawn]

I often have this issue… the last word from a verse is getting overlapped by a chorus

[quick sonnet]

Exactly my problem.

[mystic flume]

Good as in reliable? No. Extra lines help, and sometimes just putting a period or elipsis on one of the empty lines will remind Suno to take a breath. But as always, the only reliable way to get exactly what you want is to keep trying. The same lyrics will get different results with each gen, without you having to adjust anything. However, if it keeps happening, adding something between the verse and chorus will usually help. Even if it's just a "Whoa-oh!!" or whatever.

[storm pawn]

I recently had a song that was too good to scrap, so I decided to fix it outside of Suno this way 😁: https://suno.com/song/d7cc0cf1-3b47-46c1-a743-4fe1937204eb

[quick sonnet]

I got 5 times the same result.

[storm pawn]

But it’s really annoying since most of my songs had this issue, and I had to regenerate the section, take the vocal stem and replace it in DAW

[storm pawn]

It seems that sometime it is that much hardcoded for a given version of the song, that every cover follows the same pattern

[wispy laurel]

[verse]
Lyrics here
And here
And here
And probably also here
...
...
[Chorus]
More lyrics here

The ...s help a lot
Having multiple lines of ... (rather than just empty lines) also seems to help

[mystic flume]

Haven't seen much concrete info on what "Heal Edits" does, but I have noticed if you highlight the actual edit transition and click the "heal edits" option, it does seem to make a slight modification to the waveform (and a noticeable improvement).

[storm pawn]

That’s good to know! I’ll try it on some songs that need fixing

[mystic flume]

Before:

After:

[storm pawn]

@wispy laurel is there any instruction to put accent on specific part of the word? Sometimes Suno puts accent in a complete nonsense place and every cover of such songs follow the same pattern 🙁

[storm pawn]

I have some songs having weird background noises in drum/other stem, I’m curious if it can fix these

[mystic flume]

Best practice if you want specific syllable pronunciation is to use dashes between syllables. This can mess up the actual pronunciation though so sometimes you gotta change spelling too.

I couldn't get it to sing "reprobate" for the life of me, until I did "rep-ro-bate" for instance.

[wispy laurel]

Capitalise that syllable
And possibly also separate syllables by hyphens as well as capitalising the one that's supposed to be emphasized

[sleek stump]

Protip: Never rely 100% on online service to keep your stuff safe. Always copy what ever you find important to a document on your device.

[storm pawn]

But if it accents “ro” in reprobate, how to make it accent “rep”?

[wispy laurel]

REP-ro-bate

[storm pawn]

Yeah just noticed your reply 🙂 I’ll give it a go

[mystic flume]

A couple ways: First, the best way is to always take the gen you like best and then do a cover of it with a more specific style prompt. Often the high-hat cymbals will irritate me, and I put "high hat cymbals" in the exclude which seems to help.

If you insist on using the stems, then I'd download them, then use another service to clean them. Often you can just separate a given stem to get the noise out. if you use Suno to create Stems of Stems, it might work, but usually because of how they do their stems (basically covers of the existing) it can deviate from the original enough that they don't match anymore.

[mystic flume]

In that case, I'd do "rep-robate" and see what happens. If that doesn't work, I'd try "REPP ro-bate" etc.

It's usually a matter of trial and error. Just yesterday I was having issues with an old song that I was trying to remaster using covers, and Suno just will not reliably pronounce "cartographer" for me. I tried a half-dozen different ways of spelling / structuring it, and it ended up just being randomness that got me what I wanted. I left the word "cartographer" and just kept hitting create until it said the word right.

[mystic flume]

Another approach is to re-examine the entire line. Can you rephrase it to better match the specific meter the model seems to be expecting? That can be a lot easier than trying to force it to do what you want, and it can often indicate a problem with how the lyrics were originally written. What we hear in our heads when we write a poem is not always the way Suno will interpret the words themselves when they become lyrics.

[tawdry scarab]

has anyone trying to stress certain parts of words just used the accent diacritic

i would not be surprised if it didn't work but from a typesetting standpoint there is a diacritic (modifier to a character) for that purpose and it's just called an accent mark

[mystic flume]

I have tried, and no it doesn't seem to work.

I've also tried using the international phonetic alphabet, which Suno doesn't seem to understand.

[daring tangle]

why does my prompt keep getting deleted? 🙁

[mystic flume]

Are you adding a persona? If so, it'll delete what's in the style prompt and replace it with whatever that persona's prompt was. It's a really irritating system and I just use copy and paste because the "undo" button is unreliable and the "save prompt" bookmarks don't seem to work very well at the moment.

[daring tangle]

nah I posted a prompt in here and it was deleted from the discord, it doesn't break any rules so idk why someone keeps deleting it

[mystic flume]

There was a message from Woodhouse about using too many capital letters.

[daring tangle]

ahh i can fix it to not have caps mb

[worthy igloo]

is the download option on it? download / upload

[small dirge]

How can I get SUNO to distinguish between Bōw and Bŏw? It keeps calling everytbing a bŏw.

[viral pond]

no, but i did download it before this happen tho

[humble verge]

70 songs in a row - including remasters and covers - with rapidly deteriorating sound quality after the first minute or two. This last update completely broke SUNO. They had this bug almost fixed, and now it's worse than ever. I wish I could pause my membership until they get around to fixing this thing, but that would require for them to acknowledge there is a bug in the first place and they don't seem particularly interested in doing that.

[desert vector]

i havent seen that can u screenshot?

[scenic coyote]

I'm not seeing that today. are you using a certain persona or something?

[scenic coyote]

It's on the bottom bar of the new editor. you can change it from fixed to dynamic tempo and set the tempo. that's at least how it worked last I checked

right side of the bottom black bar

[humble verge]

Nope. And I doubt most people even notice it unless they listen to the song for a minute and then jump straight to the last third. The ear tends to follow along when the change is gradual, but if you compare the sound of the instruments from the first third to the last third, it becomes glaringly obvious. I didn’t even catch it myself until late last evening after saving a ton of songs I had done. Listening today, I can hear it in all of them. It’s consistent - especially with drums and percussion. They go from sounding fine to sounding like someone drumming on a cardboard box in a bathroom by the end. It's super depressing, because I made some great blues songs last night and they're all ruined.

I wrote SUNO support and their reply was basically "have you tried turning it off and on again?"

This bug was big in 4.5, but they more or less fixed it with 4.5+. Now, over the past 48+ hours or so, it's back in full force, and worse than it was before.

[wispy laurel]

I haven't so much noticed a drop in quality per se, audio clarity is fine for me, volume and mixing definitely inconsistent in the second half of every song today though
Guitars/whatever else is playing lead, starts out loud and gets louder as the song goes on
Everything else either steady volume or dropping as it goes. And mixed really weirdly sometimes too

[lusty juniper]

the create section still buggy. has that delay when typing in the Style section.

[humble verge]

The clarity is fine. It's the sound itself - the boxy or tinny texture that happens gradually. A high hat goes from sounding fine to sounding like someone dropping cutlery into a pot. A drum goes from sounding like a drum to sounding flat and too loud. With 4.5 it happened once in a while. 4.5+ basically fixed it. But something seriously broke with this last UI update, because it's been every single song since yesterday.

[scenic coyote]

ill see if i can hear

[desert vector]

on this page?

[graceful grove]

I tried this, and again, no tubular bells. It swapped the style. I tried Bracketed Guidance as well, but with no luck. I think it legit doesn't know what tubular bells are. I can't really justify using up any more credits to do more though. 🤷

[lucid drum]

Have you ever accidentally invoked a band? This

Absynthewave. Dark, psychedelic synthwave track. Distorted, acid-soaked TB-303 bass synth, distorted pads, a hypnotic, off-kilter beat, a feeling of nostalgic disorientation. eerie, distant whispers.

Somehow resulted in the Pet Shop Boys. No complaints though, I love them

[neat nexus]

Save the generation in your hallucinations folder and cover it with the lyrics you wanted. You'll be surprised at how good the results will be.

[lucid drum]

sorry wdym? Write custom lyrics for it? I haven't really looked much into covers, I've got a lot to learn!

[neat nexus]

Yeah, using the hamburger menu on the song you got select Remix\Edit -> Cover.

Then copy back in that style you posted above in the style prompt and put your original custom lyrics in the lyrics field and see what you get back.

[lucid drum]

shi, I'm out of creds today, but first thjing tomorrow! Is that a good hack to make songs even better?

[neat nexus]

It was for me.

Of course the one I got was pure hallucination. I have no idea what it was singing but it was beautiful. So I tried covering the hallucination in the same style with my lyrics and it came out just as good.

[half dragon]

I think they meant - In the editor - not create page
This page - at the bottom there.

[lucid drum]

Ah nice! I have a few songs I'd love to create variations on. Do you find you can use a previous song to make basically a new song by the same artist/ band?

[neat nexus]

With some work it should be possible.

Have to play with things and probably build a persona too so you can nudge it. And don’t sleep on the sliders for that!

[desert vector]

right, does that mean when i prompt an idea a song idea itll make it to the same tempo and key?

[marble cobalt]

Is there a place to view a definitve guide for 4.5 creation with tips and tricks

I couldn't find one and I was working on one myself.

[half dragon]

Some one shared this some where - might be helpful?
https://jackrighteous.com/en-us/pages/suno-ai-meta-tags-guide?srsltid=AfmBOorFdPD8IKB1IYQnXZtxZp099Ci6220BwRg-WW_T705kQ9fGY6UM

[scenic coyote]

there's a couple links in this thread that are really good. not sure if they are pinned but there's a couple documents people shared that I uploaded to a custom GPT and now it's a wizard

[scenic coyote]

right after I read this I don't know if it was luck or not but I just started getting underwater sounding generations. like clearly not subtle... normal prompts.. coming out like absolute doo doo.

very very very inconsistent results right now

[glacial wing]

Hey guys ! anyone has a clue how to fix that annoying hissing noise in 4.5 . 99 % of the track have this super annoying hissing sound in the background... I generated like 450 songs there is maybe 5 songs that don't have the background noise.

[scenic coyote]

in the exclude prompt type "lofi, lo-fi, hiss, noise, bad quality, background noise, static". can't promise it'll work but it's definitely 100% worked for me many times

[scenic coyote]

I have a persona that accidentally sounds exactly like Men at Work. It's so sick lol. it started from a guitar loop that I played/recorded directly into the browser recorder

[sullen marsh]

I am writing to express my disappointment with the recent changes. With the removal of an important tool and the fact that Inspo and Persona no longer work together, a core part of the creative experience has been lost.

This limitation severely impacts the creative workflow and takes away a feature that many users considered essential. Instead of an improvement, what remains is frustration.

I urge you to reconsider this decision, or at least provide a transparent explanation as to why this step was necessary. User-friendliness and creative freedom should be top priorities — but right now, it feels as though these needs are being ignored.

[worthy nexus]

Just want to second what pulation has said here.
Noticed this immediately today when i went to create a new song. Big disappointment!
Being able to use Persona and Inspiration together allowed me to use my real instrumentals as an inspiration playlist to give a perfectly consistent "band" and production sound to the generation while still having the original voice and overall style of the Persona.
Removing this pairing is a big downgrade in potential for making more creative and consistent generation.

[pastel plaza]

hi guys , why song covers gives only a 4:24 minuts out of the 8 minuts song ? and how do i fix it ?

[cursive jetty]

its a colossal downgrade the tool was enabling a much higher level of consistency 🫣

[rigid folio]

How can i redeem my credits for the happy contest?

[cursive jetty]

anyone know why im not getting the 'submit remix ' option when remixing the 'Perfect place' comp in 4.5+ pro

[soft gyro]

Hi! In the last UI we could choose between automatic lyrics (simple Ai generated) and "man made" (elaborate AI lyrics generator) lyrics.
This seems to be gone now. Am I right ?

[glass wagon]

Is it possible to make a persona off a persona? because sometimes the persona changes a fair bit and i want to use the new version of it

[west hare]

guys how do I do a duet?

any specific prompt?

[quick sonnet]

What exactly you mean by duet? Multiple voices at the same time, or diferent voices for diferent parts of the track?

[soft gyro]

for duet you write "duet" and "male and female voice" in the prompt, but it is not 100%. Another method, safer, it is to write "male voice" and "duet" in the prompt but set the switch to "female". Of course you can and also commands in the lyrics prompt

[hearty laurel]

Hey guys, Quick question. What is "Hooks" under the side home menu? I can't anything about it anywhere. thanx

[half dragon]

I'm not seeing this, post the screenshot?

[worthy nexus]

Hooks appear to be a mini platform for short form videos. When creating a hook you can upload a video, i think it said 5 secs - 2 mins long? And then you can attach one of your suno songs as the audio

[balmy hound]

Is there any way how to recreate exact same voice from the uploaded song with accent and everything? I have song in German and I made English lyrics and I just want to replace lyrics with English one keeping original song. Tried Extend feature but that just generated 2 second clips. Tried cover but it constantly generating completely different voices (v4.5 model). Created Stems from original song and from vocals I made Persona and attached it to my cover but that generates completely different song.

[half dragon]

I just see it now - it's like "shorts" for your songs.
#📢┃announcements message

[graceful condor]

couldn't agree more. There is always a "thank you for your feedback" but I do not see any improvments. It really got really bad lately.

It is the worst since I started using Suno. The first month was a blast but the new UI ruined all. And I do not get it why they do not roll back and test the new UI on test servers until it is 100% bulletproof.

[neat nexus]

Have you tried playing with the sliders to try to nudge it with your persona and cover? Try German accented English or vocalist with German accent in the style prompt to see if this nudges things further for you?

[coral plank]

the new suno layout sucks i’m def not renewing my subscription after this me and about 15 friends are not renewing our subscription after this month and i’ll look for another ai software to use suno is just not where it is anymore after that update

[graceful condor]

So true But Sadly there is nothing better right now. As far as I know.

[agile imp]

same here 😞

[coral plank]

that’s true i’ll rather save my money until it’s fixed or something better comes along i haven’t created since the update i just don’t have the urge anymore i was on it 24/7 now i just listen to my old stuff and log out

[raven turtle]

Same here. The old issues are still there baked into the models too. I’m downgrading until a decent update comes along or another platform comes up through the ranks

[hard beacon]

Well... see you 😉

[graceful condor]

Totally understand it but I invested too much time and effort, to build all this persona and sounds and materials and yeah this is the issue if some party has a
monopoly position. Still it is just sad and disappointing. I really hope it will improve somewhere soon. Because spending Premier plan is a lot of money for me.

[cursive jetty]

ive done this several times - and also adapted the personas along the way through the versions for that reason eg. the AIman persona and the Ailé person (the inspo feature was a great way of assisting but suno appear to have dropped it >,<)

[glass wagon]

Why can't i download any of my songs lol

[tranquil leaf]

There's a post on the subreddit right now about using emojis for defining individual voices in a song. Has anyone actually had any luck with this? The times I've tried it, the AI tries to sing the emojis and spouts gibberish anywhere they are in the prompt.

Just wondering if I'm missing some method of making the AI obey it properly.

[shadow pier]

I fear Inspo and Persona never worked together. You mind sending over a song you created with both and we can check on our side to see if both influences were actually used in the creation of the track?

[tribal fern]

Olá. Paguei 4 dolares e recebi 500 créditos, porem, não esta gerando minhas musicas, da erro. Qual o motivo?

[orchid notch]

any one knows how to make suno generates choir acapella? i managed to get it a few weeks ago but right now when i try to generate it only uses 1 voice and also adds music and drums that i don't want

[gaunt tundra]

is anyone else having issues creating hooks? i made a mini lyric video for the chorus of one my songs and i line it up perfectly with the song, but when i post it, it plays like 5 seconds earlier into the song than i wanted it to, if that makes sense. idk if im doing something wrong, its my first hook lol

[graceful condor]

Wth the background sounds really weird like if the whole musical track is a glitch as a whole. Never had that before but other issues.

[raw compass]

ok so maybe a dumb question, but I "saved" my lyrics, then accidental refreshed the page but clicking on something else and my lyrics are gone. But I did click the Save ribbon icon in the lyrics, so how do I get them back?

There use to be a + button that had your saved lyrics and styles

then just load them that way

[eternal dirge]

Anyone else have MASSIVE lag on Suno right now?

[neat nexus]

To ensure that all inquiries are addressed efficiently and without duplication of effort, we kindly request that you post your question in a single, most relevant channel.

This practice helps us to track conversations more effectively and provide a consolidated response. We appreciate your cooperation in keeping our communication channels organized.

[half dragon]

Supposed to be this button - but I found it's not working properly - I can't save more to it.
Style and lyric should be saved to that list - but it did not work for me.

[frigid willow]

I want to show off how I use Suno but my username is all over the site, is there a way to hide my username while I use SUNO like a streamer mode for example?

[neat nexus]

Thank you for this valuable suggestion. To ensure it is reviewed by the appropriate team, please post it in the #suggestions-layout channel.

[hollow trellis]

i was on github and looking for a copy of suno/bark with a yml file does anyone know were i can find a .yml for it?

[flint haven]

Beware: there is NO official Suno API

🚨 Suno does not have a public API. Any claims otherwise are unauthorized.

[shut ferry]

Best way to create long guitar solos. They are usually to short, I have tried using meta tags to dictate how many bars but usually gets ignored.

[neat nexus]

The effective method involves a combination of [Metatags] and the strategic use of spacing elements to structure the prompt.

Alternatively, you can also specify desired instrumental features, such as 'extended guitar solos,' directly within the style prompt.

Older example: [Removed links]

Another effective technique involves initiating an extension from a point prior to the track's conclusion to create a longer, integrated solo. However, this method has not been frequently utilized in recent workflows by me.

[shut ferry]

For me, extensions only work at the end of a song, whenever I try to add a section, that section ends up sounding very different from the rest of the song.

[shut ferry]

What does (...) do?

[wispy laurel]

Creates space or delay between lyrical sections
It tells the AI that there's something there, but that the something doesn't include any lyrics

[neat nexus]

A technique for creating instrumental breaks or extending musical passages involves using an ellipsis enclosed in parentheses (...) within the lyrics.

The model generally interprets parenthetical text as backing vocals, and this input can function as a temporal spacer. This often prompts the model to fill the extended duration with non-lyrical vocalizations or an instrumental section.

While its effectiveness can be context-dependent, this method provides a practical way to influence the song's structure and pacing. You may find it useful to experiment with this approach.

[shut ferry]

Thanks. learn something every day. Any tips on how to have less screaming in rock/metal songs?

[neat nexus]

One potential workflow is a two-step process. First, generate an initial track using "divorcecore" as the primary style prompt, specifying the musical details with [Metatags] in the lyrics. Once a suitable vocal performance is generated, use the "Cover" feature on that output. In this second stage, apply a more descriptive style prompt and make fine-tuning adjustments with the "Sliders" and "Persona" features.

Achieving a specific outcome often requires iterative experimentation and persistence, as there is an inherent element of randomness in the generation process.

[north sleet]

I kind of dig the new interface. It is more iPad friendly when using the browser. ++karma to Suno.

[undone lotus]

less screaming? in rock/metal?!

What kinda guitar solos are you referring to? like yngwie malmsteen arpeggio noodlecore or like, carlos santana?

[shut ferry]

Looking to create longer solos that increases/decreases in intensity. Also, every line doesn't have to end in a an inhumanly long scream. Tool's 'Fear Innoculum' won a grammy as a metal album and you can probably count the number of Maynard's screams on the entire album on one hand.

[undone lotus]

I understand wanting more epic ebb and flow, like what kind of solo? or who do you have in mind when you're thinking of your inspo? if you're higly specific, I can guide you in the direction to get it nearly exactly how you want.

Oh yeah, and I was joking about the screaming in metal thing, you can do whatever you want in music. I just went and saw the nerdiest metal band in the world (Ayreon) and they have very little screaming and it's composed like Opera and goes against a lot of the conventional ideas of what a lot of people think metal should be

though I do love me some technical death metal screaming with non-stop gravity blasts and noodly guitar

[cosmic birch]

Are we no longer able to work with the writing AI to generate a section of lyrics at a time?

[cosmic birch]

In Edit Lyrics mode if you select something it will only replace what you select rather than rewriting the whole thing, in case anyone else is trying to figure this out.

[lost holly]

It would be nice, if that would be true 😄 depending on how you phrase it, the ai will replace more -> all.

[thorny tinsel]

Hello Hello. Since a Couple if days i can Not add Songs to playlists anymore. Anyone Else had this experience?

[sweet mesa]

Hi everyone, does anyone know why the user interface has changed? The auto lyrics option has also disappeared and the quality of the performance has decreased in many elements. Is there a way to return to the old interface?

[prime prairie]

So, I've been playing my Playlists on repeat at night. And tonight, that option has disappeared! Why am I paying for shyte?

[proper token]

Has the amount of characters you can use as a prompt dropped significantly on the app?

[sweet onyx]

Hi

Need help

Got credits and im getting internal error

[echo prawn]

After Suno's update, the cover songs became more and more unstable

In the original case, I turned the style and audio influence to 100% and still got a satisfactory result.

Now it just generates some messy poop

And now generating "a few seconds" of songs will not return credit

[graceful condor]

It is so unbelievably annoying lately that even extends are cut off again and are way too long most of the time 9 out of 10. I need seconds only seconds and removed already everything from the prompts expect [Fade out] [END] and it still creates long extends and singing again. This is so frustrating and wasting credits for nothing. Another issue that got worse lately. It worked way better before the "great" changes we have now to deal with

[lyric kestrel]

how would you describe a prompt for this song?

https://youtu.be/b7tRRauk7ys?si=FWPx6BBxKB48Fntf

[graceful condor]

Give Grok the link and it will write it for you

[lyric kestrel]

most of the time it just makes it up tho.. not accurate sometimes

[graceful condor]

[Start]
[Prompt:An aggressive,hypnotic,and high-energy instrumental drift phonk track in the signature style of JVDXN's "Stay Focused".The entire track is built on a foundation of a heavy,distorted,and saturated 808 kick drum and the iconic,relentlessly driving cowbell pattern that defines the genre.The sound is dark,aggressive,and repetitive,designed for intense focus,driving,or workouts.The "vocals" are dark,repetitive,pitched-down Memphis Rap vocal samples,which act as the main hook.The production is modern,bass-heavy,and intentionally distorted and blown-out,not a clean hi-fi sound]
[Mood:Aggressive,hypnotic,driving,intense,dark,energetic,focused,confident]
[Production:Modern,bass-heavy phonk production.Intentionally distorted,saturated,and lo-fi on certain elements.The mix is powerful,centered,and dominated by the kick,cowbell,and bass.It should sound loud and powerful on mobile speakers]
[Performance:The instrumental performance is robotic and precise,locked into a relentless groove.The vocal samples are used as a percussive,rhythmic instrument]

[Intro]
[Arrangement:Starts with a heavily filtered version of the main cowbell pattern and a dark,ominous synth pad.The sound is distant and muffled,as if coming from outside a car.A reversed cymbal swell or a synth riser builds tension before the main drop]

[Main Groove / Drop 1]
[Arrangement:The beat drops with full force.The distorted 808 kick drum hits hard and heavy.The loud,driving,iconic cowbell pattern is the absolute focus,playing a constant,syncopated rhythm.A sharp,trap-style clap or snare hits on the backbeat.Fast,tizzy,programmed hi-hats provide high-frequency energy.A deep,rumbling,distorted 808 sub-bass provides the massive low-end]

[Vocal Sample Hook]
[Arrangement:The main groove continues relentlessly]
[Vocal Sample:A heavily pitched-down,distorted,and repetitive Memphis Rap acapella sample enters.The phrase is short,rhythmic,and hypnotic,acting as the main hook.The sample should sound dark and aggressive,like "Stay Focused" or "Get Buck"]

[Verse Section]
[Arrangement:The main groove continues,but the vocal sample drops out,leaving space.A subtle,dark,and simple synth melody might enter,playing a few haunting notes]

[Breakdown]
[Arrangement:A classic phonk breakdown.The heavy 808 kick and cowbell drop out for a few bars,leaving only the fast hi-hats and a heavily filtered,echoing version of the vocal sample.This creates a moment of tension before the beat crashes back in even harder]

[Climax / Drop 2]
[Arrangement:The main groove and the vocal sample hook return with maximum energy.A siren sound effect or a distorted synth riser might be added in the background to increase the intensity.Everything is at its most powerful and distorted]

[Outro]
[Arrangement:The song deconstructs.The cowbell and hi-hats are slowly filtered out,becoming more and more muffled.The beat stops abruptly,leaving only the deep,rumbling 808 sub-bass to vibrate and slowly fade into silence]
[End:Fade out]

Style

{Tempo:Allegro,140 BPM,fast and driving}
{Genre:Phonk,Trap,Electronic}
{Subgenre:Drift Phonk,Memphis Rap Revival}
{Instruments:Distorted 808 Kick,Cowbell(essential),Trap Snare/Clap,Tizzy Hi-Hats,Rumbling Sub-Bass,Dark Synth Pad,Siren FX}
{Vocals:Pitched-down,distorted,repetitive Memphis Rap acapella samples.No traditional singing or rapping}
{Mood:Aggressive,hypnotic,driving,intense,dark,energetic,focused}
{Mix:Bass-heavy,mono-compatible,loud,powerful,centered}
{Production:Modern phonk production,saturated,intentionally distorted,lo-fi elements}
{Form:Intro-Drop-Hook-Verse-Breakdown-Climax-Outro}
{Performance:Robotic,precise,relentless groove}

[lyric kestrel]

Ok thanks i’ll try that see if it works. Thanks

[graceful condor]

Not made with Grok, I did it with Gemini 2.5 Pro turned out way better to use for Prompts. Grok is only good to analyse Mp3 or Links to songs like a YT link. But I discovered that Gemini if you name the Artist and Song does a better and more detailed job than Grok.

[lyric kestrel]

Ok thanks for the info dude 🙏

[lyric kestrel]

What prompt would I use? I put all that into my lyrics

[graceful condor]

What do you mean?

[quiet solar]

That's been an issue for quite a while. Suno doesn't play nicely with super short generations, and will make up something to go in the space if you just tell it to end straight away. You'd be better off feeding it sections to work with leading up to a Fade Out / End that you can then remove using the Editor.

[graceful condor]

I work always with sections but SUNO rarely stick to them. Even lyrics are still going on it cuts it off middle of the word etc. For exmple I use Sections like this in Style for example: {Form:Intro-Verse-Chorus-Bridge-Coda(Choral Climax),dramatic dynamic build} or {Form: Intro-Verse-Chorus-Verse-Bridge-Outro}and follow the same route in Lyric prompt but SUNO does what it wants at many times. Like totally learn resistent or command resistent.

[graceful condor]

{Tempo:Allegro,140 BPM,fast and driving}
{Genre:Phonk,Trap,Electronic}
{Subgenre:Drift Phonk,Memphis Rap Revival}
{Instruments:Distorted 808 Kick,Cowbell(essential),Trap Snare/Clap,Tizzy Hi-Hats,Rumbling Sub-Bass,Dark Synth Pad,Siren FX}
{Vocals:Pitched-down,distorted,repetitive Memphis Rap acapella samples.No traditional singing or rapping}
{Mood:Aggressive,hypnotic,driving,intense,dark,energetic,focused}
{Mix:Bass-heavy,mono-compatible,loud,powerful,centered}
{Production:Modern phonk production,saturated,intentionally distorted,lo-fi elements}
{Form:Intro-Drop-Hook-Verse-Breakdown-Climax-Outro}
{Performance:Robotic,precise,relentless groove}

This part is for the style section. The larger one above it is for the Lyric one. If you see those [..] they are always lyrics section and {..} are stlye.

[blissful rampart]

Question. If I upload a full song I've written without singing, how do I have Suno (I pay for pro), simply sing over it? I uploaded a 3:32 song before and it changes it. Changes key, adds solos, makes the song 4:50+. I just want vocals over it.

[torpid otter]

hi every one...after last update/upgrade i ve noticed some issues....for example if I want to create a cover song the output doesnt resemble the track that i ve uploaded...even the style or the instruments.....have you experienced anything similar?

[quick sonnet]

It's a gamble, but if you write in the styles promt, what you want to get, it increases your chances. Also try to increase the audio influence, as pro user you should be able to do it.

[graceful condor]

I said it ones and say it again one of the biggest loss after we got this new UI is the absolut or nearly absolut destruction of the use of Covers. Before it adapted the melody of the used audio regardless how the sliders were placed or how detailed your prompts were and now it nearly always ignores the Audio especially if your Lyric and Style prompts are detailed. It is like if they completely devour the Audio and ignore it. It makes using Audio nearly worthless anymore.

You have to remove the Persona and reduce all your prompts to the absolute bare minimum like only Lyrics without any directions and commands for the AI and then only then you might get to hear the melody of your uploaded Audio you wanna use. Really heartbreaking whats been going on. Sad times

[wide pond]

Anyone know where stored prompts live now? I'm aware of the bookmark button, but done see a way to recall them?

[tepid moth]

Hmmm is that not what the "Add vocals" is for? Haven`t tested that myseft, but I guess that is the way.?.

[torpid otter]

do you know any further details for that ? are they goingo to resolve the issue?....even if you try old versions , nothnig works.....i have lost all my credits and I bought new credits for this....

[polar spade]

@hearty jackal Check your billing history first to see if they have a request to make payment.

If you had stopped payment (block card, have no funds at time of payment, remove payment info) the payment processor does still make at least 3 attempt (or more) attempts as a courtesy at retrieving payment from you last source given. But as you say, our account DO NOT CANCEL merely by auto-pay not receiving a payment. We have to cancel early (at least 7 dsys) before and let the refresh date drop us down to which tier request by us.

So updating payment will likely just continue your account to renew on the 9th again.

[lost forge]

I experimented a bit how far I can go with voice attributes using Suno only.

https://suno.com/s/mRjozFaywGL228BW
Now if I only could get the echo effects to work.

[graceful condor]

Sadly no they not really communicating with us. So it is hard to say.

[winged peak]

Had some good results by mentioning not just 'echo' but an echo-type like 'tape delay' or 'space echo' in the style-prompt, at the beginning of the lyrics and right before the phrase you'd like to apply an echo to. The more specific you tell Suno the kind of echo (=delay) you're looking for, the better the chances it might work.

[lost forge]

Thanks will try!

[winged peak]

At the bottom of the style-section where Suno randomly suggests styles, the first button is the |\\ - one, where you can access your saves prompts.

[winged peak]

As I'm using Delays in probably 95% of my tracks, maybe you'll find some inspiration. I left all the prompts in there so you can copy&paste and give it a shot. Good luck! https://suno.com/playlist/b8484a54-ceea-4e74-98a7-b7d97e483a7f

[polar spade]

".....to sing over it...." (and change nothing of your original instrumental) is the rub. That's just adding a Suno track of vocals that sing your exact melody over the actual track of the instrumental.

To do that presently, you would use another program to add vocals to (replace) an already existing melody line (and may need a MIDI version). Then sequence that back in with your instrumental, if the online AI program doesn't do it for you.

Suno cant do that within three steps from what I've seen.

[wispy laurel]

If you want the exact original instrumental, Suno can't do that in any amount of steps
If the Muses are smiling and the Cover function is working properly today, you might be able to get Suno to generate an instrumental that's almost identical to the original
But if you're committed to actually using the original recording rather than Suno's imitation of it, you'd have to split the vocal stem out of Suno's version then use other software to combine it with the original file. Suno currently can't incorporate uploaded audio directly into its output, the closest it can get is to try to copy it
(Add Vocal/Add Instrumental should do what you're trying to do, but they actually don't)

[polar spade]

@winged peak Hey. Thats what I said too . . . And I didnt gets a thumbs up? 😢 j/k 🤗 🤣

[polar spade]

For fun , I uploaded a song, made a melody line out of the vocals and added vocals to the melody line to get a new vocals with the melody line and lyrics of the original. Just to go around my .... foot... to scratch my ear. Useless, but a fun test.

Still cant add the original instrumental part onto the augmented vocals within Suno.

Granted , with Suno Studio coming on the horizon, this may look more promising. 🙏

[polar spade]

That is to say: its likely not an issue. Not something to be resolved or fixed. But simply putting the oness on us seasoned users to re-up our game and require a bit of a new learning curve with each update.

Otherwise, new users, as said, wouldnt know the difference, learn the sweet spots, and fly passed us old school folk. 😄

[torpid otter]

thank you for your answer, but believe me I have tried everything(sliders, styles, input audio, lyrics, exclude style...) and nothing works as before the last update/upgrade.....I spent like 2500 credits to experiment for 2 covers only and I was like 60% pleased for two only outputs....thats like 1% success overall....a simple example is the one that i ask for male voice and the output comes with female voice.....I think they have changed their policy towards covers but I dont know why....or something went wrong after last upgrade....they have to resolve it or they are going to lose......

[polar spade]

Are you specifically aiming for the vocal to be cloned with near same characteristics of say, your own voice or at least the same vocal on the audio? And you got that before, but not now?

Or, as long as its male and , say, a baritone, then it doesn't matter?

[final hazel]

I get what you’re saying.. making music through prompts can feel different from “traditional” art. Many of us here use Suno because it helps us express ideas quickly, experiment with sounds, and even spark inspiration for art we might later create by hand. Both paths are still art, and both have value.

[polar spade]

sniper? I guess. Darned, and I had a nice retort.

[final hazel]

He wasn't here for a fair discussion, he was in more channels 😉

Discussions are fine.. disruption not 😄

[polar spade]

Your response is more eloquent. I'll claim you my proxy.

[proven plank]

Using a copilot to generate songs is a must, as Suno's copilot is not up to par. There's nothing more to say.

[polar spade]

You mean Copilot from Microsoft -- to write lyrics ? Just clarifying.

[proven plank]

Copilot is philosophical.

[proven plank]

I put your song in my song writer copilot
https://suno.com/s/J8bA6HVuZiozLhYb

[quiet solar]

Further to Dasvon's comment, it's pretty much all in the sliders and the prompt. I just did a couple of tests, and Suno had absolutely no issue producing a cover that sounded very close to the original - although it was clearly a reinterpretation. It will never perfectly copy the instruments, vocals or music - especialy not if there are a lot of complex things going on.

If the original audio was made in Suno, then ideally use an identical style prompt and go for something like Weirdness 5%, Style Influence 30% and Audio Influence 95%.

If the original audio comes from outside of Suno, then it helps to have a style prompt that's very closely attuned to the song you uploaded - re-using Suno's interpreted prompt from the upload can be a good starting point, but it doesn't usually give enough of a steer to replicate an entire song accurately, so there will inevitably be a number of changes. However, if you crank Audio Influence up and Weirdness down and play around a bit with finding the sweet spot for the style prompt then you should be able to get something that shadows the original.

[sly spruce]

Anyone able to help think this is a bug...

Im not able to see a button to bring back up lyrics that ive saved after pressing the bookmark save button, is this a bug? i see the book looking button to load style prompts that are saved but nothing for the lyrics.
when saving lyrics it shows "saved prompt" at the bottom but i cant find how to access those saved lyrics, any assistance would be awesome

[polar spade]

Well , thats a new one..... ( my post vanished? ) meh 🤷‍♂️

I made an edit to the post so thats what zapped it. The line spacing.

[polar spade]

@oblique pond . . . Hello Guys!,
I have really short question i guess... I would like to create song in two diffrent language but keeping the same style ( ye, it might be a little diffrent) but still i want it as similar i can get. Is this possible ? or i need to create new one using same prompt. . . ?

Since my last post disappeared....I' ll post a shorter technique:

A short way to get this done is paste the lyrics of both languages in the lyrics , that is, one full set in one language , then the next. Let it sing the full song to pick up the melody lines and metering of the stanzas.

Then crop out each song of one language and the song portion of the other language.

Make a cover of each.

[final hazel]

Sorry Davson. It happened because you had more than 5 empty lines 😐

[polar spade]

I didnt look at the whole thing when I added the second bit, so didnt realize I added more spaces than needed.

Damn, it sux to be old.

[final hazel]

You are not old.. Woodhouse is 😉

[polar spade]

I eat plums presently. Its only a matter of time that my tastes change.

[neat field]

I wonder how afraid Suno team is of Qwen audio team...

Eagerly awaiting open source to catch up to even Suno v4 levels

But Suno still a great deal for inference/storage/remaster, team will probably still get my money sometimes to try their v5 and up

[polar spade]

The library/books icon on the bottom left of the styles prompt box should bring back up the saved prompts.

As much as I can tell, though, It has a few bugs at the moment after some updates.

But I dont believe it is set to save lyrics anymore, nor has it retained the lyrics from old saves. Just from my clicking on past saves from my saved prompts list.

Whatever may fix the bug, may change actions regarding lyrics. Just a wait-and-see thing after reporting the bug.

[sly spruce]

yea it use to save the lyrics and it even has a individual save bookmark for the lyrics prompt. it shows "saved prompt" when you press it for lyrics but no way to reuse now.. crazy how they didnt tell us

[polar spade]

The models themselves, that are not beta, are released models. Interfaces, though, are always in flux and considered perpetually temporary. 🥴

Mention in suggestions box to keep that feature. I liked it myself as a crutch while editing songs and working out combinations of lyrics and prompts.

[proven plank]

Have some ressource about Qwen?

[polar spade]

Through the years, a VHS cassette tape was basically the same, but the VCR lost alot of nice features ( like the ad skip feature) that kept them cheaper to manufacture. Even the cellphone lost their earphone plug socket and literally no phone company has yet mentioned it. 😆 it was a good feature to help fish out your phone by the earphone cables if it drops in the toilet.

[blissful rampart]

I tried doing that, but because Suno changes the song so much, the vocal stem doesn't line up with anything. 🤔 I wonder if Suno studio will have this ability.

[wispy laurel]

Weird. I can usually get it to give me a cover that at least matches the tempo and melody. And then if necessary can edit the vocal track, add or remove silence to get everything in the right place
If it won't match that closely though, unfortunately there's not really much else you can do.
Hopefully this will be a thing in Studio, it seems too obvious to not include it. Although recent updates and changes have made it pretty apparent that the devs don't understand what their users want and need. So I'll believe it only when I see it, sadly

[blissful rampart]

One can only hope. They advertise someone playing an instrument and putting the rest to follow. How accurate that is, we won't know until it is released.

[neat field]

Not yet, I hear rumours - qwen did just train a ASR model that can label music though - easy to find that work of theirs if you wanna dig

[proven plank]

Thanks, I will have a look

[prime prairie]

Why the HELL was the Repeat button removed from the Android app?

[polar spade]

I find it bottom right while playing a song or Playlist. I was not able to make it NOT appear by changing display formatting. Android.

[prime prairie]

Not there!

I have been using it at night when I go to bed. But the past couple of nights, it's not been there.

[half dragon]

probably an issue with the way it renders the screen.. mine is super small and the other stuff got cut off too..

[prime prairie]

I haven't had a problem before. As you can see, mine isn't cut off. I just want my d@mn replay button back!

[half dragon]

Report to #📱┃bugs-android if you haven't. Seems like an issue.

[prime prairie]

I'll try. But they won't do anything about it.

[polar spade]

Update was September 16th to ver 1.28.1

The button Is to the right of the player's timemark for mine and VRKing. Your player [timemark] extends all the way right, so the replay button is likely lost off to the right (or bottom) of the widget panel for your screen per your phone. If you have ver 1.24.1 or below....you may just need to update the app so it formats the display properly again? Check your version.

If the update still has your player and its timemark all the way right, try changing your phones display, at least to test if a smaller font and/or display format will squeeze the player small enough to show the replay button.

[vivid violet]

...

[prime prairie]

Changed the font size. Still not there. I can't scroll lower, or to the right. The effing button is just GONE!!!

And yes, the app updated yesterday!

[cyan jewel]

@serene delta my songs all end up looking like this. you can see near the end where it's just oops all loudness.

dynamic range doko?!

[serene delta]

Inai yo,Tanuki

[cyan jewel]

xd

[polar spade]

Sorry to ask again... but screen zoom is also smallest ?

I zoomed display up and was able to wipe the button off the screen when any song has an added caption. Or squished if the title wraps more than one line. The rest were more or less normal. And your photo didnt have either added things in the lyrics/title. So im out of ideas.

Ive got an s22, s24 and Fold 5.

But I still cant get the player banner to widen horizontally on its own, so that the button goes off the edge.

[polar spade]

And no biggie to uninstall and reinstall, rather than just update, to see if that helps.

[dense meadow]

Sorry, i'f i may ask a question again, but i searched for and didn't find it here or in the help channel...

I struggle with a song that has to be 8 min long as i am not able to shorten the lyrics without blowing the meaning of the song up... 60 times tried, 59 times failed to get a version long enough to fit all the lyrics (and the one... well... let's not talk 'bout)

Experimented within the prompts, but does not work. Extending is terrible, no single try brought it seemless, no healing does neither.

Do somebody find a workaround here or do i at least have to switch to producer ai to simple type in the length i need? ( if doin' a jingle for a promo, the length is limited exactly to the millisecond...)

Thanks in advance - and hope, this is the right channel.

[prisma cypress]

cam someone help me with this Uploaded audio matches existing work of art. error

[frigid orbit]

Did we lose bookmarks?

[torpid otter]

no, only male or female...thats it.....but this is the simplest thing that occurs....when you upload a file you expect the output to be at least close to the file....now its like it ignores the input file....before the last update with one try i have very good results...now....i spent like 2500 credits......hahahaha...for example i play guitar and the Am chord changes it to anything....or when it sings ..it sings another melody off tune......a lot of errors....

[torpid otter]

thank for your response, but my opinion is that after the last update/upgrade (a week ago), in order to make a cover is so frustrating....another example, you want drums in the cover...so I write POWERFUL DRUMS.....in the output no drums at all.......before i didnt have this problem.....the output is a different song ...even if I slide the weirdness and the style to zero...the output melody is something else...something bad....maybe it works better for english songs than the greek that I am trying to cover......but still i havent these problems before.....

[torpid otter]

ok....listen to me.....something wrong with suno.....the covers are shitty....I dont know why...they have to resolve it....if you know music you can notice that on a second....because we give money....the truth is that now i give more to reach in a output that is like 50% to the audio that i uploaded....another shitty example....you ask for male voice...it starts male...at chorus female....and its not one....its like 10% from all the outputs....another shitty mistake....it doesnt follow the melody at all....even if the sliders are in 0% or 10% or 20% or 50% ...even if you try all the combinations.....come on...am i the only one that have these issues??.....who ever wants....i ll give you a audio to upload and you can give a try ...i dont know what else to do....its so frustrating and money loss.....from yesterday till now...one audio with tunned guitar and sang by me(its not off)...i did like 60 covers.....no one is good...something always fails.....and ....I say it again....a week ago I didnt have these simple issues.....the only thing that i want a cover to has is...be close to the instrumental melody(80% and up), the vocal melody(80% and up)...i dont care if it plays paino instead of guitar( i care, but i write only the main things that a cover has to do)....and the cover that i take as a result is like 10% to 50% close....its another song.....

[proven plank]

I guess there is a lot of features that will never work by design.

It's helpers not a solution

[ebon sand]

I've also had a bad time with covers but couldn't pinpoint when it started

[torpid otter]

i would agree with you ..until a week ago i didnt have any of these issues.....whaen you upgrade something you do it for posotive results...right??....i dont get it anymore!!

[proven plank]

Like chatgpt and all AI generative

[torpid otter]

i am not a newbie....r u greek??

[ebon sand]

No, just an enthusiast, but I write songs in greek

[proven plank]

@torpid otter would you mind review an album made only with one-shot ? The prompt is still on the song. I will clean that after.

[ebon sand]

The covers recently didn't attempt to follow the provided song, they sound as if it would just create a new song without a cover song being present

[torpid otter]

ok...thats good for music experiment...the thing that mekes me crazy...from all people here...i am the only one that writes all these......i feel like the crazy one.....hahahhaha

[torpid otter]

so its not a cover......right???

[ebon sand]

To me it's not working as I'd expect it to, yes

[torpid otter]

ok.....no problem....

[ebon sand]

I used to use covers to polish and improve songs with little change and can't seem to get this anymore

[proven plank]

We talk about prompting right. Don't ask me if I like this kind of music 😅 https://suno.com/playlist/c107ebce-29b0-4dbf-bd34-6ba31a59da68

[torpid otter]

to me it worked...not anymore...and i am trying to figure why...because at the end we give money.. right??

[ebon sand]

Yes, I'm agreeing with you

[ebon sand]

I'm not sure what this has to do with covers

[proven plank]

This channel is tech prompt talk ?

[ebon sand]

Sure, I thought you were responding to kotsitos still, if it's a new topic then I apologise, do you have questions for that prompt?

[proven plank]

Are there any people interested in algorithmic music?

[ebon sand]

Not me personally, I just responded to cover behavior

Others for sure

[gilded mortar]

Are they going to listen to us and fix it ?

[ebon sand]

I don't know but maybe if more people hand it in to the bug report channel, they for sure won't skim this one here

[proven plank]

That apply also to voice. So the base of the algorithm improve everything I guess

[torpid otter]

so you have these kind of issues for a week or from the start??...because i have noticed that after the last update where all the web screen changed...in order you can see it all....but with that positive update , for me the meaning of covers is not the same anymore.....

[foggy oar]

hey people, anyone knows how to fix my issues please, thank you

First I cannot seem to pin any song in my workspace - it simply pins and unpins within a second by itself.

Second - I cannot find my trashed workspaces, i never intended to trash them but to simply store my finished workshops out of view for my daily workflow but now i cannot find or access those workshops that still have all the stems from those sessions

[ebon sand]

I didn't have this behaviour in the past but have been unable to create covers that didn't sound as if they were just fully newly created songs recently, when it started I don't know but it's not as it used to function

[ebon sand]

I'm creating very different music where these wouldn't apply and sometimes not even music at all, I'm not the right person for algorithm 🙏

[torpid otter]

i dont use suno for new songs.....i just use it for orchestration only(cover)...so from a (guitar - voice ) input , to take as an output something with the orchestration that I ask for(close to that)....and it was ok...but now its off....its not working at all.....

[ebon sand]

Yeah that's also how it is for me

[polar spade]

Thanks for reiterating that you seem to have tried everything, spending 1000s of credits without success. Clearly, there is nothing more I can personally suggest either or explain to improve the circumstances.

Perhaps at this point, try other fresh ideas, or integrate other outside sources to see if there is a successful combination that work best for you. I wish you luck my friend.

[junior terrace]

Hi everyone, I'm trying to create a song just like Daft Punk, but it looks like Suno does not know what Talk box is...

[torpid otter]

what if someone else try it for me???...i want to check if really something wrong or i am freaking out without any reason....

[proven plank]

No problem! And if you have a song online so I can try a facelift myself, I'm interested (I'll stop otherwise you'll say I'm a pushover)

[torpid otter]

try vocoder vocals

[polar spade]

They're of an odd sort at the moment. After a dozen tries, I dont find "bookmark" as saving properly or consistently. Only three saved after 4 minutes, and two just saved literally one letter. The timestamp of when they were saved seems to work though. 😅

[junior terrace]

Thanks I'll check

[ebon sand]

I can look for a song later I could open up but they are all v3.5, I'm very protective of my singer and have everything on no-remixes

[proven plank]

It's hard to replicate.
But sometime I prefer the new way.
https://suno.com/s/mcry5rXXzSUFmCnH

[proven plank]

I am trying to resolve the same desease right now, I understand. 😁

[lusty juniper]

are they going to fix the buggy UI soon? the create section is laggy and sometimes i cant even click to type. it closes that bar

[junior terrace]

damn that looks really hard

[winged peak]

If Vocoder does not the trick, add things like Vocaloid, Autotune, Talkbox - even all of 'em could lead to the result you're looking for. Good luck!

[cursive tartan]

would be nice to allow remixes 😇

[proven plank]

It would be better idea to regenerate the song

(all the prompt are inside it)

(I tick the allow remix)

[lusty juniper]

Suno is very... laggy... there is like a delay when loading a song, when generating a song, when typing in the styles. etc etc. I tested if its my computer, but its just Suno that is laggy

[final hazel]

It's working perfect for me.. So sounds more like it's your ISP who is having a queue on the route to Suno. You can test and solve that if you have a VPN. With a VPN it will take a different route to Suno and thus avoid the queue.

[wind skiff]

Guys

A question. I saw that the free plan now has a character limit for prompts. There's no such limit in the pro version, right?

[shut marten]

I noticed that too I am also having issues with a slight delay in the execution of my lyrics with the tempo in portions of my song idk how to really fix it inside suno

[shut marten]

nvm I fixed it XD

[shut marten]

ugh its not fixed

👀

[lusty juniper]

yeah for me it's still an issue. i tried VPN didnt work.

[rocky wigeon]

how can we use like a proffesional

is lyrics matter then style ?

and whats the weirdness and style infulience guys

[lusty juniper]

what?
weirdness will make your song different and random from what you except it to be.
i usually make mine 0% and style 50% and audio 100%

lyrics are important of course. its the message of your song. as for the style you can decide. what kind of song you want it to be. RnB? rap? hip hop reggae acoustic piano violin etc etc

[pastel plaza]

hi guys how do you clean the white noise in the songs ?

[torpid otter]

who is using suno for covers????

[lusty juniper]

me

[lusty juniper]

not quite sure what you mean with white noise

[torpid otter]

because i have some issues after the last update (i use suno in pc), i would be very grateful if you could help me to check what is wrong.....first of all ..is it working ok for you??...i use it for greek lyrics, but the styles are the same(rock, pop...)....

[lusty juniper]

since the update, it is very laggy for me. i need to wait sometimes 5 seconds before i can press the next button

send me a link of the song so i can check

[rigid falcon]

Hello. First post here. Has the download option disappeared from the android app?

[torpid otter]

i upload an audio that i have recorded...and i use cover to add more instruments after i choos also what style i want to....thats it.....in order to have an idea i have to send first my audio in order to have a base to start...

[lusty juniper]

no its still there. just click the 3 dots on the right

[lusty juniper]

yes. i do that as well. what do you need help with?

[torpid otter]

ok...let me make it clear...i have wrote that before...but the strange thing is nobody has same issues....

[lusty juniper]

in styles i put Organic, Audition, Lo-fi
weirdness 0%
style 50%
audio 100%

or i put in styles High Sound Quality
weirdness 0%
style 50%
audio 100%

[lusty juniper]

what issue though? does it sound bad? sound different or what?

[torpid otter]

yes....different and it doesnt follow the things that i want....but this started a week ago....

before that...everything was great...

[lusty juniper]

what styles do you put? send a screenshot

and also try changing between 4.5+ to 4.5

or 4.5 to 4.5+

like you switch it

[rigid falcon]

It's not there

[torpid otter]

i change between these two versions....the results are crap

[lusty juniper]

it only works if its a song you generated

if it is, and its not there, either you are logged in on a different account or there is bug cause i have the download button

[rigid falcon]

I logged in on my laptop. Is that related?

[lusty juniper]

shouldnt be. im logged in with my phone, tablet and PC

[lusty juniper]

can you send me a screenshot

[rigid falcon]

I did above

[lusty juniper]

no no not you. i mean kotsitos for his song

[rigid falcon]

Sorry just realised.😁

[torpid otter]

this the style description : Modern pop/laïko power ballad, mid-tempo (85–90 BPM). Emotional male vocal, starts with piano and acoustic guitar. Big build in pre-chorus with driving bass, lush strings, and rising percussion. Huge, explosive chorus with full band — bouzouki riffs, electric guitar power chords, deep punchy drums, cinematic strings, and modern pop production. Passionate, dramatic, radio-ready, bittersweet but triumphant

[lusty juniper]

for the specific kind of instructions, i used to do that but not in style. i put it in the lyrics section

[torpid otter]

screenshot

[lusty juniper]

like this:

[Piano and acoustic guitar]
[verse 1]
bla bla bla

[Pre-chorus build up with driving bass]
[lush strings]
[rising percussion]
bla bla bla

[Chorus with full band and energy]
bla bla

and i often had to cut and extend

i keep the parts that were good and then i extend

[torpid otter]

but i always did that....and i had great results...now i have to write everything in the lyrics section??

[lusty juniper]

no. if it worked before then it still works now, but you just need to try again and again

until you get a good one

its also part luck

[torpid otter]

ok...but before...i tried only two or three times.....now...it takes all my credits for one cover and no right results

if you could listen to the input and then the output you ll dan better understanding....i will give you an example though....i choose male vocals.....it starts with male...then you listen a crush in the sound and it changes to female....hahahahahaha....from 100 covers this applies to 10 songs.....another example...you give the key..lets say A minor....it changes to something else.....that applies to 20 covers....

[lusty juniper]

ahhahaah

exclude female vocals

that might help a little

[torpid otter]

i did....

believe me its like.....i didnt

i ask for no zeibekiko rythm....exclude...right..?....the cover starts with zeibekiko rythm.....hah

[lusty juniper]

it happens to me too. some songs are cursed i thought. but based on the prompts you put and flow of the melody, the Ai assumes a female will sing this now, so i just change some parts of the lyrics.

hahahaha yes yes it happens if you are unlucky

[torpid otter]

unlucky???....a week ago i should play lotto...

another example...it makes it 8 minutes long.......8 minutes!!!!!.....the lyrics are not too long......whyyyyy????....i used all my credits for two covers (2500) ...unlucky??

[proven plank]

@torpid otter IHMO it depends on quality lyrics

[torpid otter]

the lyrics are greek...but still...for me a week ago and for moths ago after the cover was introduced...everything was great...what changed now????

[proven plank]

My benchmark for lyrics is I change the style radically, if it's not work the lyrics are not good.

[torpid otter]

what do you mean "are no good'??...give me an example

[proven plank]

https://suno.com/s/LWoqkMNkxXBWCSUu

https://suno.com/s/5otelfo9WdQ5bWY8

This is 2 separated one shot

[torpid otter]

is this a cover????

[proven plank]

What am I saying is there is kind of universal lyrics that can be change all way you Want. And some not.

[torpid otter]

i am talking about covers....covers follow an input audio....and changes comes from the sliders or the style description....not from the lyrics.....

[proven plank]

If they work the way you want.

[torpid otter]

yes...the lyrics are measured in metronome....so...it should follow the input audio...

in the input audio the lyrics are in there...singing....

[proven plank]

The model has learn based on a dataset.

It learn on a specific musics. So there is internal rules.

[torpid otter]

i think that i am unlucky if nobody else has the same experience.....a week ago everything was great...now....its crap...totally crap....but only for me....i dont know why...i will try to transfer everything in lyric section....

[proven plank]

https://suno.com/s/MxSZ9Gx3Zzsncyyd

[torpid otter]

by the way the first one...loved it!!!

[hollow wigeon]

Anyone here downloaded stems to try to edit on your own DAW? I drag and drop wav files all the time for freelance recording out of my studio, but no mater what I’ve tried, I cannot lock the Suno tempo into Logic. I also have tuning issues because the sounds are slightly off standard tuning. Anyone else experience this? Do you have a work around??

[proven plank]

You can compare how the lyrics are generic.

https://suno.com/s/uaRnhw4nrhCQm0TH

https://suno.com/s/uWBueZC7H9jtPAiU

https://suno.com/s/ME3zUoc8HNu4mn7B

[native breach]

I hope someone here could help me 🙏 . I have done a song in Suno with lyrics but now I want the same song as a instrumental version, is there any easy way to transform it into a instrumental track?

[quick sonnet]

Separate the stems in instrumental and vocal.

[lusty juniper]

cover the song but remove lyrics. just select instrumental

sometimes it wont work right away but eventually it will. put like piano instrumental

[native breach]

Now I see how you can extract steams, vocals and music 10credits, thank you 👍

[quick sonnet]

You welcome.

[native breach]

Thank you as well MTH, I will try that 👍

[neat field]

my weekend hobbyist research lead me here to "a suite for training, evaluating, and performing inference on state-of-the-art pretrained music models" called MARBLE on github, which also lead me to "International Society for Music Information Retrieval Conference" 3rd party page which shows that SUNO will be presenting a paper at the conference - sometime between today and Sept 25th

not posting links because of the big W automation....

Suno has not made any announcements on their website or discord they would be presenting at ISMIR conference - interesting radio silence

edit: looks like they don't have a paper there, Suno simply is a sponsor of the conference

[long lily]

Sometimes real piano is tuned to 441 or 442 Hz. And guitar tuning is far from perfect on the higher notes. Regarding to tempo, Suno doesn't generate fixed tempo, I saw slight variations over the song.

[prisma hound]

Very interesting prompting method. Similar to JSON at some point. I (lately) realised that now you can have 1000 characters in the style section, prompting it with a very precise definition of what you are looking for makes a whole better sound than just promoting raw styles. I'm wondering if this does not as well favor higher quality material in the generation.

[proven plank]

https://suno.com/s/LLJKC8A8EA0pTX4Z

It create weird things but still on the gray zone 😍

https://suno.com/s/tpBZjCdZ3xTcng5K

https://suno.com/s/on0knMqTAsLGcHmN

Still the same song. It's sometimes crazy

[prisma hound]

Thanks for sharing. It's really interesting. I tried before prompting this way but without great success, your method is well organised. Does the seed variable work somehow or not at all? This is really something which is missing in Suno

[proven plank]

I guess it's more algo seed inside the same song.

[prisma hound]

Where did you get the algo name? I checked ChatGPT and while some are in the literature others are not. But it's interesting to see if this triggers the generation into one way or not. Re the seed I was referring to the fact that in Midjourney you can use this functionality to force generations into the same style. Here what use do you have of this seed parameter?

[proven plank]

@prisma hound I guess I can't explain more 🙂 (than this song)

https://suno.com/s/dTvxDRxGJMZnOK0T

This is my project. (at least the head)

[prisma hound]

I understand, it's all a matter of exploring how the matrix works 😅

[proven plank]

I am born to do that.

[prisma hound]

It's what is keeping us awake with Suno, exploring how to crack the code is a journey by itself! But I must say your method rocks 🚀✨

[proven plank]

I feel like I am young again.

It's so much fun

I put all same song here
https://suno.com/playlist/5880b574-af99-4fc6-b807-58edaf1cfc91

[prisma hound]

Or just by poorly humming the sound and then prompting it the right way - unsure re the copyright policy however 😅
https://suno.com/s/neGtWchgHzHASM0l

[prisma hound]

Thanks for sharing! I'm bookmarking it (and already following you). Great work 👍

[proven plank]

@prisma hound I will put your pretty song in my prompt. The result will be interesting for me. I will send you for educational purposes.

(it will be less than you, it's impossible to up that. Lol)

[prisma hound]

You're welcome, if you need access to the terrible hum recording I can send you that 😄

[proven plank]

https://suno.com/s/LRX8ci4nSIwvnuBj

I only take your lyrics and styles

[prisma hound]

Using your prompt with my humming guidance rocks
https://suno.com/s/nTLYxuk9QWWfDlpZ

For the fun, using the same method, and as our tonight's exchange are around French stuff, a recent review of the Marseillaise with a brand new song and style 😄 I used as guidance an old public domain recording from 1914! But this one with just keywords for the style
https://suno.com/s/JqkEg3jPE4iJupms

[uncut tulip]

omg is remixing from uploadsd banned

[neon slate]

I’ve been trying to figure out a rain prompt forever and it never works.

[wet dune]

I've tried scrolling up, but I am still confused about what all the information is on that page. Can I get a quick update or eli5?

[proven plank]

You mean the tags [] inside the lyrics?

[queen linden]

is there any way to use negative prompts? I'm still suffering through ridiculous swooping and soaring vocals. theres got to be a way around it. I got 3/4 through what i thought would be decent song and suddenly female vocal is reaching for the stars..... gack. its getting pretty silly.

[polar spade]

Its like positive prompting, but when you want to remove the adjectives, "swooping" and "soaring" and not the noun "female vocals" then you have to think of genres that often including swooping and soaring vocals.
Start a test with these?
Opera, power metal, indie rock, middle eastern, world music, . . .

Your question was about using exclusion prompts, but if it were me, Id just add extension from the point it starts swooping and soaring, and include excluded prompts there and even remove much of the original prompts* with each extension.

• ( my theory and practice - and could be just cognative bias) we can remove the original prompts in an extension since it reaches back and seeds off of the previous laid down tracks already for the vibe. This allows the exclusion to have more pull to influence the extension.

[queen linden]

that makes sense. I go for a tracy chapman and end up with angry Idina Menzel lol

[polar spade]

Using Acoustic, Folk and femme folk? And she is also a bit "fingerstyle" and a lot "introspective"

[queen linden]

I wont bore you ,,, this was the prompt. Menacing, emotional, scorned, hurt, intimate indie folk LOW contralto, a self-aware moment of resentment,Chorus stays in the low register, no belting, no high notes, no long sustain alternative R&B 100 bpm. I now see where it would pick up the very things I wante to avoid.... I took them all out and did get a much better result TY

[polar spade]

Ok. Great. Yea. Chapman creates emotion from us, but doesnt exude it. She's pensive, introspective, melody-driven, storyteller.

We all tend to agree Suno doesn't understand "no" as well. 2 and -2 are the same distance from zero to Suno, thats all it knows. So "no trills" is just "trills" to Suno -- "no belting" is really "belting". . Try the exclusion box for those.

[queen linden]

thanks fo much

[wet dune]

Like this:

[thin plaza]

I need help with suno audio quailty?

[proven plank]

A research I conduct right now. Thank you to ask.

[neat nexus]

What are you prompting?

[rigid folio]

got a bug / a problem with the earn credits through listening, some reason my credits wont be processed and all the tasks disappeared

[daring falcon]

same here

[half dragon]

Simply download the song as wav file or wave stems and drop them into Ableton.

[polar spade]

Just taking a big swing at guessing: I would assume there may be a connection with the weekend now here and "work" being done to disburse credits. As there is limited staff on weekends, its possible this is presently manually administered and monitored. So check things in about 12 hours to see if you receive your credits then.

Logically, if I were administering this, I would shut down the tasks available until I could get caught up come Monday. It seems tasks are at zero for everyone, which makes sense. Let's see what happens in about 12 hours. @fathom crypt

[fathom crypt]

It doesn't process mine too and today I got free 50 credits lol.

[rigid falcon]

Download button still not available in the android app. Any ideas boys and girls??

[polar spade]

More indicators that one might be automatic and the other manual. 😄

[fathom crypt]

Lol 😂

[prime prairie]

Tried that. Nothing is fixing my issue! 😫

[polar spade]

Display set to smallest zoom as well?

[prime prairie]

Yep.

Don't know what to do! But it's pissing me off!

[jovial trench]

any prompt advice on getting 3.5 to sing slower and longer notes? If i write a line/stanza with lots of words it defaults to sing/talk or rap-pop no matter what I tell it in prompt.

[fair finch]

Hi!
Does anyone else have issues with playlists? I can’t add Songs to playlists anymore for some reason 🤷‍♂️

[fathom crypt]

I have an issue. I can't delete an empty playlist lol 😆
There is no option.

[fair finch]

Go in the playlist, press the …

Then delete playlist

[fathom crypt]

From trash 🗑️

[fair finch]

You mean you can’t delete a playlist from the trash?

[fathom crypt]

Yeah

[fair finch]

Why would you even want to? 😅 but yea, never tried

[fathom crypt]

Because it's empty lol 🤣

[nova sequoia]

I’m having issue adding to playlist also.

[heavy owl]

I’m adding vocals to a song that’s about a minute and a half long, but it cuts off in the middle. What can I do?

like this

@shy yew

[proven plank]

Share your prompt instead of your wrong result.

[chrome gate]

On a mission to get as dry as possible vocals to use in my productions.
In lyrics block:
[Block,voice=male_smooth,genre=solo_vocal,tags=dry,close_mic,raw,DAW_hint=reverb=off,delay=off,fx_bus=dry_only]
[vocal: single voice only, no harmonies, no layering, no choir, bone dry recording]

In styles block:
One dry male voice only, no reverb, no delay, no instruments, no harmonies, no choir, no backing track. Studio mic, close-up, raw.

Still... there's always some room / reverb, but I keep trying.

[heavy owl]

male vocal, baile funk, melodic vocals

17%

50%

77%

[Intro]
A noite acende, luzes no salão
O grave bate forte, sente a vibração
Corpos se aproximam, ritmo no ar
A pista esquenta, ninguém vai parar

[Refrão]
Sobe e desce, move sem medo
No calor da dança, eu me entrego
Cada passo é fogo, cada toque paixão
Vem comigo sentir essa emoção

[Interlúdio]
Respira fundo, deixa o som guiar
No silêncio da pista, começa a sonhar
O coração dispara, a noite é quente
Treme o chão, cada instante é urgente
Sente a energia, deixa fluir
Vai subir, ninguém vai resistir

[Refrão]
Sobe e desce, move sem medo
No calor da dança, eu me entrego
Cada passo é fogo, cada toque paixão
Vem comigo sentir essa emoção

[Outro]
A noite termina, mas fica a lembrança
O ritmo persiste, pulsa na esperança
Corpos e sons, juntos a vibrar
Essa dança é nossa, ninguém vai parar

lyric

s

[proven plank]

Why You don't help the AI with more defined tags inside the lyrics?

The AI lost often his memory, if it found an instruction near where it is currently working that is helping I guess.

[heavy owl]

wt is defined tags😭

[proven plank]

https://howtopromptsuno.com/making-music/voice-tags

I'm not saying you need complicated tags but between 0 and too much, your are at 0.

[proven plank]

I miss of one of your previous message.

My tags are experimental. I was speaking about more simple tags.

[proven plank]

https://suno.com/s/XjbcG2vAu13z2Nqh

https://suno.com/s/Q9YshiVd0kDVUFZq

[jade mist]

Anyone have information on max song duration for v5?

[heavy owl]

https://open.spotify.com/intl-ja/track/7hCUHXMZknrymWCyeMx8B6

@proven plank how to make vocals smth like this song's

btw the vocals of this song was generated by suno ai

@shy yew

especially the drop part

[proven plank]

You are allowed for example add "enimem-style" in your lyrics tags (not in styles).
In your case, you need to know who sing like that.

[heavy owl]

ill try

[quick sonnet]

wav file only avaible for pro users, but mp3 should be no problem.

[frigid orbit]

Right when I needed it.

I was planning of adding the same sample to all my workspaces.

[quick sonnet]

You mean Eminem style, right?

[storm pawn]

Were there any "secret" updates since yesterday? I generated 10 songs today, pretty much all of them has some sort of a vocal quality issue

[proven plank]

Yes, sorry

[quick sonnet]

No problem, I make stupid typos too.

[proven plank]

Prompting song is fun 😝

[quick sonnet]

What the f... was its problem?

[proven plank]

I want to do song prompting. Meaning the lyrics hypnotize the AI.

If you define a heroes in your song, that could be good if lyrics can live outside the song, if you prompt the lyrics in chatgpt.

[polar spade]

But were you aiming to finish and have a 1:30 song?

[heavy owl]

ye

[polar spade]

Did you already have a 1:30 instrumental? How were you adding lyrics/vocals?

We can upload an instrumental. Then we can "add vocals" to it from the menu. Is that what you tried?

[heavy owl]

ye

[polar spade]

@heavy owl And if the instrumental was 1:30 , originally, how long was the output once you added lyrics to it?

The output was 1: 02?

[quick sonnet]

If you are a free user, you can upload only 1 min to be used, but you still can generate 4 min of it.

[polar spade]

Thats a good first question . 😀

[polar spade]

Are you a free subscriber? Or a Pro/Premium paid subscriber?

[polar spade]

I may have interpreted the post incorrectly when they said they were "adding" vocals to an already 1:30 song.

[quick sonnet]

You can upload up to 4 min as a free user, but have to select only 1 min long part from that audio.

[heavy owl]

im PRO🤑

[polar spade]

OK. Then we keep going.....

[polar spade]

Ok. I see what you're asking.

The output didnt cut off early. The output was still 1:27 minutes. You ran out of music before you ran out of lyrics.

It did exactly what its supposed to do. Assuming that tempo was about the best speed for the vocals according to the instrumental.

If you want that same instrumental, you can extend it out enough to get one more verse in. Split stems to designate an instrumental stem. Then "add vocals" as you tried before.

Thats a bit odd way to do it. But if you need exactly 1:27 as the instrumental, you may need to cut back your lyrics one verse.

I cant tell what you wish your final goal to be.

[quick sonnet]

How about extending it first?

[heavy owl]

ok

ey

ty for response

[foggy oar]

Hey sorry guys, where can I find my trashed workspace? I accidently trashed a workspace i was in the middle of and i cant find it now haha

[polar spade]

You are not required to limit the music to 1:30 if that is not important. You can extend it first to help fit all of your lyrics in.

When you "add vocals" to an instrumental, I think Suno tries to oblige, but its priority is to keep the same length as your original instrumental. So, if a lot of lyrics, it may not all fit by the end of the music.

[polar spade]

Do workspaces go anywhere? I ask myself. I dont think they do. Hold on, ill delete one and see.

[foggy oar]

ya only my active workspace are visable, the rest were "stored" in my trash folder but before the iu change there was a "view trashed workspace" or something like that

[polar spade]

Thats a tough call. If they have removed that option. From whats there now, just three. No workspace. Hmmmm. 🤔

[foggy oar]

ouf

ye seems like they removed workspace from trash

[polar spade]

I personally have never looked. Sorry I guess im no help for that then, if it existed before. Ill look around as well.

[foggy oar]

i stored probably over 50 workspaces in the trash folder, since i do every song in its own workspace and then when the song is done i move it to a finished song workspace and then trash the original workspace for the song but all the stems and song variants are still left in those workspaces so its still gold for me

i'll email suno, im sure they can internally sort something out, thanks for checking tho!

[polar spade]

You move from a temp workspace to a "final" workspace? Or did you mean you move the finals to a Playlist? Thats what I do, at least.

[foggy oar]

I move the finished song from an original workspace named the songs title, edited and whatever needed to a workspace ive named "Finished songs" then from that work shop, i'll put the songs I publish onto a playlist but the song is both in playlist and the finished song workspace

[polar spade]

It makes sense to keep an organized list of your draft outputs, even if tyo prove your work for legal purposes. So im with you there that it helps to have a failsafe trashbin thst we can restore from if needed.

[foggy oar]

then to clear up the view i trash the original workspace so i dont have to scroll through so many, its easier on my eyes if only the 5-10 unfinished songs/workspace are on the screen since i usually mix up my song names and i dont like the confiusing xd

[foggy oar]

ya 100% the trash bin has been golden for storage :D, i might send it in as a feedback if its not only my side not showing it, suno hopefully gives it back with everything still inside hehe

imagine accidently trashing your workspace you almost have finished, poof song gone haha so i really think suno is having some backups somehow

[polar spade]

Yes. Just asking how you prefer. I put finished songs plus the best drafts if i want to revise things in a playlist, and delete the workspace. Leaving the files to float back into the Library, just in case I need to desparately search for one of the really bad drafts - provided i had not deleted it already. 😅

[foggy oar]

oh are you saying that if a song is on a playlist and its workspace gets deleted then that song automatically moves over to the main "My workspaces"

[polar spade]

If you just deleted the workspace, all of those files dont get deleted. They still hang in your Library/main workspace. But good luck finding them all. The only chance is that you spend one segment of time making the song so that the songs are chronologically listed near each other.

[polar spade]

Yes. It you ONLY click the workspace menu and Move to Trash.... ummmm... a month ago.... it just un-tethers the specific workspace. Youre not deleting the files/outputs. They always are in your main workspace /Library [same thing] until you individually delete them.... as of a month ago.

If Suno has changed that, then "bad dog" on them.

[foggy oar]

ya i will just focus one song at a time for now then to be safe, is there anyone on discord that would have quicker access to the suno team or would a email be best at this point?

i suppose since it was removed igt might just be a new ui bug or the new ui just forgot to add the button for it

[polar spade]

Can you now search the title in Library/workspace and see if the files show up together?

[foggy oar]

it doesnt work to search for any of the trashed workspaces anywhere

[polar spade]

You could bring up your issue in https://discord.com/channels/1069381916492562582/1372994553233276970. The need-help channel. They will, though, likely ask you email Support. But your main question is where is the deleted workspace folder now found, where a bin location existed before.

[foggy oar]

oh i've completely missed that channel lol, im in the wrong place then

thanks Dasvon, you're a legend for helping

[polar spade]

Youre searching for the outputs. Did you originally give a title or are they all called "untitled"? Did you have any unique prompts -- (check your finished songs) you can search prompts as well.

[foggy oar]

ya on the last ui there was a button somewhere saying something like "show trashed bin" or similar which is how i accessed it but that button is nowhere on the new ui

[foggy oar]

each workspace is just named the song title, so for example "never knew" "not the hero" and so on

[polar spade]

Search "never knew" in your Library search box.

[foggy oar]

oh wow

we're getting somewhere

songs from that workshop came up

[polar spade]

There is the trash bin icon in Library. But as mentioned, it doesnt have a "workspace" label at top.

[polar spade]

Some time ago they took away the ability to search lyrics in the search bar of the Library. So thats been the worse. I can remember some lyrics but dont always remember titles. That totally sux to scroll through pages of outputs hoping to hone in on one output I might want to go back to. 😂

[foggy oar]

okay okay we actually got progress 😄 this is amazing, so if i search for the songs name there in the library all the ones from the trashed workshop comes up and i can move the songs to a new workshop if needed, i can basically redo all my workspace and just pull the songs and stems im still using and just have all workspace not trashed

[foggy oar]

luckily i use a song title so the stems would be easy enough to find with the filter for stem in the search

[polar spade]

Yes. If only I could remember the song titles. Haha. I had a song basically titled "Catches On" . I couldnt remember the title. So... think...think. .. think.... I finally remembered i used the work "Christmas" in the prompts. Ta da et voila!

The song is in a Playlist, but I had no idea which Playlist i had put it in.

[cursive jetty]

anyone else having issues 'submitting' remix for Illangelo Remix - i amm no longer getting submit function only publish?

[graceful condor]

I am so sick of it the AI singing stuff that is not even part of the Lyrics anymore. Why do I even add lyrics and the AI decides to sing a verse or anything else from long before. This is getting dumber and dumber.

[quick sonnet]

This is a known problem.

[tepid narwhal]

The only issue i have with this is it repeats my second verse sometimes

[tranquil leaf]

I've had occasions where it will repeat verses or choruses if there just isn't enough lyrics there to fill a ~3 minute song.

as for singing things not even there... the only time I get that is if I start playing with the weirdness slider

Or miss a bracket someplace. 🙂 I was trying to get one song to put in goat bleat samples, and instead it had the singer scream GOAT BLEAT!

[fervent meadow]

Hello! I am trying to create a song from Hindu Scripture. I was able to make the same song a few days back but today it is giving error "Lyrics contained copyrighted material"! How can anyone copyright scriptures? I have emailed [e-mail] with details. Any help is welcome. Is there any way to contest the copyright?

[neat nexus]

You could try rewording the verses you're using?

Is the translation you're using under copyright or are you using one that is definitely known to be in the public domain? (Project Gutenburg, etc?)

[fervent meadow]

verses cannot be reworded as its from scripture so meaning changes. yes its well known to be in public domain but I suppose in India. https://en.wikipedia.org/wiki/Hanuman_Chalisa

The strange thing is that it worked just an hour earlier, now it doesn't and gives lyrics copyright error

[neat nexus]

I wonder what changed?

[fervent meadow]

Suno engineers would have to share how they handle lyrics copyright and if someone submitted copyright for this.

[hard beacon]

i think it will be detected via a content ID system like youtube

[fervent meadow]

Clearly lyrics of every song created by non-free users is not copyrighted, as I created the song with this lyrics an hour earlier ( and a few days back too), so it is not copyrighted in my name; so the chance that they have some other kind of copyright database which is updated from some other method; not from song making interface

[hard beacon]

well. the system is detecting also vocal parts... so if you use the same sentence as a track (for example Taylor Swift) you will get this message..

so... just change one word or the spelling and then it works 🙂

[neat nexus]

Can you maybe alter it with phonetics? The shape of the sounds not the accurate spelling of the words?

[fervent meadow]

okay trying it

strange, I change 2 words but still "Couldn't generate that.
Lyrics contained copyrighted material"

[hard beacon]

sometimes that are just 2 words of a part of the text... don't ask me why.. but that's how the system is working 😉

[fervent meadow]

works when I change every sentence 🙁

[hard beacon]

there must be a specific part in the text... do you write the lyrics on your own?

[fervent meadow]

I suppose Suno personnel will have to answer on it

[hard beacon]

but that wasn't my question 😄

[fervent meadow]

I do write my own lyrics but I am trying to make a song from the religious scripture

[hard beacon]

why i am asking.. normally i write all lyrics by my own.. if i use (for example) chatgpt for inspirations i get a copyright wanring.. coz chatgpt is combine well known lyrics..

[fervent meadow]

strangely I was put in timeout! Well, point is I am making song from material which cannot be copyrighted. Suno has to find why those lyrics are copyrighted and remove it from that copyright

[neat nexus]

Weird! I don't see anything in the logs.

...wait. I see it now. You used a restricted word. I apologize.

[hard beacon]

😄

[jade mist]

Why did the model selection get updated, now instead saying for example better song structure, max 4 min, it says basic song structure. Is this signaling v5 or something?

[hard beacon]

well... only new words to describe the models... nevermind 😉

[neat nexus]

https://tenor.com/view/soon-gif-19121692

We know it's coming soon but we don't know when exactly.

Conisder how long Studio has been coming soon and is being worked on to make it as good as it can possibly be.

[cursive kiln]

it means that they will add v5 in a moment and give an even better description to it

[hard beacon]

i think v5 will not be coming this year... i think it's coming in feb 26

now it's time for suno studio in beta.. then improvements and then next year v5.0

[cursive kiln]

nah

[hard beacon]

we'll see 😉

[cursive kiln]

studio is something else, it's a different product for more advanced users, there was a waitlist for it from the beginning and many people have had it for a long time, and the model versions are released normally without a waitlist on one day for all pro users, and it will probably be released very soon

[hard beacon]

well.. i would happy with a new model (if it's good).. but i am sceptic 😉

[neat nexus]

I like it. But then I tend to like the newer models once I get over the learning curves.

[hard beacon]

there we go... "Lyrics contained copyrighted material".. the only text is It's time to dance 😄

[jade mist]

If someone on Suno's twitter said definitely not 2 months! to a user who asked soon as in 2 weeks or 2 months, that tells you it's actually coming soon. I've learned that Suno is about the only company that actually means soon, unlike some companies.

[neat nexus]

Googling gives several results, which is not unexpected since it's a common phrase...

[hard beacon]

i know.. but claiming this one as copyright is.. well... you know.. stupid 😄

this like to claim "here we go" as copyright text 🙂

[neat nexus]

Can you get me a screenshot or something of it?

[hard beacon]

too late 🙂

[neat nexus]

I'm starting a thread on this in the mod's section because I keep hearing about this happen.

Hoping to bring it up in our weekly meetings.

[hard beacon]

if you copy a lyric-text 1:1 (for example linkin park, taylor swift or calvin harris) i can understand the copyright warning.. but.. well.. a "put your hands up" or something like that isn't a copyright abuse

let me try if the good old (RIP) Scatman John text is working 😄

[fading rover]

The word "lowlight" in the lyrics is always blocked for me, because its a producer tag?

[hard beacon]

maybe..

[neat nexus]

If you've got screenshots I can try to pass it along. Help me help you.

[lost holly]

"lowlight" and "Put your hands up" are song titles, thats probably why they get blocked with coyprighted.