📚┃suno-school

[onyx sparrow]

u mean export upload export upload?

[frozen dagger]

it's ambigous. if we say "No true peaks" maybe it doesnt understand what "no" means and reads "true peaks" so instead of avoiding it, it adds it because you named it. (thats why i dont like using it that way)

[onyx sparrow]

:/

thats a possible way

but i m a lazy....guy

[frozen dagger]

no. just create, then extend near the beginning, stacking the info it keeps all the prompt data from all previous extensions.

[sleek stump]

Yep, anything yo uput in the Styles window is interpreted as something you WANT

[onyx sparrow]

°_°

if i write -true peaks then he

in the style section

push the true peaks?

or i need to write - no true peaks

[sleek stump]

Also some advice in teh styles prompt, more is not always better. You can get so much in there that Suno just chokes and picks a couple from your list.

[frozen dagger]

(i dont think you need all this space, but if you wanted it, this is the only way i can think of to workaround the limit of 1,000 character in the exclude box) 🙂

[onyx sparrow]

paul sad

xD

[frozen dagger]

indeed. less is more. imo. Suno, take the wheel.

[onyx sparrow]

damn

all my dreams about money fame is gone now

[sleek stump]

There are a lot of people overloading too many genre and styles and they always dont really get what they expected.

[onyx sparrow]

xD

more is less but also more i more

okay tx for information i appreciate that

[sleek stump]

Smarter prompting will always get closer to what you expect, just piling in stuff doesnt really work out that well.

[frozen dagger]

don't give up on the dream.

[onyx sparrow]

no that was only a joke

i meant this sentence about fame and money

yeah thats why i need the master chief

but it is what it is

then no more space for paul

[frozen dagger]

if it helps, 5k characters isn't enough lyrics space for me (just lyrics, no prompts) Oblmous needs more space too.

[onyx sparrow]

i copied the sentence to the translator and i still do not understand itxD

what u mean?

yeah lyric space i have enough space

[frozen dagger]

5,000 character limit for the Lyircs is too small as well Oblmous (me, Oblivious) needs more space too!

[onyx sparrow]

ahhh yes

we both need the master chief

someone have the telephone number xD

[frozen dagger]

they have animation function for the pictures. that was my ask. i'm good. lol

[onyx sparrow]

xD

anyway cya

[frozen dagger]

i get on in about an hour and a half lol

[manic basin]

can i post song links for input on which style to take to my lyrics? having trouble getting suno to do a slow enough beat for the version i have in my head

[frozen dagger]

can u use a hyperlink to a song as an input? i don't believe so. you'd have to describe it, or other tricks. if wanting slower beat, try the change speed option (if available) and then try to cover?
you can use audio to direct the song with uploads if you have the rights to the music you upload. but if it's someone else's have to reverse engineer the style from sound to words.

[manic basin]

so far it did good with me specifying a BPM "range" and eventually getting it in other genres but with hiphop/rap I can't get it to go under 60BPM

and in v5 it seems to want to always rapidfire all lyrics xD no instrumentals/breaths between lol

ohh maybe I should try another version..

[frozen dagger]

you usually have to pad you lyricss with [instrumental breaks] or similar tags on your own. you could try the change speed thing it has option to keep pitch. but usually a cover still needed. Can set BPM's in studio but same with adjustments but can put things in right direction.

small tweaks

[manic basin]

i'm a 1 gen monkey i guess, i don't do anything with these songs (like post them anywhere) so thank you for a valid answer but is too hands on for my ape brain so i'll have to cope and keep trying

[frozen dagger]

in that case just try reinforcing your wording about the cadence, rephrasing. telling it the bpm range is good. anything else you can use to describe helps too. musical terms when u can:
Largo

very slow, broad

~40–60 BPM

“heavy, solemn slow-breathing pace”

⭐ Lento

slow

~45–60 BPM

“steady, unhurried cadence”

⭐ Adagio

slow, relaxed

~66–76 BPM (slightly faster than 60, but close)

“slow but expressive”

[manic basin]

i'm so interested in how the subjective prompts work (“heavy, solemn slow-breathing pace”) maybe I should try some

[undone basalt]

When you want to use a persona, how do you set the Audio Style, Weirdness, and Audio Influence controls?

I did set mine to - Weirdness: 25%
Style Influence 60%
Audio influence 100% but looks like SUNO gave me a generic voice instead of Persona's voice

[onyx sparrow]

is that possible if i write hiphop or trap or drill that i get always a choir?

because i get always a choir

[onyx sparrow]

why is there a space?

[wise kindle]

is that part silence?

if a waveform has a straight line it usually means there's no sound there

[onyx sparrow]

yes but i dont have this silence on my sample

is the sample better to upload in mono or stereo?

for more compatibility

[wise kindle]

I'd use stereo personally

[onyx sparrow]

okay

i tried 20 times to make a beat with this sample sounds not good

like a sausage

idk the word in englisch

sound like under water

ah sry not sausage

like a soup

xD

[dire swan]

Im trying to ise studio to modify the vocals. Every time, it says timed out. Whats up?

[onyx sparrow]

idk

[scenic viper]

Bonjour, dans suno studio lorsque j’essaie de crée des stems j'ai ce message d'erreur . Merci

[noble sage]

Hey, how to add [Doorbell Ringing] in the lyrics

[timber robin]

Relatable. To resolve shimmer, I EQ high frequencies for the whole track, then preset it, then sometimes a small bass boost.

[wise kindle]

it's not really capable of sound effects. your best bet is to download the song add them in yourself

[noble sage]

How about male and female vocals?

[wise kindle]

like duets?

[noble sage]

[Male]
How is your day going?

[Female]
Good how about you?

[wise kindle]

Something like that could work, but it's not officially supported so results may vary.. a lot.

[noble sage]

Is there something supported?

[wise kindle]

sure, the style prompt and lyrics parsing is pretty extensive

and it can generate multiple voices. It's just not guaranteed

the seer mentioned a clean way of trying call and response between two different singers #📚┃suno-school message

[wise kindle]

Re: the "live" pronunciation chat yesterday. I've tested lyve and l'ive and the latter seems to work great

[weary sable]

I made a song in V5 which went normally but started hallucinated at this timing: https://suno.com/s/Sl1wCBtez0TOcbGe?time=139

Interestingly, remastering it also impacts the initial lyrics which went well in the original and make the sound of the hallucination different while still following a similar structure:
https://suno.com/s/WMdqwI0VbuHSpRwA?time=139

The other variation of the same song (since the songs are created as pairs) hallucinated a bit as well and ended abruptly:
https://suno.com/s/WMdqwI0VbuHSpRwA?time=139

[polar spade]

Cool. I hope that creates a fun springboard to make a new song off of the hallucination. They are fun to make.

( Sorry, i would not know where the hallucination begins or ends as my French is quite limited).

Oh I see you timeset to the ending part where the "random" weird hallucinations are. Got it. Yes, that can happened too.

When we talk about the hallucination "songs" is where Suno actually hallucinates a full melodic output completely different from whats prompted. But sounds like a completed song in some gibberish language.

[polar spade]

Here s a small clip of a hallucinated song. It gibberish, but sounds like an actually song with a cool groove. https://suno.com/s/XHmUfpOncAGomh8y

[dire swan]

Its a call amd response. In older versions it seemed to work. I made 70 tries on 4.5/5.0 amd never worked.

[noble sage]

It works sometimes... It's not consistent

Still I wasn't able to DORBELL RING

[marsh leaf]

so I am generating some track for background scoring, I want no rhythm or beat at all which is specifically mentioned in excluded section, still suno generating rhythm full track, any tips ?

[wise kindle]

it's trained on music, not random foley sounds. it probably has no idea what a dorbell (doorbell?) is

[polar spade]

@flat scarab @noble juniper @north pawn @quick wedge

Since there were mentions or questions about any Black Friday sales:

Everyone check their subscription page for new discounted Annual plans! 🤑

Since this Thursday is Thanksgiving in the USA (and elsewhere) , November 27th, hoping everyone will have a nice holiday or enjoy a moment to be thankful.

But saying that . . . Be aware servers may be busier than normal and Support may be limited while everyone gets a little turkey and dressing to eat.

Make sure your app is updated, funds are ample in your accounts, and plan for some delays from the longer holiday weekend and influx of more users and questions to the format.

Further, any new purchases or upgrades can experience a longer bank holiday to impact approval of funds and activation of subscriptions. Rather than a typical 3-5 bank days, it may be 3-7 days, if same as last year. Thats the international banks doing that, and NOT Suno.

[north pawn]

Yes, I saw that 10 minutes ago, thank you!!!

[polar spade]

Put the sentences in quotes. That may help.

[noble sage]

It's not because of thanksgiving... It's because it's my birthday

[fallen dune]

hey how do you guys approach lyric making? this is for chill and house music

[polar spade]

Will there be cake? Or we just putting candles in the pumpkin pie?

🎂 Happy Thanksbirthday! 🎂 🥳

[noble sage]

Pizza cause i'm italian haha

[long jackal]

Hi, why can't I use my persona when editing in Studio? the singer is different to the original persona. I am a premium user account. Thanks in advance.

[polar spade]

Personas tend to reflect the nuances of the vocal characteristics of the original. It is not designed to clone an exact copy.

[long jackal]

Hi Davson. Thanks for that. I wanted to change a couple of words in the verse but the whole thing sounds very different to my persona voice. The song is unusable with that obvious difference in vocals. Shame That's why I paid for premium to keep that persona.

[polar spade]

It all depends on the upload, likely as well as to how close it my be. Is your voice already pitch corrected and mastered before uploading?

[long jackal]

Ah, but actually no. Good thought though.

[dire swan]

You more or less helped me yesterday. So how about today. I went into stufdio to do am edit. Ive tried 30 times. It keeps telling me it timed out.

[polar spade]

Just as we can pitch and master our voices in a DAW to a point it doesnt sound natural, Suno tends to overextend that , but also then adds back natural flow, breathing and the now expected AI tweaks that make it sound quite natural, albeit now not as similar to our original.

For now, you just have to find a good cloning software out there that does less AI and more or less just clones and pitch corrects the voice.

[long jackal]

Thanks, DAsvon. It's kinda of a no brainer that you would edit the lyrics with the original persona, esp when personas are a premium perk. Good idea about AI training voices. I used one that enhansed my own but yes, that is a great idea. If you do hear of any ways around this, keep me in the loop! It's a lot of effort to change 2 words. But its a process of learning, I guess. I wish I had more time to devote to learning more tricks.

[polar spade]

I have never seen a time out message. So im at a loss, other than the server maybe having an issue processing it for some reason.

[onyx sparrow]

someone know the algorythmus of suno? because i deleted some song

and...i want that againxD

[dire swan]

Ive created 2-3 other iterations. Anywhere from 4 min to 7 min. When i go to studio amd try to change tye voice stem it times out. Actually, anytime i try to change any full stem or the whole ot does. It lets me do 30-45s sections. But when i need the voice consistemt, thats a crap shoot

[polar spade]

Changing two words? Thats different than cloning a voice.

What's going on? Minimal word changes can be done easily* in the edit UI.

• easily as in the process, though getting a generated output that sounds quite what we want might take a few clicks.

[long jackal]

the Persona when those words are sung are not the same.

Hmmm

[polar spade]

Ahhh.... I havent played with stems in a week. Someone else may have experienced the same as you then, if something is amuck with the stemming program.

[dire swan]

It wasn’t just the stems. It did the same thing when I tried to remix the entire song just by changing the style prompt, but it only seems to do it in studio. I can make songs with no problem.

[polar spade]

We should always run a couple of remasters, just to see if it equates all segments to sounds like the initial vocal at the beginning of the output. It may or may not. But it evens out the vocals fo me sometimes.

[long jackal]

Interesting....you just gave me an idea....brb to see if something works...

[long jackal]

Okay so I went over just the word...that worked. EXACT Persona. but then it changed the section I didn't highlight in the trackbar.....

[quick wedge]

Thank you for letting me know about that. Yea, have seen people asking about that. I thought maybe they would’ve made any announcement about it hopefully I know what new information to use about Black Friday for when the users do ask and I can answer them correctly without messing up.

[polar spade]

Yes. It has to anchor onto something on each end in order to blend the changed words back into the output. So it may need to changed the metering, speed, inflection and melody of about 3 seconds (to a minute) worth of "voicing" left and right of your change, even though the lyrics or words themselves dont change.

[long jackal]

Thanks Dasvon, you are most helpful. I'll have a fiddle around and see if I can take out those bits that it added on. I usually go into ableton to do little changes like that and do a cross fade at those points I cut. I'm sure this might work...Thanks so much for your time.

[polar spade]

If you have :

This is the worst of times.

And you want to change to :

This is the best of times.

. . .then words to the left and right of BEST are going to need to be reanalyzed , spacing and timing becomes slightly different. Suno will try to match things before, but it may need to just generate a melody or intonation not as before in order to naturally blend BEST into the phrase..

[long jackal]

so to put it back i'd do that in Ableton, yeah?

[polar spade]

With DAWs we have had to manual splice and stretch or smash tempo and hit the spots as close possible so it doesnt click and pop like a spliced spot or sound off tempo. Quantizing may help.

But with AI its doing that over a larger area. I suppose the best term is "feathering" the edges so it has a flawless seam.

Up to you to use a DAW, first, but Suno may or may not view your glitches as a feature of the upload and ignore trying to feather it further, if I understood your question.

Understand, im also an idiot, so i dont get everything someone says. 😅

[long jackal]

Thanks, that actually makes sense. In Ableton we’d just splice/warp the clip, line up the transients, and use tiny crossfades so it doesn’t click or drift off-tempo. Quantize helps, but the real fix is manual feathering on the edges.
So what your saying is Suno is doing a macro version of that cos it' smoothing over a wider area. So if the glitch is baked into the upload, the AI might treat it as intentional and won’t feather it further for example.
Appreciate the clarification and explanation. That tracks perfectly with DAW-style workflow. No pun intended! lol

[long jackal]

Naw, I just explained myself rather clunky. Thanks for being patient with me.

[polar spade]

It my opinion that its a mindset as well. We all, including myself tend to love the first good variation of a melody, so any changes Suno does, doesnt sound quite right.

Ive often make the Suno edits and then go do something else so I dont listen to the "new" version for awhile. Two months later, I listen to then again and ofte like the new version better, where I did not before.

Maybe do corrections in Suno first, then download and make small corrections in Melodyne or whatever pitch correction you have. Thats been the easier way for me on some. Thats the same as if we sang the wrong note, anyway.

[long jackal]

I'm mashing old skool audio editing with Suno. I don't mind. But yes, feathering is the perfect way to describe it. The sound quality in Suno has improved since this time last year. It's unbelievable.

[polar spade]

Specifically, anything Suno hears as "noise " , on an upload, its default is to try to turn it into anything else than to just remove it, so splice clicks become a edge tap on a drum, an air conditioner becomes a synthesizer. Its kind of fun actually, to just give Suno the crappiest sounding upload and listen to what it changes the noises into.

[long jackal]

Hold up. You can upload audio into Suno?

Dang! I had no idea!

[polar spade]

Thats what I thought you were doing. 🙂 it may not be available to Basic users, but for paid, uploading and live recording of audio are available.

Yea.... Basic can upload/record up to 1 minute. 8 minutes for paid subscribers.

[long jackal]

That actually explains a lot to me how it's working. So Suno ,treats now sounds by reinterpreting it as musical material... that’s so wild.
Now I kind of want to test it with the worst audio I can find just to see what it morphs into.

[long jackal]

I'm a paid member since 2023

I'm clearly underusing Suno...

[polar spade]

In all seriousness, read the TOS carefully , its been update November 5th 2025, to be aware of those legal permissions you are giving Suno regarding your music/audio that you upload or record, but if you wish, try some test uploads or even just click record and ramble some words off to see what suno does wih them.

[long jackal]

Oh really?...I'll do that. TOS. I wonder if those TOS changes are due to the AI artist hitting number 1.

[polar spade]

unknown Legally, I couldnt say. 😆

[polar spade]

And thats an overstated self-perpetuated promotional pitch. It wasnt No 1 on Billboard Country charts. And wow, do those songs not sound good. Lol just news fluff to get listens. I do my couple of trips to Nashville every year. That user would be tarred, feathered and kicked out of town for even posting songs like that.

[long jackal]

You’re right. If someone played that in an actual writers’ room down there, they’d get laughed out, tarred, feathered, and put on the next flight home.
Marketing fluff ≠ industry reality

[oblique tinsel]

But these discounts don't work if you have an active pro plan and want to upgrade to the premier plan.

[polar spade]

Yea? Ill check mine. Be right back.

[polar spade]

You seem to be on the Monthly panels / No discounts for monthly. The discount is for Upgrading to an Annual plans.

But as for Annual, there does seem to be a glitch. They discounted the immediate prorated upgrade but as for the ugrade at end of the present period it is still regular price. They should look into that. Ill bring it to their attention.

[oblique tinsel]

True, the screenshot is monthly, but as you say, no discount is applied to the annual summary.

[polar spade]

Not exactly what I said.

The promotion shows on the Annual panels.

[crisp glen]

Hi i have an strange question i need to renew my sub the 1st of dec. And it says $4.80 / month thats for a year is $57,60. But when i go to my payment page it says $120.00 / year and $24.00 is the WAT.

[oblique tinsel]

I don't see the special offer signs either, like you do.(limited time offer)

[polar spade]

. . . Only that, at checkout, the immediate upgrade subtotal shows correct, but the subtotal for the optional "change tier at renewal date" remains at the regular price at the moment.

[crisp glen]

oooh dident see that you guys talked about this

[polar spade]

Wow. Yea, youre right. It may be a slow roll out. Im looking at mine from a present monthly Pro subscription on my Android (desktop browser).

[long jackal]

I was a on the premier plan and stopped for a month and went back in. Is that why?

[wary basin]

Question: how does one get suno to use only a single instrument? Cause I'm trying to do something like a primitive Orcish war chant for DnD

[polar spade]

Is what why? I've canceled down to a lower tier one month and back up the next before.

[long jackal]

I just checked my member status. I'm def on the Premier version.

[polar spade]

Is this a good thing? I may have skipped a post somewhere. 🙂

[stuck junco]

Hi there! Thanks for reaching out about this!

When you get a chance would you mind DM'ing me your suno handle (--eg @example_handle) and a song link to your account. Would like to look into what's going on here

Thanks for drawing awareness to this!

[polar spade]

Very generally -- and one of maybe many ways to do this -- you could prompt for that vibe in styles prompt box. Even just one or two prompts such as "solo taiko drums"

Do the same as a metatag in the lyrics box.

Listen to find some solo areas or even stem it into tracks to separate the track best suited.

Segment and crop the segment wanted. Make a persona of the solo.

Extend from that segment while attaching the persona to get more chances of solo events. Repeating to add more extensions and sometimes that may be enough to get a couple of minutes. Remaster when done..

https://suno.com/s/YyzbxuY1OXo1KX8R <--- example using this technique for geting a solo bass run on ver. 4.0

[wary basin]

Awesome sauce!, thanks!

Also

How do I prevent the orchestral humming/vocals/choir that V5 seems to love adding to the start of everything?

[polar spade]

unknown solution from me. Users are coming up with different ways. Certainly with Studio we have workaround by just splicing/stemming what we want and dragging in where we need.

But normally we could add more tags like put in [interlude] shortly after the Intro, so we could crop out the intro, as a quick fix.....

or..... section the intro out and try to replace it in edit UI now with the interlude to its right to influence it to not have vocals. This i have not tried yet though. So I wouldnt have any examples of it working as a solution.

[compact horizon]

I upgraded to the premier plan but cannot get stem separation to work…just get the spinning wheel of doom, any clue?

[half dragon]

Could be UI refresh issue - leave and come back in and see if they're there?

[compact horizon]

Thx will try that today

Support seems to be non responsive

[worn thistle]

give them a little bit more time, It took me 4 business days to get a response.

[compact horizon]

I’m at 2 weeks almost
Love the program
Any good links on promoting you know of?

Prompting that is

[quick wedge]

For prompting tips and tricks you can check the articles in the workflows channel https://discord.com/channels/1069381916492562582/1252267444773326848 see if there’s anything in there you wanna experiment with for your future songs

[hard beacon]

@compact horizon : you can also check the suno page, exploring the genre you like and then take a look to the prompts that were used.. looking and learning 😉

[compact horizon]

Excellent thx. I would like to create an arrangement template like I have in studio one any one seen that? I’m new to Suno but not recording

[hard beacon]

Just to give you the best answer.. what plan are into?? free? pro ? premier ?

ahh.. sorry.. read it the lines above 😄

well.. you can't compare suno studio with a regular DAW with premade templates... it's more like a stem editing tool and of course new AI-generated layers

[worn thistle]

hey, do any of you have some persona with: deep bass, fast beat and actualy good female vocals?

[quick wedge]

You’re welcome also welcome to the Suno community too. I’m not familiar with the music template recording wise but you can ask in the Studio channels and see if they might know about it https://discord.com/channels/1069381916492562582/1420792719969878028 also here’s a road map of the server’s channels if you’re not sure on where to go https://discord.com/channels/1069381916492562582/1420465001524887604

[hard beacon]

You can get a lot of infos about suno studio on the official suno youtube account: https://www.youtube.com/watch?v=qR4BefPvSiI

There are also advanced tutorials

[long jackal]

Augh. I can't get this damn word to change. It's taking over the whole verse just for that one word change and changing the lyrics.

[oblique tinsel]

ok, send

[hard beacon]

I think Mike needs a 48 hour day to handle all his DMs and requests 😉

[stoic kettle]

Hey, anyone knows why there is no option for exporting midi?
noticed that some people have that option and i dont have any.

[stoic kettle]

I dont have it

[hard beacon]

even if you are on pro or premier?

[stoic kettle]

yes

[hard beacon]

hmmm..

[stoic kettle]

I have pro

So confused right now...

[hard beacon]

maybe midi is only available for premier users...

in the compare plan section there is no info about midi..

[stoic kettle]

The devs should let us know about this.

[stoic kettle]

Midi is super important and everyone should know about this.

[hard beacon]

let's find it out by ourself... Anybody online with a pro plan ????

[stoic kettle]

you mean Premier plan, i guess
because i have Pro and i dont have it

[hard beacon]

no.. pro... wanna check out if other pro plan users also can't generate midi

[stoic kettle]

oh ok

gotcha

[hard beacon]

i have premier and i have it.. that's why i want other pro users to check it out 😉

[half dragon]

Yeah Studio is for Premiere and also MIDI

[hard beacon]

ok.. there we have the answer 😉

thx buddy

[stoic kettle]

Damn it...

@half dragon thx for the help!!

[hard beacon]

i''ve never checked the midi export before.. don't know how good it is..

[stoic kettle]

me neither, not in the mood to write midi from scratch 😂

Thx @hard beacon

[polar spade]

Sorry, I didnt ask. . .

Did you "align your lyrics" to make sure the magenta colour highlighter crosses over the lyrics at the point they are sung? It will follow the lyrics like karaoke.

Upper left of either editor, but I often use Legacy for this.

Align first before doing any editing and before each edit, check the alignment.

[hard beacon]

even after spending round about 100.000 credits i still don't know how to optimize a good track.. remaster or cover... it's a fight against the AI at all 😄

[half dragon]

I don't think there's one way to do it - don't fight it work with it 🙂

[hard beacon]

that's why i've spend these high amount of credits 😉

[hard beacon]

oh... i've just say it and voilá... there is the output i want 🙂 clean mixdown (perfect for mastering) and the lead synths i want.... well.. only 3000 credits to get it ^^

[compact horizon]

Thanks everyone very helpful

[hard beacon]

you are welcome..

[ashen dock]

“I’m trying to create a 3-minute rap song where Suno.ai only sings the chorus. The chorus should appear at the beginning, in the middle, and at the end, while the rest of the song is just the beat without vocals. I can’t get it to work and I’m really wondering how or if it’s even possible.”

[frozen dagger]

id create just a gen of the vocal. when i had something i liked id create the instrumental, then id place them together, either tracks in studio, or just splice the singing into the instrumental, and then cover that.

[tired forge]

Whomever renamed a song of mine to Drowning from whatever dumb name it was I wanted you to know that I posted it on my YouTube channel.

[ashen dock]

How do I create a vocal track that sings without any beat?

[frozen dagger]

you could try prompting tricks to do that, but easier way is to just split the vocal stems to get the vocal only to add later to the instrumental you want

[compact horizon]

Being new to Suno I seem to be having an issue getting stems…premier version on gives me the eternal spinning wheel of doom, any suggestions…thanks

[shadow rain]

Does Suno have an API or ability to automatically send lyrics and hyper parameters to the model and get an audio track (eg .mp3) or video with caption / sing-a-long lyrics (eg mp4) version back? Or is there a similar structure that is possible with the model?

[half dragon]

No API.

[compact horizon]

I think I found them

[ashen dock]

“That’s going to take years and 10,000 credits to get the right result. Separating the vocals from the beat creates noise, so that’s totally out of the question. I’m going to cancel my subscription and wait until Suno releases a proper chorus feature. Otherwise, it’s useless to me.”

[frozen dagger]

more than one way to skin a cat. i don't know exactly what you have already or what you working on. just tosssing ideas at ya. studio add track to the intrumental you already have and just highlight the section in the new track where you want vocals and add them, it will use the beat as the reference.
Others more versed than I may have other, better, suggestions as well. The tool is generative tho so gonna be some credits spent seeing what it do.

[flint haven]

Beware: there is NO official Suno API

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

[shadow rain]

Thanks understood. And to check there is no other way to say hook this up to a tech stack / production pipeline right? (Eg we would need to use say an open source model like a music version of Llama etc )

[junior holly]

yea you might consider open source models and do your own training

you should have multiple GPU though

[half dragon]

That's right, currently you can only create songs via suno.com and its mobile apps.

[junior holly]

heads up.. training takes weeks

[fluid musk]

Hi All - I'm trying to add a Suno vocal to an already completed instrumental track that I uploaded to Suno . . . I added the lyrics in the prompt - but can't find the "instrumental" toggle switch - to let Suno know not to create an instrumental - but just add a vocal to my completed instrumental track - I can't find that "instrumental toggle" - where is it? - - - I followed the directions exactly and it says to turn off the "instrumental" option after "create" etc . . . but still don't see that "instrumental" option. Thanks!

[frozen dagger]

in teh create tab when the audio is added there is a dropdown, is this what you're looking for?

[half dragon]

There's no longer the instrumental switch or toggle. Instead, you'll find Lyric Mode (under Advanced Options) : Auto or Manual. If you have lyric then just use Manual Lyric mode.

[fluid musk]

Thank you - I had no idea there was no longer an instrumental switch or toggle available- I’ll try the manual lyric mode.

[sleek stump]

I don’t know if this still works, but you can get stems (or do a 1:1 cover) for the uploaded instrumental, then use the “add vocals” option with it.

[sharp kettle]

i laughe so hard but wasited my daily credits.... but apaparently the ai bot with this prompt and my voice thinks i sound like green day lol

1960's Rock & Roll, Punk Rock, Dance Rock, Baritone, 140bpm, make sure the music genres are blended
Use D major G, major A major and B minor Drumbeat tempo and rhythm,
Use the D major chord, G major chord, A major and B major chords mix.
Hand Drums Should provide the main music. The electric guitar and bass should provide backup melody and rhythm.
Hand drum stem, electric guitar stem, bass guitar stem.
vocals stem.

[polar spade]

It may help to let the room know your setup. Are you Premium and have Studio? Or Pro?

[ashen dock]

PRO

[polar spade]

How does "drum stem, guitar stem and bass stem" work as a prompt? What's the expected result?

[sleek stump]

Suno would disregard most of that prompt.

[sharp kettle]

2 poor for for pro right meow but eventual im using free all.. i thought using my daily credits was big hint lol

[polar spade]

I had asked that to another "blue" avatar user, but hey, its good to know everytime someone asks a question in here. So thanks.

[sharp kettle]

I was just testing a prompt.. yo see if it could combo some rock generes. but the ai model seem to like to replicate green day, good charlottle, simple plan, and it did fall out boy, the last song was def my chmecale romance... but apparently i sound mostly like green day

[sleek stump]

Suno more or less sees that prompt as...
A fusion of 1960’s Rock & Roll, Punk Rock, and Dance Rock at 140bpm, with driving hand drums at the forefront, blending vintage and raw punk rhythms. Electric guitar and bass weave syncopated backup riffs using D major, G major, A major, and B minor, creating infectious grooves. Hand drums propel energy throughout. The baritone vocal stem delivers with punch and swagger. Song moves between high-octane verses and catchy danceable choruses, each stem clean for remix potential.

But, Suno doesnt create stems in its initial rendering, thats a completely different process.

[sharp kettle]

well thanks thats good to know but i still thought it was hilerious that i sound like the green day singer.. he was very clear he hates ai though...

[polar spade]

Billie Joe Armstrong, if we're curious

[sleek stump]

Most established musicians and vocalists hate AI lol.. its just how it is for a while longer.

[sharp kettle]

im bad with names i bary rmemeber band names autism and all that

[sharp kettle]

i don't understand the hate foir ai its been around since 1954 and was 90% intregrated in almost all of societies tech by 2012...
the modern version is just sky nets momma the first version was the grand mama

if you do anything creative on a computer there is a 90% chance you are using an ai model of some kind..

[sleek stump]

If you are referring to algorithms, sure, but what we call AI is light years beyond what algorithms do.

[sharp kettle]

ai render assist
ai predictive model
ai photoshop model come about in 1984 wasnt wild availble until 1990
microsft ai algorythims are use in OS i forget the date on that one

[onyx sparrow]

today i get only hardstyle beats yesterday a short time rock this is a very strange algorithm

[sharp kettle]

holly would has been using ai program since , but i don't actually know all the models or year except for cgi and audion enger software lol

[onyx sparrow]

is this real ki or only a collection of randomized generated sounds

i use a trap prompt and i get techno hardstyle beats

thats funny

xD

[sharp kettle]

finding the right prompts for certain song styles is it or mist i had to rwerit my kat song like 50 time befor i got the beats i wanted

[onyx sparrow]

its like a slotmaschine to get 3 cherry...

xD

[sharp kettle]

New Horizens Nation national anthem 2023 it got the prompt on the first try but it took like 10 song generations to get a nice tempo

[onyx sparrow]

....

[hollow hare]

It helps if you get more specific like specifying the correct subgenre (using established names rather than what you think the subgenre is called), the year, mood, etc.

The more of these details, the more likely the AI will narrow the learning pool, which typically improves clarity

With the predominant terms used first in each set. Actually, not sure if we even have sets anymore since transition to the long prompts, hm

[sharp kettle]

i find it help to narrow down the pool if you pick a year of the genre if you wanta specific era of music... my baraitone is inspired by elvis so i often have the 1960s rock and roll prompt

[onyx sparrow]

its enough to write elvis or also the titel?

[sharp kettle]

if you use 1960s rock and roll pioneer in 4.5 all. it often give an ai render of his voice lol

[polar spade]

What are the good points of the ones youve created so far that still have music in the rapped parts?

Do you want ANY thing else besides a voice? Eminem for instance is going to have a very basic rhythm guitar and beat with the rap section , then an orchestrated chorus.

Since we can't paste separate tracks as in Studio, we are left with lucky prompting or specifically extending each section with prompts that give just rap or vocals/music.

What are you having the best luck with?

Things you may already know: lyrics for rap work best (for suno to not add music) when it is written in continued sentences or phrases, and have recognized rhymes thst can help suno set where the meters divide. Writing in 8 to 12 syllable lines, and stacked like a verse, signals Suno to sing it.

Quotation marks help distinguish spoken sections, but not necessarily.

Capitalization is fine to start a sentence, but try without periods/full stop so Suno decides where the breaks and breaths go.

[onyx sparrow]

okay

[main hazel]

so i generated a suno song, but instrumental part is missing, some singing parts i want to make them as instrumental melody, so i keep typing in "instrumental" in promt - extend song option, still no melody

what can i do?

[sharp kettle]

for rap music you need 1980s rap, or 1990s rap, in the style of 1980s rap artists oe in the style of 1990's rap artists..

any thing outside that two tends to generate rap crap.

you also need to make sure you lyric are in poetry form and using the correc [line read] .....
i dont do much rap generations so i don't know the correct bracket forms for correct reading/ singing

i usally ad sing like your reading a poem, and rythmic 4 beat count lol... it worked fairly well

[half dragon]

Let's see if I could understand what's going on here.
You generated a song, but some instrumental parts in the song are missing? meaning it wasn't generated enough instrumental breaks between the lyrics?

[polar spade]

To save confusion, i'll step back from continuing. But leave what ive posted to be considered and tried. Hope it helps.

[main hazel]

Not exactly. The song itself is fine – I generated a track in Suno that I really like with vocals.
What I’m trying to do is this:
I want the same vocal melody to appear again later in the song, but as an instrumental-only part (no singing), like an instrumental hook/outro that repeats the melody without vocals.

[half dragon]

Ah, now that's clearer - so you like the part of the singing melody to be played later in the song as instrumental.

[sharp kettle]

whoops didnt mean to do the whol thing sorry hit the wrong key

[polar spade]

Are you Basic or Pro?

[main hazel]

pro

[amber oasis]

i need to download stems and wav when im a pro member i cant

[sharp kettle]

so i actuall do have the problem as well some of my song have two or more instrument breaks like this one.. its a nation national anthem.. but it often skips this break right here

The businesses are regulated so they can never-ever:
Steal away or un-stabilize your eco-nomic life.

[Interlude Instrumental]

New Hor-rise-zens Nation founded in twenty-twelve.

(esc liffer free 4.5 all for now lol)

[half dragon]

That would be difficult via prompt.
Here's what I would try but require Premiere sub 🙂
You'd have to have Studio to attemp this or your own DAW.

1. Cover the song, but clear out the lyric, leave the song structures though, like [Intro].. [Verse]...[End] to generate the instrumental versions
2. Then extract stems, bring them into Studio
3. Cut and slice into the previous song to add the melody instrumental part
4. Export to new song.

[main hazel]

i do have fl studio tho and can also isolate tracks from vocals

but that vocal has the melody i want, sadly not the instrumental

[half dragon]

Yes, the trick is to generate the instrumental version of the melody then cut and slice it into the other song

[safe hearth]

Any advice for getting relevant usable generations in Suno Studio? I've tried everything from simple language to AI generated prompts but it's like i'm typing wearing boxing gloves.

[sharp kettle]

im really good at geeting most prompts to work. post an example. ill do a quick rewrite and tips, since im boored. try it out and see if it works

[half dragon]

I always use [Melodic Instrumental] or [Instrumental Interlude] and they work more often than not.

[sharp kettle]

oh so just switch around the wording to dyslexic version got it ooh melodic instrumental thats a new one . what breaks would you recomend if i'm make a song for exotic dancer

[main hazel]

are there any good tools online for that? as midi song maybe? or even better?

[onyx sparrow]

is the version5 limited in instruments? synths?

because its beta

?

or has v5 full access to all sounds?

[safe hearth]

A full, sorrowful string orchestra playing warm, sustained legato lines with deep emotional weight, High violins carrying long, aching melodies; mid-strings supporting with gentle trembling motion; and low cellos holding heavy, tragic chords, Immediate emotional intensity, like a heartbreaking final moment. No buildup.

[half dragon]

To generate the intrumental version of the song? I'd use Suno Cover feature like I list the steps above, have the Audio Influence slider high around 80%.. The generated arrangement might be different but melody and beat might be as close so you could use it as part for the other song. Might take a few tries but it's option I could think of.

[half dragon]

I honestly don't think it matters; they're not genre based tags

[main hazel]

any other promts than instrumental needed?

[half dragon]

Prompt to tell Suno generate the instrumental of the verse melody parts? You could get lucky [Melodic Instrumental] but I don't know a way to prompt reliable for that, almost impossible.

[sharp kettle]

orginal
"A full, sorrowful string orchestra playing warm, sustained legato lines with deep emotional weight, High violins carrying long, aching melodies; mid-strings supporting with gentle trembling motion; and low cellos holding heavy, tragic chords, Immediate emotional intensity, like a heartbreaking final moment. No buildup."

Step 1. you need to add a genre.
like orchestra
step 2.
there are billions of orchestras to pull from.
so you def ned to get specifc in the year you want pulled from
step 3 added a sub genre of orchstra music helps this prompt is def operatic and tragedy
step 5 your prompt is actuall quite good. just a few changes

changed
A full, sorrowful string orchestra playing warm, sustained smooth and connected notes, deep emotional melody, High violins carrying long, sad melodies; mid-strings supporting with gentle trembling motion; and low cellos holding heavy, make you cry sad chords, Immediate emotional intensity, like a heartbreaking final moment. No buildup. pull from 1986 orchestra, operatic, tragedy melody.

try that and let me know the results.. i had to look up a few words but it was actually a mostly good prompt...
sometime ai 4.5 all up has trouble with fancy words or song terminology

uou can subsitied the year for which ever year you orchstra heard is based off of

[safe hearth]

Thanks, i'm gonna try that. I had it set to orchestra. Main problem was that it's the final(intense) chorus and you could fall asleep to the results.

[sharp kettle]

i reccomend 1986 because that is miceal crawfor and phantom of the opera and im assuming that the kind of heartbreak orch you want

use danny elfman years btw if you want epic orcstra music lol

[sharp kettle]

any one got some prompt tips for good halloween music i'm haveing trouble writing the songs though maybe some sample of hallowen musics might help me lol

[quiet solar]

That's going to be very difficult just using prompts.

If was going to attempt that through prompting, I'd do an Extension from partway through the song and take a crack at a few different variations of [Instrumental Verse] or [Instrumental Chorus] in the lyrics depending on where your melody sits. With luck, you may get something workable. Since you don't have Studio, you'll need to take that workable instrumental to FL Studio or Audacity or something similar and insert it into the original song wherever you want it. Doesn't matter if it's scrappy, since the next step is to reimport that edited version into Suno and run Cover or Remaster on it with a low-to-mid audio influence to try to get Suno to naturally clean it up.

If you're going to do that exercise anyway, I'd honestly suggest humming or non-singing vocalising the melody yourself and then doing the same thing - inserting it where you want it in the song, and inviting Suno to interpret that into a Cover/Remaster based on the melody it's hearing. It can be surprisingly good at that, and that avoids headaches with prompting.

[hollow hare]

You can check out the two recent Halloween contest entries for ideas:

• https://suno.com/playlist/945356ce-bac4-443e-aa11-19e5613acdd6
• https://suno.com/playlist/8f603965-54fb-4b2a-988f-791ee89eebeb

For the most basic prompt I used to do something like: Dark Halloween Cinematic thriller, eerie theremin

Of course, it depends heavily on the genre you want, but having the term Halloween tends to give it that Halloween vibe (and allows theremin to appear more readily, at least in past versions)

[sharp kettle]

thank you so much i wasnt getting much results with the traditonal prompts like spooky vibes, chillers and thrillers

[hollow hare]

Oh right, should mention that the first playlist is based on a remix contest whereas the second is not

[sharp kettle]

my brain is not in hdf mode so ill probally do the one 160 list for today but i booked marked both for laterz meowsa

[fluid copper]

Is there ANYWAY to stop this (I'll explain it the best I can):
Female singer. Gets to last line of a verse or a chorus and goes into cutsey innocent or sultry voice and the music ofter stops with maybe a bass drum.

(the music stops as she sings the changed voice)

[sleek stump]

Likely has to do with the genre you are trying to generate. Suno tends to do what its trianing was based on. Using additional genre/styles can change the way vocals are delivered.

[fluid copper]

Style is 70's Hard Rock, W=0 / A=100 / S=100

Would it take something like: "No musical stops"? Then there's the voice change

[sleek stump]

Are you trying to cover a song or an uploaded audio?

[fluid copper]

Covering one of my songs created here.

[sleek stump]

Protip: Never use 100 in any category, Suno tends to do really odd things when you set sliders to 100.

[fluid copper]

ok.... I'll try that 🙂

[sleek stump]

My goto are W=60 S=80 A=85

What model are you using?

[fluid copper]

v5

[sleek stump]

Try to describe how the track should progress in the styles window, 4.5-5 benefit from more descriptive prompting.

[fluid copper]

okay. I'll try that after dinner 🙂

Thank you

[sharp kettle]

Is there ANYWAY to stop this (I'll explain it the best I can):
Female singer. Gets to last line of a verse or a chorus and goes into cutsey innocent or sultry voice and the music ofter stops with maybe a bass drum.

[verse 1 female singer/ describe tone, / sing line rhythmically or melodically](/ no bass drum/ no fade out- negatives sumtimes works but it generate wonky effects) in lyrics
try that in v5 i dont have it so i dont know if it will work

and dont forget to add the back up instructions in the style prompt.. almost forgot that

[quiet solar]

@fluid copper I'd be wary of including a negative prompt in the lyrics box: "no bass drum fade out" is very likely to be interpreted as "bass drum + fade out", since Suno often doesn't read negatives.

That aside, @sharp kettle 's suggestion is good - if the tone of voice is part of the problem, prompting a different emotion for that line in the lyrics [e.g. rebellious, intense] may help to steer the model to something different.

[cloud plank]

Add “Uruguayan Plena” as a supported style
This is a rhythm from Uruguay with specific percussion: zurdo on beats 1 and 3 (“a tierra”), tambor piano steady base, repique calls, reversed cowbell pattern, and a specific hi-hat charleston. Current Suno outputs tend to resemble salsa or folkloric rhythms, not plena.
I can provide audio examples if needed.

[sharp kettle]

I used [ latino love song voice, or rock and roll pionner tone] it sound like elvis and ricky martin]
no i dont have any of the generations i was just playing with the system.. it works well when you prompt it to the voice type and way of singing most of the time

i personall use voice cover and pros can have a full song with audity and picth perfect you can basically get any ai voice mimiced under the cover feature

i use it to test my soprano voice.. couldnt tell if it was man sing in soprano my buddies thought the song was a women lol

[sterile yacht]

#🐣┃suno-showcase message an experiment… any suggestions what could make this work better (free user)

the emotions seem to be picked up okay 🙂

[frozen dagger]

looks like did a good job already. would take the vocal and literally drop it the a couple semitones with an editor like Studio, or some other way, since free plan. (The stage play stuff with Suno is fun.)

[slender reef]

hey guys, [instrumental only]
is there a prompting way to control when each instrument starts & the chord / progression they play? if there is, please guide me or give me an example. thanks

and can i create an album based on a concept / theme & story?

(im on premier plan)

for example, can i use custom mode and use meta tag in lyrics to control the arrangement of the song?

[wise kindle]

I'm still testing it out myself, but from what I can hear this seems to work in the lyric section

[Time Signature: 6/8]
[Tempo: 168 BPM]
[Chord Progression]
Intro: E-B-D-A-B
Verse: E-D-C-E-D-E-A-G#m-A-B-G#m-C#m-A-B-E

You can also use [Key: D Minor] and roman numerals for the chord progressions if you're more musically trained

[slender reef]

yea im a producer myself, i use suno to testing out ideas & find inspiration. but it seems hard to control

[wise kindle]

yeah Suno is the perfect tool for music producers (especially with Studio). due to the random nature of AI it can be tricky to get exactly what you want, but with enough persistence you can generally get there

sometimes it surprises me and gives me something I didn't know I wanted

[slender reef]

have u tried controlling instrument? eg: i wanna introduce the guitar after the first 8 bars of piano, how can i do it?

[wise kindle]

again, still testing this instrumental stuff out myself, but I've been playing around with

[Intro — Instruments: solo flute]

[Verse 1 — Instruments: strings, hand drums, flute]

with varied results

[wise kindle]

just did a remix and it threw an electric guitar in there, so yeah, definitely not 100% lol

[dire swan]

My beta has updated to v1.1.0 amd now doesnt work. Wont play songs

For studio

[dire swan]

Songs load and play in create mode. But not studio mode

[half dragon]

I'm in Studio and music still plays for me
What browser are you using?
Sorry I assumed you're on browser - you mentioned your beta has updated - what exactly is your beta? mobile app?

[half dragon]

Suno is having Black Friday Sale... just so you know.
#📢┃announcements message

[onyx sparrow]

yo guys whats up

if i want double time rap i need to write it in the lyrics or in the style?

[frozen dagger]

generally speaking place in style for apply tto whole song, use [brackets] in lyrics to have it apply to song where you want it in lyrics

u can tell it in style to do it to a specific section in style too like i.e. Double Time Rap during Bridge

but [Double Time Rap] in the lyrics section where you want it is usually what u want

[onyx sparrow]

okay tx bro

sry forgot to accept the friendship

request

i tried to transform a german rap to a american style rap that sounded so funny

xD

"funny" -> not good

“Write highly lyrical rap lyrics with double-time flow, complex rhyme schemes, dense multi-syllabic patterns, and fast technical delivery.” i added this to the prompt he dont do it

my god lol gg

okay looks like i m to dumb to write prompts i give up

xD

†

https://tenor.com/view/rip-brah-rip-rest-in-piece-rest-in-peace-brah-gif-27633084

[frozen dagger]

remove the "write highly yrical rap lyrics" put that in part in the lyrics generator if you do not have your own lyrics and it will generate highly lyrical rap for you lol. the rest of the prompt stay in style. reinforce with what you looking for. It doesn't know what it is just word association.
Tempo & Speed

prestissimo, allegro, molto allegro, vivace, double-time flow, triple-time flow, rapid-fire cadence, high-BPM delivery, accelerando passages

Rhythm & Cadence

syncopated rhythms, staccato phrasing, semi-staccato articulation, percussive syllables, machine-gun cadence, rapid-syncopated cadence, polyrhythmic phrasing

Flow & Delivery

chopper flow, tongue-twister delivery, high-density bars, dense internal rhymes, rapid multisyllabic schemes, nonstop phrasing, continuous metrical flow

Articulation & Technique

crisp consonant attack, tight diction, clipped syllables, breath-controlled phrasing, micro-phrased delivery, syllable stacking

Energy & Feel

aggressive momentum, kinetic vocal drive, hyperactive phrasing, adrenaline-tempo delivery, high-intensity vocal rhythm

[onyx sparrow]

°_°

https://tenor.com/view/spongebob-rip-sad-sponge-bob-stephen-hillenburg-crying-sponge-bob-squidward-gif-12966776

[quiet solar]

@slender reef @wise kindle I'll be completely honest, I've seen variations of this approach for feeding chords into Suno come up a few times and I've never known it make the slightest bit of difference to the output.

I just ran six tests on an example song at 80% style influence, with the prompt in both lyrics and style boxes, and the only thing Suno consistently picked up in all 12 generations was the tempo instruction. The key was consistently different and the chord progressions bore no relation to what was prompted - and it wasn't an unusual chord progression for the genre, either.

I suspect it's too detailed an instruction for written prompting, and that Suno isn't able to look it up effectively in its musical library. So it's something you'd need to feed it as an audio prompt if you really need the key and chord progression to be exactly as you want it.

[wise kindle]

yeah the key is always off by a step for me whenever I use chord tags [E] or the [Key: scale] method. without properly anaylzing the chord progression I can't say how accurate it is, but repeat generations yields a similar sound, so I'm guessing it does something. Writing a quick track yourself and covering it with Suno is clearly the ideal choice

I think the time signature instruction works too (also sometimes)

[polar spade]

I agree there is something there to saying that adding specific helps to nudge nuance. But has never resulted in anymore than 20% of it following such prompts, usually by shear "like vibes" . Prompt for pop and suno will often output a song in the 100bpm range anyway, regardless that we put "100 BPM" as a prompt. Yet we think we nailed it.

The trick for me has been either changing variables , such as prompts, literally generating for several dozen runs and compare all to find, for instance, the one thst hits the right key, or just do, whats likely, Sunos limit for now, then pitch correct it offline or realize its fruitless for such set of prompts and move the momentum another direction.

For months I would take someones massive prompting , whittle them down to about 5% of the volume, ending in 5 to 20 prompts, and the result were nearly the same.

Its a needed chore for some, so if it helps, go for it. Many say , "i just like making songs", so its not about the commercializing or pitching the song, but the fun of creating one with 2000 characters of prompts. If its a win for them, thats what works for them. Well done.

As it were though, like covers and personas, it (long JSON STYLE prompting ) which works well for AI video and artwork, does physically take up space when making new outputs in generative music. Sunon still pulls from all sources in the chain. That hasnt seemed to have changed. Find smaller broader prompts that when combined work just as well and leaves room for Suno to be creative.

[wise kindle]

"like vibes" is probably the best way to explain the results. I haven't gotten it to the point where it's "that's exactly it" when listening to them, it's always "ehh it's the kind of vibe I was going for, but this isn't what I wanted". Like, if I ask it to use phrygian dominant I'll get a slightly middle eastern feel to the song, but it's not fully phrygian dominant. If I ask for Lydian I often get that floaty fantasy vibe. It's almost like the scales are vibes, not actual scales. Keys are often close - usually off by 1 or 2 semitones.
Still, it's interesting to experiment.

[open jolt]

Guys I'm a pro user and I'm happily coming back to 4.5-all cause it's more creative and follows my old prompts and get my style of music. Suno V 5 can't. It gives very mainstream music and can't understands my music and what I want while 4.5-all (not the pro one) nails it perfect. It gives more creativity, more freedom with the music, the vocals. It actually TRIES stuff. Nevertheless, I'll cover some with V5 for the sound quality.

Give us V5 sound quality with 4.5 for creativity and musical freedom, then we got it.

[wise kindle]

Totally valid approach. Some people have even gone back to 3.5 because they liked the generations better then just remastered in the newer model

[open jolt]

Yeah. It was more surprising

[cyan lantern]

I'm sure this has been asked before but: Is there any 'predictable' way to prompt for oblique vocal harmonies at specific points in my lyrics? It seems pretty hit or miss and never really where I want them.

[quiet solar]

Not really, no. All prompts are essentially suggestions rather than commands, but anything involving multiple vocalists can be particularly troublesome.

The best suggestion I can give is to include that you want vocal hamonies in the overall style box, and then try to prompt for specific points within the lyrics box using style prompts, e.g. [harmony] [lead and backing singer duet] or other potential variations just before the line you want harmonised.

[cyan lantern]

Thanks for this. I've been having a heck of a time with it. I've been prompting for the oblique harmonies in the style box already. But every time I do what you suggest in the lyric box, I usually get the words oblique harmony whispered in the song... lol. I'll keep at it though. Thanks again for the response.

[wintry folio]

As anyone been able to create a "radio dj" personality? short radio bumps/intros/outros?

[quiet solar]

If you're putting the prompts in [square brackets] in the lyrics box, typically Suno shouldn't be singing them - it does occasionally get confused, but it shouldn't be happening multiple times.

[cyan lantern]

lol. It happens almost every time tbh. But it's all good. I'll just keep trying. I do use the [] to put that 'prompt' in the lyrics as you suggested though. It's all good. Thanks again for responding.

[quiet solar]

Sort-of.

Lizi Annem shared an example of a radio short that you could take a look at. Apologies - I would send a direct link to it on Suno, but the automod had issues with that. Should be easy to find a link if you Discord search for it.

However, Suno's not really designed for producing a punchy, coherent 5s-15s jingle of the type I think you're after. So you might want to expect this will involve some editing work, rather than just being something you can bring into existence with a prompt and a click on the Create button.

[hard quest]

Hi. I need your help. I'm trying to subscribe and I'm approving the subscription through my bank's app but it's still not allowing me to subscribe. I have enough money in my bank account as well, so that's not the issue. Anyone experienced something similar?

[manic basin]

is there a way to prompt music samples that's public domain due to age?

[polar spade]

5.0 has been decent in doing this. But 3.5 , 4.p , 4.5 all can do their version. Usually its more of the whole package voice and bump at the same time. But Ive done a NPR -like soft DJ voice on its own as well.

The trick is to attempt hitting exactly 5, 9, 15 or 30 seconds. But often get 22 sec or 58 seconds etc. So some have to be spliced/cut down.

But all that said, watch a few "Radio DJ Dude" on youtube and get some nice ideas from him. Thats my lazy way of suggesting a how-to.

https://suno.com/s/55nTY1UvVQZ8febr

[lean crest]

Hello everyone!

I am trying to extend an audio I uploaded (instrumental), an audio of about 2 minutes, when I try to extend it by including the song lyrics, a large part of the lyrics is ignored in the extension. (The lyrics are a bit long, but it works normally, except when using "extend")

So far in v5 this is the first time this has occurred, but in this combo of the same audio + same lyrics, it ALWAYS ignores the lyrics, regardless of the extended part. I tested extending from more than 30 different points in the audio (e.g. starting from 1:00, 1:10, 1:20, 1:30, etc...)

I don't know what to do. I ended up using the cover instead of the extend temporarily and it solved the problem, but I really wanted to be able to solve it with extend!

If anyone has any tips, they would be welcome.

Lyrics structure below:
[instrumental intro] [intro] [verse] [verse] [verse] [pre-chorus] [instrumental] [chorus] [verse] [bridge] [verse] [verse] [verse] [final instrumental]

[polar spade]

What do you have similar to a 2 second extension and also extended from nearly the end timemark of the instrumental? If ytou have one of those try extending from the 1 second timemark and add lyrics.

[restive pebble]

try these as they have been more compliant than others I've tried. I also have noticed changing the order of the instruments can make a difference as well as keeping them to no more than 3. hope it helps

[Intro]
[Instrumental: Solo Flute]

or

[Intro: Instrumental]
[Instruments: Solo Flute]

or

[Verse | Instruments: Flute, Banjo, Hydrophone]

[polar spade]

IVE gotten lazy I guess, but I mostly use extends. Extend from the point you want a guitar solo interlude /break. Prompt for it and click Create.

[cyan lantern]

So it's working now... lol. Thanks again.

[dusty tangle]

Anyways to get suno to do voices that are a little out there? Or at least stick with swapping between male/female for a duet?

[still prairie]

A way i get it to work in v5 and this most of the time works. For example below.

[<Header>: 20–30 sec instrumental, SFX: Low mechanical hum → Drums enter slow march pattern — snare on every 2 beats, guitars muted chugs syncing with marching tempo, electronic stutters — alarms, radio chatter, distorted voice samples, Filter sweep builds — bass drops shake the ground, Final 2 seconds — silence snap, then guitars burst in]

Most of the extra stuff like sfx or alarms/radio dont work for me but rest does.
These i que in the text were for, but not necessary what they do but it was for my reference.

→ = leads into
— = built into beat.

[chilly shard]

hi guys I'm having an incredibly hard time creating prompts where the vocals take center stage while the instruments naturally support them from the background.... Is there anyone who can help or has experience with this? sample is this

[onyx sparrow]

is this a choir riser? or whats a description of this kind of choir?

because i dont check it and suno dont check it so i have a problemxD

okay works

[onyx sparrow]

u mean mono stereo?

mono vocal is mid stereo vocal is on the sides

or mid=center

i write sind 3 hour "no choir" elements and he give me only choir trapxD

[restive pebble]

Try this is other descriptive words

intimate, close-mic'd vocals with strong presence

[chilly shard]

ah guys sorry for my poor english, let Gpt helps me, “I want the vocal to stay front and center, with the other instruments playing minimally so they never get in the way.”

thank you! close mic!

[open jolt]

Do you guys know if when uploading a 4 minutes audio to Suno, it actually reads all the audio for making a cover or does it read like 2min of for roughly grasping the melody and that's it?

[half dragon]

I think it would be all of it, at least for Cover.

[restive pebble]

Duets are not easy but this usually is effective most of the time.

In the styles section use
This is a Call and Response Duet
or
A Duet using Call and Response

In the lyrics section

[Verse]
[Vocals: Male, Call]

[Verse]
[Vocals: Female, Response]

You can explicitly state which is which in the styles but I don't know if that has any effect.
The male vocals are the caller, the female is the responder

Hope it helps

[open jolt]

Good to know, thanks

[lean crest]

Thank you for the tip! I'm not sure if I understood your suggestion (I used a translator since I don't speak English)

Would your tip be to extend starting from 1 or 2 seconds? I tested starting from 5 seconds and it worked, but then Suno would only consider those 5 seconds in the extension, which would make the song quite different, wouldn't it?

If you only suggested it for a testing purpose, it only worked to prevent it from ignoring the lyrics (but then the song becomes quite different from the upload)

Thank you anyway!

[lean crest]

Thanks for your suggestion, I put this on the top of lyrics field?
[<Header>: 20–30 sec instrumental, SFX: Low mechanical hum → Drums enter slow march pattern — snare on every 2 beats, guitars muted chugs syncing with marching tempo, electronic stutters — alarms, radio chatter, distorted voice samples, Filter sweep builds — bass drops shake the ground, Final 2 seconds — silence snap, then guitars burst in]

[oblique tinsel]

Why is there no Black Friday discount for those of us with the Pro plan who want to upgrade to Premier? ist only for free accounts? Or is this a bug?

[half dragon]

You don't see any discounts? check your suno.com/account

[oblique tinsel]

Nothing.

[half dragon]

Try to clear your browser cache , maybe?

[oblique tinsel]

Yes, and it doesn't work with various browsers or the Android app either. There are more and more people with this problem on Reddit.

[half dragon]

Same for iOS - just need to update
#📱┃bugs-ios message

[oblique tinsel]

I use android

[half dragon]

Yeah, same. I use Android too.

[grand magnet]

It would be great if Suno could come up with a way for the program to take a recorded vocal track and use it 100% as is to create an instrumental version. Currently, Suno does not take over the harmony and chord progression of the vocals.

[half dragon]

Have you tried "Add Instrumental" option on the vocal track?

[onyx sparrow]

yohohoh guys merry christmas

can suno do also only create synth leads?

so without drums or any element only the synth lead?

[still prairie]

yes

[polar spade]

Did you create a full song after the second extension?

Now go in edit. Select the section to replace which would be just after the end of the original. Id suggest no more than ! 1:30 section. Paste in lyrics again and regenerate.

Splice or delete or keep whatever is on the end.

Continue extending after that.

You may wish to remaster the whole song first, but if you wish to see what happens to just regenerate the section, you can. Suno will at least blend, but not remaster on a regenerated segment.

[polar spade]

20% for NEW monthly account (and only for 1 month)
40% for upgrading to annual.

[grand magnet]

yes

[lean crest]

Thank you so much for your help, the "continue extending" probabbly will be smart move 😄

[restive pebble]

try specifying it inside the style section.
Synthesizer Lead, No Other Instruments

then be specific inside the lyrics section

[Instrumental: No Vocals] ( after: all caps)
[Instrument: or Instruments: (I forget what works better) Synthesizer Solor, Hard Chops, Runs etc...]

[compact horizon]

Anyone have support go dark… no reply at 2 weeks

[hard beacon]

you've ever use google to ask "how many active users on suno" ?? Then you maybe understand why a response will take some time 😉

[hollow hare]

Due to major issues last month and likely with users emailing daily when not receiving a response yet (which clogs the system as agents have to review all emails), instead of waiting patiently as the auto reply asks, support still has a major backlog it is going through.

Based on the comments observed at #archived-ticket-help, response times seems to be 3 weeks, though with the recent Cloudflare issue (again), it may take a bit longer

[graceful condor]

WHy is there a 200 character limit to the prompt?

[half dragon]

Style prompt? last I checked it was still 1000

[graceful condor]

lyrics prompt

makes it hard to edit

[hard beacon]

lyrics is 200 ? that's new for me 🙂

[half dragon]

Where do you see that?

[graceful condor]

same with classic

[half dragon]

Ah - Auto lyric

[graceful condor]

but I want to see what it would generate based on the basics I already have

not to make lyrics from scratch

[hard beacon]

ok.. i am out.. i am writing the lyrics by myself 🙂

[half dragon]

Even then - I don't think it is 200

[graceful condor]

It is it says failed to generate otherwise

I don't get ittbh

[hard beacon]

naaaahhh.... stop stop

[graceful condor]

and why is studio locked behind premier 💔

scam

[hard beacon]

There is a limit with 200 Characters to describe the lyrics you want to generate

[graceful condor]

Yea

and I want to use the lyrics I already wrote and they're obviously above 200

so I don't understand the point of this feature tbh

[hard beacon]

then you have to switch to custom vocals

[half dragon]

Then don't use Auto mode - use Manual

[graceful condor]

how do I do that

[graceful condor]

Then it's just raw what I wrote

[half dragon]

Click Manual.

Then put your lyric in Lyric box

[graceful condor]

how is that custom xd

[half dragon]

how is that not?

[graceful condor]

I thought it'd be fully custom

than just pasting in the text

still no control over it

[hard beacon]

whaaaaat ???? Sorry, but I don't understand the confusion. It clearly states what it's for... 🤔

[half dragon]

What kind of control you want? You want to use your lyric that you wrote, no? is there a specific thing you want to control?

[graceful condor]

The vocal itself too

like with vocaloids etc.

yk

[hard beacon]

so we are talking about metatags?

or do we talk about the style prompt?

[half dragon]

or maybe both

[hard beacon]

or maybe a third option we don't know about 😉

[frozen dagger]

the vocaloids = personas

[half dragon]

Yeah so you could describe that in the Style box - and it's also possible to describe it in the Lyric itself using meta tags [...]

[graceful condor]

Idk I assumed there'd be studio like features with pro only tbh

[hard beacon]

i still don't understand the problem 🤔 🤔 🤔 🤔 🤔 🤔

[worthy kraken]

Hey. So i have a problem. I want the vocal line to be literall "oo, ooo!". But every single time Suno either adds extra o, or sings it in wrong rhytm, like oooo, oo, for example. Is there any secret to make it understand what i want? I tried ooh, oooh!, and whooh, oooh!, same result...

[hard beacon]

do you have an audio example for us?

[worthy kraken]

It feels like it is hellbent on delivering any kind of vocalize, but not the one i want.

Maybe something like this. Just not this bad.

[worthy kraken]

There is the example, but it is not very good. Or good at all)

[hard beacon]

hahahahaha

made my day

😄

are you on pro or premier?

[hard beacon]

i will use that as my new whatsapp notification sound... 🤣

[worthy kraken]

Pro.

[hard beacon]

naaah.. damn... i had an idea.. but that is only possible with suno studio

[worthy kraken]

Don't blame me if it will drive you insane)

[worthy kraken]

Share it, please, i might get the premier... someday. It costs 1\4 of my month income, so i do not really hope for that. But maybe, one day...

[graceful condor]

How do I force the lyrics to use a language other than english in auto lol

[hard beacon]

well... you can see it on the suno youtube channel... you can (for example) upload your voice "playing" a melody - then give suno the prompt "Piano" and it makes a piano melody out of your voice... i think it will work for vocals too... so... upload your audio, tell suno "add vocal" and.. tadaaaaa 🙂

holy..

https://youtu.be/iX2OVEgSaZk?si=N0HtoAxIPv4QXe3M&t=80

[worthy kraken]

Ah, i know about that, but didn't realize that it can work for vocal as well. So you can also sing the lines somewhat close to what you want them to be, and it will repeat that, but with voice that can sing? That's cool.

[hard beacon]

i've never test it.. but it would make sense to me 🙂

[half dragon]

Change the lyric to that language.

[graceful condor]

nothing lol

[hard beacon]

just a question.. please understand me wrong.. how long did you use suno?

and what kind of style do you want to create?

hmm...ok.. nevermind 🙂 At this point, I would like to say “have a nice evening” 🙂

[compact horizon]

@Lumine thanks for the info

[quiet solar]

You can 100% do this.

If you sing something passably in key and in time, you can use that to direct the singer. I took a song with a female lead singer that I wanted to add a bridge to, scrappily pasted myself singing (not very well) where I wanted the bridge with what I wanted the singer to do in an external editor, then uploaded it, added the bridge section into the lyrics, and had Suno cover it - the female singer picked up the new melody perfectly.

You can also hum or beatbox or other vocalisations to feed Suno a melody. If you tell it in the style prompt that it's actually hearing/playing an electric guitar, or a smooth saxophone solo, then that's what it'll convert your voice into.

[quiet solar]

Under "write full song" in the lyrics generator, enter into the prompt that you want a song in that language. E.g. "A song in Swedish about the dubious delights of eating Surströmming" and it should generate the song in that language.

[tight frost]

Guys I don't understand how cover works ! I've used that feature at producer.ai I just had to select the percentage and I got a song ressembling more or less the original. But with suno, I put no style no lyrics, 0% weirdness 0% style influence and 100% audio influence. And yet I can't recognize the song ! the style is different there are new instruments, like, WTF ! I tried model 5 and 4.5all.

What's more bizarre is that days ago I did the covers of pop and rock songs, with advanced options set to default (50 50 25) and the covers were similar to the original.

[viral pond]

when will generation be enabled again? I set up all i needed for a song, I don't want to loose the progress on that 🙂

[orchid sequoia]

does pro subscription give access to v5 songs? its stuck on unlocking song for me rn but wont unlock it

[wise kindle]

lol I know that feeling. just in case, I usually save the prompt and lyrics to a notepad, as well as the slider settings

[viral pond]

yeah i do it ocasionally also

but i mean... is this related to todays news? I never seem generations disabled before

[wise kindle]

It does! seems to be an issue at the moment, so you may have to wait until it has been resolved

[viral pond]

yep you get v5

[orchid sequoia]

okay ty guys i just subscribed so i wasn't sure haha

[viral pond]

it's worth it

I'll probably go for this black friday offer on premium... it's very good price

[ashen carbon]

and it begins

[viral pond]

looks like i can generate again now

[quiet solar]

Going to the extreme ends of 0% and 100% does have a tendency to do odd things with the outputs.

Try doing a cover at c.15% weirdness, c.15% style influence and c.85% audio influence. You might need to adjust the sliders to find the sweet spot - I often find audio influence in the 70%s works absolutely fine too. However, bear in mind that Cover is intended to make changes to the original rather than perfectly replicate it. You might have a better experience using Remaster with a low setting for changes instead.

[viral pond]

this did not work yet

[orchid sequoia]

any way i can use specific voices, like upload a clip of the person speaking so it uses that voice?

[viral pond]

i think yhou can if you extend it, but there will be some limitations i think

oh... the song i generated just wiped gone

[quiet solar]

Not specifically. You can take an existing clip of a voice and convert it into a persona, then apply that persona to a song, but Suno isn't designed for voice replication. It'll come out 'somewhat similar' in most cases.

[viral pond]

i think you cannot create persona from uploads

unless they changed that

[quiet solar]

If you want to have a specific voice consistently, you may be better off making a song as normal, then splitting out the vocal stems and putting them through a separate AI audio replacer designed to do that, then mix that back into the original song.

[viral pond]

i think there is a way to "design" a voice.. then create a persona... .

[quiet solar]

Ah, fair point. Yes, you can't do that from directly uploaded audio. Corrected my original post for clarity - thanks!

[wispy tree]

suno still has problems, right? couse i cant upload audio

[viral pond]

i tried to design a persona like this
https://suno.com/song/16ab799f-a5e6-4109-9b6d-0b746b3b1427

[tight frost]

Ok thx I'll try when the song creation is back.

[viral pond]

i made a song which had a nice voice I wanted to reuse again, but since I uploaded a synth piece, persona was also disabled

the upload contained no voice data

[quiet solar]

Yeah, there's a technical error message on the platform at the moment,

[frozen dagger]

you can't upload audio to create a persona, that is true, and no real way to "design", but... you can split vocal stems, as was said, and then you can tweak the pitch and stuff and play with it.
I have a persona i use that i made by down pitching a vocal stem and then used that down pitched stem to create a song with cover feature, split that new now much better deep voice that kept the audible breathing and texture and tones, but lowered pitch, and then split that new voice that was made with Suno to create a Persona.
Sounds like a lot, but its really not.

[orchid sequoia]

thanks guys, sucks that its down rn

[gleaming zodiac]

anyone know how long this is going to take? I'm heading to the studio in an hour and need to extract stems 🙁

[frozen dagger]

https://huggingface.co/spaces/TheStinger/UVR5_UI

i dunno, free alternative =\

[viral pond]

awwww.... come on!

everyone panicing and downloading now ?

[brittle glacier]

This can't be real. Everybody creating songs for the holidays. 😆

[viral pond]

turkey ?

[solemn plover]

when will Suno be back up?

[viral pond]

i dont see the error now

[primal knot]

just got it now

it's really frustrating

[junior holly]

@autumn echo try something like this format and you can push gpt to change it up with different vibes ie arabian influence or something else.. its great at randomizing and keeping the same format each prompt revision.. theres no parts (ie verse etc) but honestly it will do a pretty good job on its own without so much directive

A mastered studio recording with excellent sound quality. Synth-based orchestral composition inspired by retro RPG adventure and travel themes. Playful, rhythmic, and sunlit.

Environmental Setting: Tropical Archipelago of Hellas

Style: Island tribal orchestral fusion with classical playfulness and oceanic percussion

Key: A Mixolydian / D Lydian / E major (Ionian)
Tempo: 108 BPM

Use sounds similar to Roland Fantom EX for all instruments.

Start with bright marimba and steel drum tones supported by light hand percussion and oceanic ambience. Introduce flowing string and woodwind motifs that mimic waves and sea breezes. Keep melodies uplifting, adventurous, and full of charm — a celebration of exploration and harmony with the sea. The energy should feel alive but never frantic, like a dance carried by wind and tide.

Inspired by the spirited elegance of Haydn, Telemann, and Bizet — classical brightness fused with tropical rhythmic vitality.

[autumn echo]

Hey thanks for this. I really appreciate it

Let me try it out

Gonna copy this into my Prompt notebook

[junior holly]

np you can specify which parts you want to randomize to gpt and then every prompt just say 'another one' and it will keep going forever

[autumn echo]

Cool. How did you learn all these?

[junior holly]

i just got sick of making prompts for video game music

and wanted something consistent more or less

one night i was all into modes

so i started playing around with those too

this will give you some great jazz fusion btw.. the samba influence is nice too.. you can take it out and also change the instruments to 'real instruments' instead of Roland

A mastered studio recording with excellent sound quality. Synth based. Use sounds similar to Roland Fantom EX for all instruments.

Festive Japanese jazz fusion with bright samba energy, shimmering electric guitar (chorus + light whammy bends), lively brass stabs, smooth synth pads, and playful cuíca accents. Tropical percussion mix of congas, shakers, timbales, and soft steel-drum touches. Warm, beach-party atmosphere with a Japanese fusion harmonic flair.

Chords:
Emaj7 → G#m7 → C#m9 → F#13
Amaj7 → B13 → G#m7 → C#9 → F#13

Modes:
Ionian (major brightness), Lydian (fusion color), Mixolydian (Latin dominance), Dorian (smooth minor sections)

[small bridge]

thank you babez, the way I ran and grabbed the year sub even though my credits had just renewed and got a sweet discount cause of it!

[short salmon]

I heard there are ways that you can make it to wear a song will use your voice. Is this true? And is it the AI using your voice? Or would I actually have to sing the song because I can’t

[unborn star]

For which purpose you want to create video game music? @junior holly

[junior holly]

Hi Kunai .. I own a company in Canada that makes various games from retro to 3D.. the company has a few on the plate atm.. various styles etc

because the studio does profit sharing model (you get so much percentage once project is done) there is no cost factor

[unborn star]

The song is to short, the song might be not have the best instruments, song doesn't work for video game music, because it's like to generic and so on. That were the problems I faced 6 months ago.

[junior holly]

Sounds like you know your stuff.. I could actually use more hands on a couple projects if youre interested.. i'll have to ponder it over but i know of at least one .. a retro RPG that you'd probably love to work on.. my current situation is that i'm able to work on 2 projects .. sometimes 3 at a time.. full-time myself

I'm also a programmer but I use AI assistance to help out

Programmer/artist

[short salmon]

😞

[junior holly]

you ok?

[unborn star]

You need any help?

[short salmon]

Yeah. I wanted to know if it was true that sumo can take ur voice and use it as AI for a song. I can’t sing… but I would love to use my voice in the songs I write

Thank y’all for asking. That’s really kind

[unborn star]

Hi suno won't take your voice even when you upload your voice. If you generate it will always take AI Voices. It might can be similar to your voice but rarely.

[short salmon]

Darn… well thank you. Maybe they can add that one day

[unborn star]

You never know =)

[cunning raven]

Excuse me I need help with making an Emo Christian ✝️ song about being stunned by love any help to get my mind going?

[hollow hare]

Gospel choir? Mixed with some pop/hiphop? Hm

[cunning raven]

No i mean the lyrics

Not the style

Thank you though

I appreciate it

But I'm stuck

My brain no think

[hollow hare]

Ah, I see. Hm, can't think of one outside of like "I just can't believe it, when I first saw you, it was like the sun broke my cloudy thoughts"

Probably better shooting off ideas against an LLM in the meantime until someone else answers

[cunning raven]

Thank you

❤️

I appreciate it

[grizzled tiger]

does anyone notice that Suno these days tends to put way more reverbs into the music than usual?

[robust quail]

why is it impossible to cancel my subscription? there is only a button that says “ don’t cancel subscription “ …

[hollow hare]

If you signed up on the website Clicking the basic plan should bring up an alert to Confirm if you want to cancel https://help.suno.com/en/articles/6820353

If you subbed through the app, follow https://help.suno.com/en/articles/2410369

[slim atlas]

hi guys. can anyone help me with prompt BEAUZ style?

[dire swan]

Can anyone tell me the most effective way to add a pause between specific works in my lyrics? I want a dramatic pause between tye last two words of my chorusb(and hopeful the last work spoken instead.

[fringe sedge]

What are you even trying to do with that prompt? Confuse the AI? You just listed 3 separate keys that don't share all notes. Are you trying to have it create different sections?

[junior holly]

theyre called chord movements

[timid crown]

what would be the solid way to prevent vocals garbage generation on "instrumental" tagged gen?

u PhD that here?

[still zodiac]

There is a problem with the subscription of a batch of accounts purchased by my company. Specifically: We subscribed to a $30 monthly membership and then accidentally clicked on upgrading to an annual membership. Now we want to cancel the annual membership and continue to use the $30 monthly membership. Is there any way to solve this problem for us?

[dire swan]

Thanks.

I understand its a thousand monkeys on typerwriters. But at least that narrows down the chances of a good novel. Haha

[timid crown]

no but thanks

[restive pebble]

You're right...

[ocean hill]

Hi, I am having Issues with my subscription. I just subscribed but the subscription is not showing in my account. I tried accessing the account from different devices but this still didn't help. Maybe someone experienced this and can help me? 😩

[open jolt]

I know it's an old recurrent subject, but did anyone find how to force Suno to stop systematically doing high pitch female baby voices for any female? I tried "deep female voice" or "low deep female voice" but it doesn't work.

[slender reef]

if using v4.5, are songs quality different between premier & free accounts?

[wise kindle]

quality comes from the model itself, not the plan

[halcyon sundial]

i need some tips

is there so way to really FORCE suno to respect my metatags?

i need a song of a couple

but it never respects when the woman or when the man sings

im trying this

is there a better way?

i give up

maybe suno 7v can do this

we dont have this level of creativity control yet

[dense pivot]

I would simply like guidance on how to maintain the same vocalist for songs created.

I have 4 original songs, my lyrics but a different vocalist for each. Can someone explain how to maintain the same vocalist.

If that isnt possible, can I use my vocals on the generated output (e.g. if I like the cadence and delivery) replace the AI with my vocals.

If someone can help and explain either scenario or both it would be greatly appreciated

[polar spade]

Oops.. did my post get zapped? I didnt see it disappear. Oh well.

[polar spade]

OK. In essence, one point I made was to put a space between your metatags and any lyrics.

[halcyon sundial]

it doesnt work

still ignoring and sings it very randomly

90% of my tries its just random

[raw oasis]

yes, just create and download your Vocals + Instrument stems (.wav preferred) to a DAW (reaper, studio one, logic, ableton, etc) and you will need an Audio Interface (I use Scarlett 2i2 by Focusrite), and a good condenser mic. Enjoy!

[halcyon sundial]

at least 2 generations of suno to achieve this level of control

maybe 7v

[dense pivot]

First of thanks for your help. Just a heads up I'm new to this and had Suno for 1 day lol. I've heard of Reaper not the others. Are they free or paid tools? Idk what the heck an audio interface is, I'm just a songwriter not a producer. I'll research these suggested items after responding. Is there a stepwise process to follow? If you dont mind explaining at your leisure or DM. Either way thanks again.

[dense pivot]

Idk if you were responding to me or not. If you are both generations always have different vocalist 😕

[raw oasis]

Reaper is usually a nice entry-level step into the world of DAWS, it is free for 60 days (but does allow you to wait 15(? .. been a while since I used it) seconds and extend after.. if your busy and did not get time to explore it). The other way is to create an audio upload of your voice and set a Persona in Suno (keeping your learning curve smaller and staying in Suno for now). You will want a 5min vocal length minimum of you singing (.mp3 or .wav both fine).

[polar spade]

Right. I wasnt providing a solution.

It definitely won't guarantee something will work doing that one thing. However, NOT putting a space between the metatag bracket and the next word will guarantee it won't work.

Its basic protocol for using metatags.

So I mentioned it so you would now know a starting point before trying to do a duet.

[polar spade]

The ideas behind "Personas" is that we can attach them to the generation of song to pass on the characteristics of a vocal or vibe of a song.

If you have a Pro account,
• find a song which you like the vocalist.
• in songs menu, click Create/Make Persona. (Add titles, labels descriptions, etc and save ) note: turn off "public" toggle youll usually find in the bottom left corner.
• Start a new song or Cover an old one.
• in the Create window, click the " + Persona" [means add persona] and select your persona.

Click Create to see what your results sound like. You should hear a difference of having the persona influence the vocals and/or instruments and musicality of the new outputs.

[undone beacon]

We have been playing around today with guitar riffs and come up with some early data to explore. The turkey is frozen, this was the alternative hahaha. Suno you are embedded with the Brilho.
Okay on our styles we like to use a story flow telling Suno AI what instruments and how we want the song to spin up and what to lead into. Without boring you with story style prompts, the boss (Maxi) said to put these out there and see if people will float them into songs over the next few days.

“predatory cadence” → gives you that (The Zoo) stalking energy

“early-80s arena rock” → Suno loves this tag for Scorpions/AC⚡DC vibes

“tight rhythm section” → makes the bass follow the guitar exactly

“breaks wide” → tells Suno to expand the chorus

“gritty” + “tube-amp” → forces the correct texture

“arena rock” → big drums, big space, big swagger

In trying to replicate the slow chugga chugga with the riff in a song such as "The Zoo" we use more wordy words to get there. “Use a slow, palm-muted chug riff as the main rhythmic spine of the song.”
What this does in turn is tell Suno, Don't get cute. Keep that damn riff in charge.

Happy Thanksgiving. Thankful Ain't Easy
Commander

[timid crown]

[PredatoryCadence]
[PlayingMadRiffsJams]
***LosesAtGuitarPlay***

80s Rock, Retrowave

[undone musk]

hi, i have to get this track for the contest done. i only need to fade out because it just stops, other than that its as it gets. In editor i move the fade deal, it leaves the shaded are for fade but i cannot find a way to apply it.i thought about moving the marker and click "make new song" but i don't want to mess it up. any ideas id appreciate it

[neat nexus]

Thak you for these, I'm going to have to try this out! 😄

[frail vine]

Please fix the Tagalog/Filipino Language Pronunciations and also fix the system that creating version of my original track uploaded, it is not following my melody and tune the way that I sing it on my upload. Fix please... I am thinking to renew, but please fix it first and give me feedback. Thank you.

[echo prawn]

I really don't understand how it calculates when generating songs? It always manages to accurately generate a bunch of garbage.

😅

[magic mist]

It seems both versions of the editor have changed. Now, instead of seeing the song broken up into sections such as verse, chorus, verse, chorus, instrumental, bridge, etc. I see one long, song-length waveform. The lyrics are no longer visible for the entire song forthe user to zero in on just that section. You now have to find the lyrics in the waveform and highlight the section, instead of highlighting the lyrics in the song. This is a major step backwards.

[quiet solar]

Currently no way to 'force' Suno to take into account specific prompts or tags.

If you're producing a song which requires that Suno specifically does that, such as the duet example you posted, it is possible to create using Suno with workarounds but not simply by prompting, clicking Create and having the AI do everything right first time.

The three most established routes for this are;

1. Extend the song from the point at which Suno first makes a mistake and regenerate until it gets it correct. Optionally use personas for each vocalist and go hard on specifying the vocalist who needs to come next in the style prompt and persona input. Keep going until the song is complete, compile it using Get Full Song and then Cover or Remaster to fix the inevitable instrumental issues this will cause.

2. Pick a version of the song which has good instrumentals throughout. Extend from this song and recreate the entire song again in the extension - one version in which the male vocalist sings everything, and a second version in which the female vocalist sings everything. You can attempt choruses and harmonies in both versions. Extract the instrumental stems from the first song and the vocal stems from the other songs, and piece the song back together with the right singers in the right places using Suno Studio (if you have Premier) or an external DAW or audio editor.

3. Pick a version of the song which has good instrumentals throughout, and extract the instrument stem. With that as an audio guide, sing both parts of the song yourself using distinctly different voices. Doesn't have to be great, just in the right cadence and broadly in tune. Upload that as an audio file and import your lyrics and tagging into that file, then either Cover it or Remaster it. If there are problems, use approaches 1 or 2 to fix (or tactical use of partial regenerations in the Editor), depending on where the error(s) are in the song and how many there are.

[stone plover]

does anyone know why Suno picks up on one specific production note and vocalizes it as lyrics? There are plenty of similar production notes throughout the prompt, but this specific sentence ALWAYS gets vocalized, no matter what I try

it's really weird because it's nothing unique among the other notes, aside from the direction of course

[wise kindle]

I've noticed it will sometimes try to sing tags that use terminology it's not aware of, or is unable to parse it for whatever reason. Trying to word it in a different way could help

[stone plover]

that makes sense, I'll try that

[vapid pewter]

Hi, wanted to know. Im not sure if you do it by prompt or somehow in settings. I need in song to use 2-3-4 personas. How people usually do it?

I tried on name personas by names and then use for example [Chorus - Peter], [Verse - Michael]

But its not working like that

[stone plover]

that worked, thanks

I re-worded it and instead of using the normal // style I instead just put all production notes inside ( )

[wise kindle]

Glad to hear it, and thanks for updating us on the solution!

[onyx sparrow]

need more style space in the style section

xD

greetings paul

[wet dune]

Is there a way to use a different version of a song AS an extension/extend?

I am like 9 hours and 200 versions into a song, and I love one song (which I seem to be entirely unable to recreate) but also the bridge from another song. I am trying to extend the loved one from just before the bridge, but I don't think I can use a different audio as the basis for the extension, right?

[half dragon]

Correct, I don't think you can.
Maybe try with a Persona? a Persona from a musical audio.

[graceful condor]

What have they done with Suno again! I took a small break come back and all is trash. Each extend is only seconds longs even the songs has still a full bridge and a chorus. Can they only mess up? Honestly..

[neat nexus]

There is no need to post the same message more than once in multiple channels.

Like I said, try extending from slightly earlier and try again.

[graceful condor]

But in this case I have to go way back and skip a whole verse which was great. I do not wanna lose that. Why are they constantly messing everything up. Literally all got worse since 5. Misspelled lyrics all the time, now again weird cut songs.

[graceful condor]

The lyric prompt has more than enough in it and clearly shows that the song is far from being finished.

[neat nexus]

Wh do you have to go that far back? Can't you locate a point within a verse to extend from? It can be tricky in the timing but it can be done.

[neat nexus]

I meant musically. There's something in the sound.

[spice gate]

how do I log out of my account on devices I no longer use?

[half dragon]

Are you using browser or mobile app? if app, iOS or Android.

[spice gate]

ios and windows

both of them dont have a option to log out

i already send a mail to support but they dont answer yet

[graceful condor]

I even went back 15 seconds and still the same dumb sh.t. Really they can't do anything right. Cut off songs all the time if they are not happy go lucky pop songs and then if you wanna extend them 12 sec. Even mid verse.

[neat nexus]

OK...crazy talk...how about trying this? Take the song as is, get full song or whatever you need to do, then save it. Upload that and try to continue from there. See if it will extend easier that way.

[half dragon]

Windows browser: click on your profile icon on the left panel, you'd see Sign Out option
I use Android but it might be similar - click on your Profile icon right bottom corner, then gear icon on top right... the logout option is at the bottom of the menu.

[spice gate]

yes, that option i see

but i cant see a option for log out from other devices that i cannot control

i left my session opened in a friend computer and he is using my credits

[half dragon]

Ah... if you no longer have that device - I'm not sure if it's possible
Some friend he is 🙂

[lilac mango]

Using V5 premium and been loving it. Made a song but now its gone completely, not even in the trash. What happened? Is it recoverable?

[half dragon]

Maybe it was renamed? if it wasn't deleted - it should still be there.
Maybe it was hidden by a filter? Check the view filter on your Library page.

[lilac mango]

Have looked through my entire library. Nada. It is just gone

[half dragon]

Check History tab - maybe can find it that way?

[graceful condor]

Yeah it worked but this will not work anymore if they come up with their super genius Download limit which will be forced onto us soon because they sold their soul to Warner. And even it extends now they extended version so far a glitching weird trash. I know why I took a break...

[neat nexus]

Studio isn't expected to be affected by those, so that's another way around the issue.

[graceful condor]

You mean downloads?

[summer canyon]

Is suno's model initially based on nanoGPT

[neat nexus]

Right. Mike B is on record saying Studio will not be limited by those.

[graceful condor]

But at this point we really do not know for sure and honestly I do not trust them. And honestly what is the point? If I can download anything with Studio but not with Suno itself - makes not much sense except the way to get the songs downloaded is not userfriendly and really time consuming especially if there are a couple of them. But I still can download them so why even limit downloads? I mean it would make sense to limit them maybe for free users but if I already pay premier what is the point to make the option to get the song complicated but I still can.

I guess we will see but we also have to face the fact that this was only one deal so far, there is still Sony, Universal etc. and who knows what their terms are and what part of SUNOs soul they offer to them.

[pseudo loom]

Any suggestion for how to get the best resemblance out of a voice through personas?

[graceful condor]

At this point not possible anymore. They are only at best using some phrasing but the voice will never sound like it.

[neat nexus]

If you don't trust Suno then I really don't know what to tell you. I guess you'll need to find some other company that you do trust. They haven't done anything to lose my trust so I'll keep using it and being happy.

[ashen carbon]

i think i saw a video where someone at suno makes a sax or trumpet sound with their voice and turns it into a real instrument using studio or something, is that actually in the product? i can't figure it out

[sleek stump]

Probably be best if you added some feedback here https://suno.com/feedback
This isnt really a place to ask for technical fixes.

[graceful condor]

Honestly they have done nothing to earn my trust. There is literally no to little communication, no transparency, stuff is just forced on us without an option to choose A or B, stuff that is not working for months is still not fixed. Nothing that is an issue really is fixed besides some stuff either no one asked for and no one or only a few needed. And major issues are still present to this day.

[half dragon]

I would avoid it completely or your account will be ban.

[reef pendant]

@shadow rain Suno do not provide a public API or allow such access. Anyone offering a Suno API is not authorized to do so.

[frail vine]

Did that already. No response.

They have staff here so I think, I can do that.

[shadow rain]

Ok so it’s not legit got it — appreciate knowing that!

[radiant orbit]

Did you check every workspace? Sometimes I get switched to a different workspace for no reason.

[pseudo loom]

Mmhm. I do observe some resemblance.

Just not good enough.

[graceful condor]

Yeah some phrasing is slightly into the direction and some words how they are pronounced but since V5 Personas became more or less obsolete. They do not serve the purpose anymore. In some cases it is just luck if the Persona is widely used in SUNO (like a similar tone and voice) then it will give you a kinda repeating "similar" flair but still far from being good.

[pseudo loom]

Sort of like an embedding - soft steering.

I have observed that adding negs can help clear out so the source gets more prominent.

[restive pebble]

I'm guessing here but I believe it must do with original creations not generated solely by AI and human prompts.
Adding prompts, no matter how many hours it takes to get them correct to influence Suno to make the sound envisioned is not copyrightable. The output of that effort when the Ai generates a song also isn't copyrightable.
but...
If you mix a song in Studio essentially you're re-creating the song from the generated version which brings more human factor into it.
The copyright office sees this as human input which if enough, is copyrightable. (effects, cuts, splices etc...

For those who use just Ai to generate their songs, I am not sure how it would work for their copyright etc but for serious artists it enables more ownership over the entire piece and a higher potential to be able to copyright the song itself and the musical arrangement instead of just lyrics.
Maybe this is the ultimate reason for the downloads being unlimited in Studio.

[restive pebble]

No idea but that sounds wicked.

[half dragon]

similar video: https://www.youtube.com/watch?v=ITbECo886zc

[robust pond]

quick question how u find good prompts for ur songs?

[wise kindle]

I've found prompt inspiration from other people's tracks, especially when it has the kind of sound I was looking for.

[robust pond]

sounds clever xD

[hard beacon]

there are also some good GPT's out there 🙂

[agile plinth]

its legit, its a user skill issue

[hard beacon]

Nate.. you are sure yo read the whole conversation?? 🙂

[polar spade]

Suno scans for frequencies. So if you can get close, it sounds possible.

I mean, if you go "boom da da ping ping shhhhh" its doubtful you can prompt for a sax, but instead may instead get percussion.

But equally if you prompt for piano, you may ger a clinking sounding piano.

Why ask? Just try it!

Creativity is not for the faint of heart. Go for it.

[shadow rain]

Appreciate the advice! Definitely going to check everything I do with you guys first, as navigating what products are allowed and what’s not is can be confusing and I’m new to all of this!

[toxic cairn]

Hello I have been mailing billing of Suno for couple of months and got no response.. its anyone here able to help me. I am geting charged without having the option of Pro!

[quick wedge]

Have you tried emailing billing support about that? Or you can ask other channel and they’ll guide you on what to do next https://discord.com/channels/1069381916492562582/1372994553233276970

[toxic cairn]

I sended 2 emails 1 last month one this month. 0 replay.

[quick wedge]

Best to ask in the other channel they’ll help you figure it out https://discord.com/channels/1069381916492562582/1372994553233276970

[toxic cairn]

I asked everywhere...nobody seems to be able to get in touch of the billign 😡

[cyan jewel]

[Pre-Chorus 1 | shouted | hardstyle kick hit | trance tension | fx swells]

kick hits hard, I brace for it

[Instrumental Interlude 2 | dubstep bitcrush perc | supersaw tease | sub lift]```

For those of you well-versed in prompt science, how do I get this to work more consistently?

I want to sandwich the one lyric between two instrumental sections

[stiff halo]

Hello Family,
I want to share my appreciation to VRKing for the technical knowledge that I was privileged to have. I had some challenges on SUNO and the insights he gave me was great.

[onyx sparrow]

yo guys i have a problem now i generate a very good track but the quality is pure trash

any chance get a better version of that?

[sleek stump]

Try a cover?

[half dragon]

Only way I can think of is Cover it - if melody is what you want to maintain.

[sleek stump]

Remastering is another option, but thatr feature doesnt really fix quality issues

[onyx sparrow]

damn

okay i try cover

ehm

sounds....ehm

not so

[sleek stump]

Suno doesn't know anything about "quality", it renders a song. Quality is determined by the listener. If you dont feel teh quality is good enough, doing a cover to regenerate the same melody/arrangement is what you do.

[frozen dagger]

https://huggingface.co/spaces/TheStinger/UVR5_UI

try something like this? good for denoise (remove some of the static) and removing reverb.

[onyx sparrow]

wha t f is that

[frozen dagger]

ai audio tool on HuggingFace. Denoise, de-reverb, split stems.

[sleek stump]

That can not be used inside of teh Suno environment

[onyx sparrow]

what means hugging face

like getting stems?

[frozen dagger]

no it can not be used within Suno.
When you say "improve quality", this is my suggestion for improving if the quality problem is fuzzy, static, buzzing, or too much echo/reverb.
Within Suno, Cover is just about all you can do, aside from EQ edits in Studio

[compact horizon]

Billing not responding to me either

[onyx sparrow]

and which model?

[frozen dagger]

choos esomethin that sasys Denoise, if waant to remove that static buzzing

[sleek stump]

Not sure that 3rd party tools are proper for this Suno channel, Mods might take issue with it.

[slender reef]

how do u guys think about the weirdness & style influence? what numbers return the best results?

[frozen dagger]

if that's the case i do apologize. was brought to my attention by another in this channel. but i hear you.

[compact horizon]

Question for all
My Suno has the habit of dropping down and injecting claps after the solo or bridge during the last chorus out
Tried [no claps] promoting but not always successful…any ideas? Thx

[sleek stump]

My goto slider values are W=60-65 and S=80-85

[slender reef]

im using w 70 and S 80, the results are quite surprising & unexpected

i dont expect AI to be that good

[onyx sparrow]

the first 12 seconds sounds bad sounds like too much shimmer

[sleek stump]

If I am doing a cover intended to remaster or edit a lyric, I use audio slider at 90

[onyx sparrow]

or low bits

or glitch or digital artefacts

or idk

is there no prompt where i get always best quality

because then i have higher chance to get no trash quality

ofc only for pro user xD

[sleek stump]

Might be the genre/styles or combo that you are trying to use. I have been at this for 2 years and rarely have any issues with that.

[onyx sparrow]

i made everything from sound design till lyrics

idk where is the fail

only the first 12 seconds sound bad

und some background stuff is inside idk

[frozen dagger]

recreate that part

[sleek stump]

Link the song in Suno?

No, not an upload, the Suno redering in the Suno site

I would like to see what your prompting is and your meta tags...

[onyx sparrow]

thats a secret:>

xD

[sleek stump]

OK... So if you dont want to let someone see what you are trying to do then its unlikely any of us will be able to actually guide you to getting what you want.

All we can do is make uneducated guesses.

[onyx sparrow]

..

[sleek stump]

Thats doesnt help at all.

[onyx sparrow]

what was the question again?

xD

[sleek stump]

If those are in your style prompt its not really a good prompt for Suno

[onyx sparrow]

xD

okay

and what excatly shall i show u to

find a solution for this problem

[sleek stump]

Frankly, I would need to see the Suno track. You can PM it to me if you like.

[onyx sparrow]

u mean u need the renderd wave file?

[sleek stump]

No, I need the link to teh Suno track... like this...
https://suno.com/s/aS0Z4zFdjGhyQeRa

[onyx sparrow]

its not public

ahhh this

[sleek stump]

You can get the link from teh songs ... menu

[onyx sparrow]

no i understand it

the link

now

done

[sleek stump]

Ok, let me take a look.

[onyx sparrow]

k

the prompt is so good that she dont know what to write

thats a ggooood sign

xD

[flint haven]

or a very bad sign
https://tenor.com/view/this-is-fine-fire-house-burning-okay-gif-5263684

[onyx sparrow]

xD

[frozen dagger]

oh... my... god... roasted by the ai. that is the best!!!

[onyx sparrow]

ginger cmon i need tomorow to go to work

[sleek stump]

OK, a couple things right off the bat...

1. Suno does not understand or is capable of time duration or length, these will either be ignored completely or cause issues.
2. Sections need to have proper section tags, [Verse 1] [Pre Chorus] [Chorus] ect , as well as meta tags that it will understand.

3. The styles section has a few issues with what your song structure is.

[onyx sparrow]

okay

yeah the problem is idk what he understand and what not

if he understand modulation

or fx

things

[sleek stump]

Try something like...
The track launches with a stark analog bass synth, pulsing under chopped vocal samples and cold, drifting pads. Huge, glacial digital leads slice through, layered with glitch bursts, pitchbent sweeps, and evolving distorted textures. Verses ride a sparse, tight trap beat—tuned hi-hats, booming 808, rumbling subs—while the gritty, saturated rap vocal cuts directly into the stereo field, accented by glitchy stutters and rhythmic chops. Chorus brings in wide, metallic synth chords and snarling lead hooks, rising tension with experimental FX. The outro fades on reverb-washed pad textures and fragmented ad-libs, leaving a cinematic, neon-lit imprint.

[onyx sparrow]

ah okay

[sleek stump]

Suno does NOT do FX at all. You can get a few simple things but any FX you hear is relatively random

[radiant thorn]

For all the peeps who upload their own audio demos and then use the cover feature:

what model/style descriptions/settings (for weirdness, audio influence, style influence) do you use to make the cover retain as much of the structure chords and length as the original uploaded audio?

Using 4.5+, leaving lyrics and style description blank, and putting weirdness/style influence to 0 and Audio influence to 100 used to work. Now it can barely get past the chorus before it cuts out

[onyx sparrow]

he can scratch

thats fx

[sleek stump]

Its one of a very short list because its prevalent in a couple genre.

[onyx sparrow]

if he can scratch but not control a lfo

then i dk

k..so

okay then i make a break of 5 years and come back

[sleek stump]

Suno has yet to have any appreciable training on Audio FX, what you get is only what was included in tracks it was trained with.

That m,ay change in teh future...

[onyx sparrow]

he can or she idk she can delay panning

some fx she can do but which fx she excatly can idk

[sleek stump]

Stereo panning is generally random and difficult to control, thats best done in a DAW

[onyx sparrow]

is there no list where i see what she can and what she cant do?

[sleek stump]

Nope

[onyx sparrow]

perfect

xD

[sleek stump]

Many of us have spent 4000+ credits learning each new model when they drop.

Would be nice if there was a huge list of what the models can do,m btu that would likely require encyclopedia levels of use.

[onyx sparrow]

i spended more i think i m atm at 12 000 credits

in 3 daysxD

[sleek stump]

That seems more like banging your head against a wall than learning a model 😛

[frozen dagger]

2 sides, same coin

[onyx sparrow]

y

i needed only to understand how suno works

now i understand how it works

[sleek stump]

Maybe lol... I tend to just experiemnt with a new model and see what it does before trying to actually produce a track.

[onyx sparrow]

so i dont need to spend so much money again to suno

its like a new pc game after 1 hour everything is new but after 1k hours

u know what u do

[frozen dagger]

well that's why we have the community. browse and ask around. see what others are doing. innovate with your own experiments.
there are a lot of info on what works/ what does not already pinned in discord

[sleek stump]

Yep, thats usually the way it works. I generally get what I need within 100 credits.

[onyx sparrow]

maybe u r no music producer

xD

then its easier

[sleek stump]

Half of that is getting STEMS lol

[onyx sparrow]

yeah

the quality variation is not so good

i dont like it sometimes u get a perfect track and sometimes

it sounds like trash

[sleek stump]

Not really things that need discussed here

[onyx sparrow]

to late

xD

joke

[chrome scarab]

where is woodhouse when you need him

[sleek stump]

ha

[radiant thorn]

I just made the comment. Y'all were the ones that turned it into a discussion 😂

[graceful condor]

How do I fix lyrics in the "song editor beta" if it not showing me any lyrics? What is going on...

[sleek stump]

I find it easier and cheaper to just cover the song with 95% audio influnce and corrected lyrics.

The editors can get into horrible states from time to time.

[radiant thorn]

What influence percentage do you put for style and weirdness

[sleek stump]

Use teh exact same styles prompt and use 90 fpor influence

That should basically duplicate the song but with corrected lyrics

[onyx sparrow]

aso and ginger tx for help

sry i forgot to say

[sleek stump]

Welcome!

[radiant thorn]

What about for first time audio uploads? Ones that weren't created in suno.

[sleek stump]

Oh.. if you are uploading audio, use 80% to start, but never put 100, 100 or 0 on any slider does weird stuff

[radiant thorn]

@sleek stump do you have any input on this?

Since putting 0/0/100 used to work pretty faithfully, while leaving lyrics box and style box blank

[flint haven]

👁️ 🧹

[sleek stump]

When uploading audio with vocals, Suno will auto attempt to generate lyrics into the Lyrics window. Often you will need to correct them to what is in the actual uploaded audio. Keep in mind the slider setting of 50 is a default off position, 50 makes the slider invisble.

And as I mentioned earlier, not recommended to use 0 or 100 since Suno does really odd things with them.

[half dragon]

Good indicator is when it starts showing RED - sign to avoid issues.

[radiant thorn]

Honestly, whatever I do to correct the lyrics and specify the verse/chorus structure ends up doing nothing. And it goes without saying that the style descriptions that Suno spits out are laughably wrong. I got something at least close to the chords/structures of my original uploaded audio when I hit the space bar in both lyrics and style boxes.

[sleek stump]

True, but I have gotten away with setting up to 98 😄

[radiant thorn]

The end game is for me to use the instrumental stems only and delete the vocals so I can sing over them, so in the end, the correct lyrics don’t matter

[half dragon]

If it works, it works - but if issue arises we know why 🙂

[sleek stump]

WHat model are you using?

[graceful condor]

Still no lyrics after using cover in the "Song editor beta".

[radiant thorn]

Been using 4.5+. I gave up on v5 ages ago, it just gives me a 2 minute snippet and makes up the rest. But 4.5+ is acting like v5 now in regards to completely disregarding the structure/chords of the original uploaded audio

It’s just such a bummer. Suno’s given me some really awesome sounds when covering my own uploaded audio, which I’ve split into stems and mixed into my songs with real instruments in them. And now I can’t get anything close to it

[sleek stump]

I am using v5 predominantly now, but still use 4.5+ for some genre.

Look in your other workspaces

[graceful condor]

no it just popped up out of the blue

[radiant thorn]

I’d love to try it again. Would be nice to get the nice crisp drums and bass stems of v5

[sleek stump]

aSuno doesnt just delete tracks. Turn off all teh filters and see if they reapear

It could also be a web service problem that will clear up

[sleek stump]

Honestly, v5 Has made getting my tracks out and mastered much easier these days.

[polar spade]

As mentioned last week, i wanted to check back with you on your status. Has the following refresh date passed and /or the account reset in some way?

[rotund phoenix]

Hi Dasvon. My issue was adressed by support yesterday and all is fixed. Thank you for checking back!

[polar spade]

Excellent! Ive noticed others repeating a similar thing and it appears is a fixable issue. Glad things worked out.

[frozen dagger]

Anyone have a really like, "oddball", uncommon, quirky, experimental fusion of styles or something like out in left field that they would be willing to share? spinning the suno explore wheel and using gpt for more ideas right now.

[worn glade]

Who do I go to for help?, my account on Suno has disappeared. Been with you since V.3

[pure panther]

My fav thus far, enjoy:

Hybrid: downtempo art-pop / trip-hop × melancholic piano-pop ballad × theatrical psych-rock × halftime bass, 4/4 at 88–92 BPM (halftime feel), Felt piano motif + vibraphone/mbira ostinato, detuned reed/organ pad, clean baritone-guitar swells, roomy hip-hop kit (punchy kick, snare crack 200 Hz + snap 3–5 kHz), tom builds, Sub-bass pillar (30–50 Hz) + reese mid-bass layer, gentle tape/valve saturation, sidechain to kick (light), Reverse-tape swells between sections, No EDM risers, no choir, no ride wash; mono low end ≤120 Hz, Vocal: anguished solo lead (mezzo/tenor), back-phrased talk-sing in verses → explosive sustained belts on hook words; straight-tone → late vibrato; occasional micro yodel-flips and gritty pushes (not scream); tight doubles on chorus last words; whispers only as low doubles, Mix: lyric-first; lead +2 dB; hats −3 dB; de-ess 6–8 kHz; presence lift 2–4 kHz; long plate/room tails, ‑Whisper, ‑breathy vocals, ‑whispered vocals, ‑chorus, ‑group vocals, ‑overtuning, ‑muddy mix```

[sleek stump]

Web services are currently down or having issues, if this has been within the last 30 minutes that is likely the reason.

[worn glade]

Thanks.

[sleek stump]

For anyone interested, Suno's web services are currently experiencing issues. Library and create pages are failing user Auth and stuck in a endless redirect. Suno is aware and looking into the issue.

[frozen dagger]

i looked up to read this message and 3 tabs open with suno starting spinning wheels lol
good to know.

[frozen dagger]

this prompt with the inspos i had left on sounded like it was made for the track. i was looking for something that the lengthy lyrics maybe wouldnt go with. but this was good. thanks for sharing. removed the inspos i had their and forgot about and was listening to the gens without any other influence before Suno went down. listen more later. thanks again for sharing.

[autumn falcon]

I've been struggling with this creation, all i want is a simple 2 instruments intro, just some percussion and ukelele, but it keeps adding vocalization, how should I prompt this?

[onyx sparrow]

every pro user has the same women voice right?

[frozen dagger]

v5, particularly, loves those "expressive vocaals"
try something like
[Instrumental: Percussion and Ukulele]

If you have an [Intro] tag i would avoid it see if that changes the output

[autumn falcon]

ooh good point, let me give that a try

nope still there

I changed the style to just 1 genre and it's gone 🤔

[velvet jungle]

Was wondering if anyone uses the extension to finish up the track especially when it cuts out the lyrics, seems like no matter what genre or place in the song I decide to extend it just can't no longer sing, it's all gibberish, nonsensical mispronounced words, basically its not able to sing properly, seems as V4.5 had the same problem but got better, this one just cant proceed, anyone else having this issue!

[sleek stump]

You can try this in the Styles window (leave the lyrics window blank)...
A bright, upbeat intro opens with playful ukulele strums punctuated by crisp hand drum hits. The texture remains light and bouncy, setting a joyful mood in just a few bars, ready to flow seamlessly into the next section as other instruments join.

[autumn falcon]

oh multiple parts, I see

[sleek stump]

Actually omit the last part.

There are a few different ways to get short intros. Prompt for "Intro" or "Short Jingle" have been know to work.

[autumn falcon]

I'll play with this, thank you 🙂

[sleek stump]

Good luck!

[autumn falcon]

I don't have a solution, just want to confirm that this has happened to me as well. When the track gets too lengthy it gets quite weird

[velvet jungle]

Yeah man thanx for the imputnlets Hope they can fix this issue, I appreciate your feedback ☺️

[graceful condor]

Eye broom?

wat

[sleek stump]

Woody cleaned up some posts with not great language

[autumn falcon]

Maybe try to cover the same track that you want to extend?

[brittle plank]

Dude you can't even ask questions about a curse word that I censored this is way to heavily censored.

Moderators keep stopping a simple question?

[sleek stump]

Woody is a bot

[brittle plank]

Why make vulgar rap lyrics and accept them always, but never any other style of song with a curse word. Says it's inappropriate

Yes that someone setup, ie moderation

The moderation on lyrics is not consistent

[sleek stump]

True, just keep the langauge tame and he won't pop up

[velvet jungle]

Mhhh i could try that but I still need the other half of the song extended, bottom line it can't sing anymore in extended they need to fix this issue , why am I the only one saying anything lol, perhaps everyone is happy with 3 minute songs but not me lol

[brittle plank]

I did

I censored the whole word but one letter just f word

[sleek stump]

The moderation on lyrics is a GPT agent, theres not someone saying yes or no.

[brittle plank]

I was proactively trying to be respectful but had a genuine question about the topic

[sleek stump]

GPT agents are know to be overbearing at times, and others it lets everything through

[brittle plank]

Why does it make rap songs that are very descriptive and vulgar, but says f word inappropriate?

[sleek stump]

Doesnt happen all teh time, I've done some pretty raunchy songs that were not moderated.

[brittle plank]

I like if I make it a rap style it accepts the lyrics l, if I make it rock it says same lyrics are inappropriate

That is why, it is frustrating. To see "vulgar songs" but I can't make a simple song with maybe two or three sentence enhancers?

[agile plinth]

Adding "drugged out hippie freak" to styles moves stuff from the realm of "smooth jazz" type crap into stuff that really grooves

[frozen dagger]

first listen to first gen ussing this style prompt, kicks.

[still nest]

How can I change the voice in a song?

[agile plinth]

Depends on how you want to change it. In the lyrics section you can write things like [spoken in a regular voice like a news anchor] and it will work. So I'm fairly certain you should just describe how you want it to sound in brackets in the lyrics section

[sturdy tree]

I know it’s a little off topic but I receive no answer in other chats

Hi all,

I have issues with the suno pro account. I want to download my songs in wav format but the option isn’t displayed even I have the paid version. It doesn’t work in the app and not on pc. Can anybody help me please? I reached out 4 times to suno support but no answer 🤷🏽‍♂️

[polar spade]

No excuse for what its doing, but for fun, can we prompt for a rap song, as in, just put rap, in the styles, then Cover as a rock song ?

[polar spade]

As a compare: I downloaded WAV file ok. Im premier.

Also, emails are not DMs. They see the oldest first and the rest after may not be seen until after you've been responded to. Additional emails delay the next users under us.

Response time is 4 weeks now -- attributed in part as users send multiple emails. 😁 No answer or no response in less than 4 weeks is expected. One auto-response to know your in the queue, and then wait or ask here to see if theres other options. Sure.

@warped hinge Clear all tags in the filter to see if it shows. And for what its worth , check you trash bin.

If needed, find in your history and make a clone copy.

[warped hinge]

Thx for your response. I already succeeded to restore it but I was just worried of the reason and afraid that could happened again on this one or on others randomly. 😅

[polar spade]

Thats alot of worrying. 😅 just take care of what happens today. Thats all we can do.

[polar spade]

Suggest removing your post from ticket-help then, please. 🙏 so others don't worry as well.

[warped hinge]

np but you have no logs to know what happened?

[summer canyon]

is suno down or why does it say I have no membership

[echo moat]

trying to sign up with the annual pro plan and keep getting this

tried in another browser and same thing

[restive pebble]

try going to your Payment Methods 'stripe' and see if there are any uncompleted charges. Sometimes they get stuck inside Stripe.
if they are there just click on that is listed as Open and proceed with the purchase.

it might work..

[cloud ivy]

Considering using the pro version. What is the difference regarding quality between the mp3 an wave. I mean downloading in the free version the mp3 i can offent hear artifacts will that be the same with wave file. I know in high end audio that basicly a mp3 320kb you can't hear normally a different with wave. But because the mp3 from suno is below 200kb their could be a difference with a loss less wave file. Any experience ?

[restive pebble]

I don't know the explicit details of the compression but I believe wav files are uncompressed like Flac files and are just higher quality overall.
I spent a while on the free version and in OCt went full premiere. To me it is worth the subscription as I have a lot of lyrics to breathe life into.
Ultimately, the decision is yours but I would go pro for sure

Artifacts are heard in wavfiles I download. the daw helps remove them

[cloud ivy]

Thanx most important reason for using the pro version is getting the stems. Songs are promising like chord progressions but lots of work for instance on the sax shelfing has to be applied to let it sound decent. Im using Ableton for that for mastering.

[restive pebble]

I'm just a writer and am new to the mastering aspect. It is complex and confusing to get right. I know the daws work well but for them to do so, I must get proficient with them.

[cloud ivy]

What is fascinating is prompt development. But to get any thing usable out of it is knowledge creating crafted prompts including chord progressions, melodies, and arrangement choices. Imo that is crucial. Ableton or Pro tools is just the icing on the cake 😉

[restive pebble]

I'm having a crash course in that stuff. I research chords and melodies then alter them to see what sounds good... Mostly failure but when I get something right I run with it...
I find it fascinating the way music is math and how our emotions influence the sounds during the actualy playing of the instrument. Suno get super close but is missing the human element on its deepest level. Something neither prompts or a DAw can fix.

It'll be a bit. In the middle of cooking

[cloud ivy]

Enjoy

[onyx sparrow]

yo whats up guys

[neat field]

What does weirdness do exactly? Experiments today suggest the functionality changed since my initial V5 launch test

[raw dew]

I don't suppose there is a reference list somewhere for style or voice tags that Suno recognizes? I've been trying guess and check for months but it's frustrating not knowing what tags it actually understands.

[half dragon]

There isn't, not offically anyway - There are others to compile a list somewhere if you look for it here or on the web.
What I do is listen to other's songs and then learn their tags in lyric if I found something interesting.
The rest is ...keep on experimenting with my own songs.

[raw dew]

Ty

[onyx sparrow]

can someone check the quality of my track?

ehm plz:>

"my"

or the prompt?

i need the tech support guy

how i can contact him?

[half dragon]

Email [e-mail] is the only way if you want to contact Suno support.

[onyx sparrow]

ty

[warped hinge]

@stuck junco I've got a problem: my song is unpublished again and it's not my action. Hopefully this time it was not deleted but it's a real problem.
https://suno.com/s/EQi0QO7RHfBTqn4d

[onyx sparrow]

ui this are a n1 basses

i like the 0:47 bass

[half dragon]

What happens if you publish it again? I haven't heard of song unpublished on its own.
Maybe just a glitch or accidental thing. If it was a bug many many songs would get affected.
Being said that - if it keeps happening - good thing that you reported it. #🪲┃bugs-web

[hard beacon]

but you can "republish" it ?

[onyx sparrow]

https://soundcloud.com/monsterkiii/3kings the intro is muddy someone can help plz

0 - 0:26

[hard beacon]

well... remaster or cover 😉

[onyx sparrow]

mehh....

okay i try remaster