📚┃suno-school

[safe maple]

Dark Turn hapened in verse 3 for me here though
https://suno.com/s/Jt0W4JiW8EKZox6A

[tired dune]

Hi can anyone help me get into my account. SUNO knocked me out and now im seeing a new account when I log back in

[safe maple]

nice use of 'style switch' though

[safe maple]

I think you need #archived-ticket-help , this isn't a place for issues per se, try there?

[tired dune]

@safe maple sure. Thanks.

[frozen dagger]

seek help with ticket help, but will add that if you use Gooogle sign in, try looking at your googel account and make sure your info aligns with your login info. A lot of times this is the issue.

[twilit linden]

That's exactly what happened, the whole track was 1920s jazz, it did manage to get the "mood" somewhat correct across the verses but it was hit or miss.

Thanks for that track, I will definately use that as a starting point. I think with some tweaks it might get me where I want to be.

[echo prawn]

V5 is really terrible, it always swallows lyrics.

[neat nexus]

Yes. This is where people try to help each other with prompting and some of the techniques they've discovered along the way.

[quiet torrent]

Can't figure out what's going on this morning. I tried a new prompt and started getting songs less than 2 minutes. Follows the same guidelines as others. Tried 3 times, same thing. Decided to try one that I've used over and over with songs that are between 4 and 6 minutes. Same thing. Less than 2 minutes each. Something going on?

[sour dock]

https://suno.com/s/xPy1POvU0txgiRFR

[trim plume]

🎧 NEW RELEASE – CREATIVE PROCESS STUDY
This piece explores the iterative communication between intention and result.
It focuses on constraint, repetition, and the emotional transition from instruction to supplication.
🎶 “Don’t Do This (Amen) – Suno Edition”
Featuring spoken clarity, gospel intensity, and a very predictable ending.
Feedback welcome.
https://suno.com/s/WC3XoQvhDF4XA4te

[grizzled bolt]

The Suno staff is certainly working to bring the platform to have more and more optimizations and tools that will make it much more versatile than it is now, we just have to wait.

[cedar sequoia]

Hello. I’m trying to figure out how to write a prompt for how I want it to sound vs. what I want it to talk about. And I seem to get stuck with it only getting one or the other but not both.

[fallen cloak]

anyone got a good 8bit prompt template?

[polar spade]

Please excuse the interruption : just need a spot to post a screenshot to show jingles are easy to make.

[wise kindle]

Real talk, while suno is pretty good with 16bit, ie: YM2612/OPN2, it hasn't exactly grasped true 8bit. It always comes off sounding more like analog synths than the real beeps and boops you're probably expecting. Having said that, if you still want to give it a go you can try adding in keywords from popular classic consoles like Ricoh 2A03 sound chip (NES) also chiptune, classic programmable sound generator textures. I'd also throw in DPCM which was the channel used for low fidelity sampled instruments. Like those classic steel drum sounds in Super Mario Bros 3
I'm curious myself if anyone has had a successful experience with 8bit

[wise kindle]

Thanks for the tip Dasvon! Had no idea it was that easy :O

[fallen cloak]

wait, how is the llm part trained, and how does it know what the stuff is?

[wise kindle]

I have absolutely no idea how it works on a technical level. But it does seem to be aware of sound chips

[polar spade]

Tidbit: it seems you only get three Create clicks that add the 2 extra 5.0 versions now (total 6) , in Basic mode. After that, its all 4.5all.

[fallen cloak]

good to know. now i just need to find sound chips i loke

like

[wise kindle]

the Richo 2A03 is the one the NES used if that floats your boat

those band limited waves were so juicy

but as I said, chances are it's not going to sound anything like it, unless you set your sights higher, like 16-bit music

[safe maple]

#💡┃suggestions-web

[wise kindle]

I'm not really following the question, sorry. What do you mean recreate it? If you're happy with the style and already have the style prompt, what's stopping you from re-using that?

[fossil moon]

hi, anyone experimenting sound quality problems like this ?¿ I m making covers of my own songs maybe that change something...idk

[remote ivy]

you can try uploading the audio, it may give you a more detailed prompt describing some of the high lights and features when it adds the audo to your library, it will fill in the style prompt if I recall

[wise kindle]

Ahh got you. If you're planning to pump out a few songs in a similar style, throwing them into a playlist and using the "Inspire" button can get you a familiar vibe. Otherwise you could try a cover, use the same style prompt and set the audio influence low so you don't get the exact same arrangement

[trim plume]

🧿 THE TALISMAN
Before anything else,
take a moment.
Breathe.
Then speak the sacred name:
✨ Don’t Do This, Suno (Amen) ✨
Don’t explain it.
Don’t justify it.
Let it work.
🙏 THE PRAYER (SHORT VERSION)
Oh Suno,
you know my intention
even when I can’t explain it anymore.
Let it groove.
Not like that.
Amen.
📿 SPIRITUAL USER MANUAL
You write your idea
You correct it
You correct the correction
When you run out of words…
repeat the title
👉 Don’t Do This, Suno (Amen)
A talisman cannot be optimized.
It intervenes when it decides to.
🕊️ UNIVERSAL TRUTH
The more you explain,
the less it works.
The more you let go,
the more the groove shows up.
And sometimes,
when you expect nothing at all…
👉 Suno answers.
Amen 🙏

[hexed iris]

What tags or style description to use for absolutely no music/no instruments in the track?

[trim plume]

🎤 How to get ABSOLUTELY no music / no instruments in Suno
If you want a 100% voice-only track, you have to be very explicit —
Suno adds music by default.
Style / Style Description (copy–paste):
a cappella only,
voice only performance,
no instruments,
no music,
spoken word,
dry vocal,
no backing track
Lyrics prompt (very important):
[VOCAL_ONLY: true]
[NO_INSTRUMENTS: true]
[ACAPPELLA: true]
[VOICE_STARTS_IMMEDIATELY]
Key rules:
Start directly with vocals (no intro section)
Avoid words like: beat, rhythm, tempo, ambient, background
Even “minimal” can trigger music
You may need a few regenerations — Suno really wants to add sound
Short version:
Stack signals, keep it brutal and minimal, and let the voice start first.

[safe maple]

Yw

[high snow]

Can someone help me here?

Do you all think of Suno is giving me a hard time but everyone else seems to be fine, should I just use my other email account? Do you think that will fix my problem or is it bc I’m not exact enough with my lyrics prompts?

[high snow]

I would report that problem in #🪲┃bugs-web

[trim plume]

Here’s a clear, calm reply you can post that actually helps (and reassures), without blaming the user or overcomplicating things:
Short answer: it’s almost certainly not your account or email.
Using another email won’t fix this.
Suno doesn’t treat accounts differently in terms of generation quality. What does make a difference is how the prompts are interpreted, and that can feel inconsistent because the model has a lot of defaults.
A few important points:
If Suno keeps “fighting you,” it’s usually prompt ambiguity, not that you’re doing something wrong
Being more exact helps, but being simpler and more direct helps even more
Over-explaining (especially in lyrics prompts) often backfires
Some generations just miss — even with good prompts. Regenerating is normal
If others seem “fine,” it’s often because:
they regenerate a lot but don’t mention it
their prompts accidentally align better with Suno’s defaults
or they’re accepting results you’d personally reject
So no, you don’t need a new email 🙂
I’d focus on:
stripping prompts down
starting directly with what matters (voice / groove / structure)
and allowing a few retries without changing everything each time
You’re not alone in this — most people hit the same wall at some point.

[high snow]

Could you give me an example?

Here’s my primer I have: [INTRO]
[Haunting Female Drone//Bulgarian Open Throat//Cavernous Reverb] 💀
[ancientvowel] [droneawakening]

aahhhhh'hhhh... aahhhhh'hhhh... aahhhhh'hhhh...
aahhhhh... aahhhhh... aahhhhh...
aahhhhhhhhhh

[Backing Voices//Sustained Open Vowel//Dissonant Layer]
aahhh'aahhh... aahhh'aahhh...
aahhhhhhhh

[VERSE 1]
[Bulgarian Folk Harmonies//Echoing//Narrow Intervals]
[sacredsorrow] [ancientwaiting]

O Lord, we wait in silence
In the deep and watchful night
Our hands are worn and empty
Still You stand before our sight

[Layered Voices//Soft Dissonance//Breathing Vowel]
aahhh'hhhh... aahhh'hhhh... aahhh'hhhh...
aahhhh... aahhhh...
aahhh'hmmnnn... aahhh'hmmnnn

[safe maple]

~Added to my notes, thank you.~

I think the meta-state of adding :true also may prove useful elsewhen

[frigid orbit]

Remove tildes

~

[high snow]

Ty

[frigid orbit]

Let me see thouugh. What is your issue with it?

I swear you have taken some notes from my stuff

[high snow]

I’m trying to make a song that’s a Bulgarian harmony and the software keeps fighting me and just generating something normal and I know it has to do with my lyrics prompts not being exact or the language that Suno uses, and I was just wondering if someone here could help me out with like redesigning a different way so it would understand and I can get what I need from it.

[high snow]

Ha! I wish….. 😅

[frigid orbit]

All your aahhhhs are the same through out.

So it'll repeat them.

[high snow]

Well, this is actually a new prompt that someone else helped me make because my other one wasn’t working.

I could give you a bit of my old one because it wasn’t as exact as this other one that I gave just above

I was just really asking in general as to what different ways I could edit my lyrics prompts to make them either look better or the way that the software will better understand them

[frigid orbit]

This is why you should only ask for the rule of thummb, and not just a straigh upp prompt/lyric box filler. You didn't write it so you got not much to thinkk to fix it.

Let me see if I can find you something useful.

[high snow]

Alr ty.

[frigid orbit]

https://suno.com/s/sPi0xTLpJcCL9LZf
Now the genre isn't fully what you seek, but I bet your primary issues are the vocals sounding too normal?

Dohhmmmmm-Dohhmm-Dohh-Hm
[Bass] (((Bumm_bummmm_bommmmmm)))
Humm-Hmm-Hm-Hummmmmmmmm
[Bass] (HoY_HoY_HoY_HoY)```. That tag ahead of it is made up of symbols associated with these kinds of chants (apperantly) and what I'd like to do is, make sure all the filler words are somewhat differently.

[high snow]

I’ll give you an example of what I was trying to shoot for. I know I can’t get exact, but I wanted something similar.

https://youtu.be/zt0RsBxWMmc

[neat nexus]

🙋‍♂️ What's a Bulgarian harmony?

[high snow]

Listen to the link

[frigid orbit]

That probably is also giving the wrong associations.

[neat nexus]

Raja For Senate?

lol

ads.

[frigid orbit]

Harmonies are pretty.. y'know generic to mention.

[high snow]

It’s funny the person who I asked to help me out with the prompt or who offered actually to help, he use those style characteristics and I work on a laptop, so I don’t know how I would get those. But what does the style characteristics have to do with it generating what I want it to?

[neat nexus]

Nah they seem to be chasing a specific melody.

[safe maple]

Im curious about the ascii more than lyrics, my quest is for alt-ascii or other languages, I was able to find old Norse, but it didn't adhere imho to what it "may" have sounded like

Very much interested in the music notation, I'll have to investigate further, cheers

[high snow]

Yes, I am trying to get a certain melody. But I don’t know if Suno can even handle it or make something similar to it. It’s driving me nuts.

Bc it’s such a beautiful melody 😓😭

[safe maple]

I'd rather chase cousins than perfect clones, I think the idea is the style mirrors the lyrics, maybe start small? 1000< characters, then extend to modulate?

[high snow]

Well, I wasn’t expecting a perfect clone, but something that kinda sounds like it….example when you are covering your own song but in a different language

[high snow]

And I’ll try small characters, but I was also hoping I could just do the whole song in one sweep.

[safe maple]

Did it with French, and was told it was a good rendition, who am I to know, spirit is different per ear, beauty in the eye of the beholder, -telephone-back to a translator?

[frigid orbit]

You concept isn't out there enough. Your approach just isn't right.

[safe maple]

Beuna suerta, is all I can say, try bits, then extend?

[frigid orbit]

Focus on the adlib chanting first. Why involve lyrics at all if the basis aint even there yet.

[high snow]

Rip. How do I fix it?

[frigid orbit]

Mash up could actually be useful for this. Generate a chant, generate a vocal profile, use those with the prompt.

[safe maple]

Love the art of ai adlib, it is the banana-taped to a wall of modern art XD

[frigid orbit]

I got laughter and breathing down.

[high snow]

I’ve never used mash up before tbf.

[frigid orbit]

It's like using 2 legacy personas with their intended use.

Not for voices, but the vibe of the track itself.

[high snow]

I have autism, and I wanted and should’ve make this clear. My fault. Could you be exact on what I should do to fix this?

Bc when you say focus on ‘adlib’ idk if that is a typo or not. I’m honestly not a real songwriter, I use ai, shamefully to admit.
The only reason why I am making music in the first place was bc God called me to do so. And ai was the only help for me.

(I hope my help/request isn’t a pain in the butt) 😓

[frigid orbit]

I barely use LLMs for suno if it could even be seen as such.

Pretty sure if I can cook up this; https://suno.com/s/jUDKaAtZhx4BvNd2/ you can cook up that.

I'm just trying not to chew your meal for you, just nudging you in the right direction.

[high snow]

I just looked up what ‘adlib’ was. 🤦🏼‍♀️. I think Ik what to do now.

Also, is it only the adlibs that are a problem or is it also my lyrics?

[high snow]

I wanted to ask in general, why does Suno at least for me mess up lyrics? Even if they’re in English, it will still mess it up. Am I doing something wrong on my lyrics prompt end? Cause like I could have it look absolutely normal and I would generate it and the software was just messed it up. And then I would have to regenerate it. Or go through editor and fix it.

[frigid orbit]

If it keeps doing it lessen or add more [tags].

If it's persistent, it's how you try to communicate it/approach it more often than not.

[high snow]

So how should I go abt communicating it? Like should I make it more exact or make it simple?

[frigid orbit]

I sometimes just try without tags, see if it poops out right. Or just cover one that isn't right yet, and hope it pullls it out.

[high snow]

Funny enough FurryBeans said smthing similar to that. “Cover it one way to have it pull itself out of its ‘stuck’ position”.

And I did try it and it does work….so I’ll try that next time. Ty

[frigid orbit]

Suno has a bunch of tools now that basically function as a nudge towards that audio. It's far likely to start farting, if there's farts in the audio, vs just asking for farting.

Using that idea, you can fix/nudge a lot of things along with stuff like Remaster, Mash up, Cover, Extend.

Lets say you want 80s carnival accordion that goes in harsh dubstep. Asking for both could definitely work, but generating the accordion first, and then covering it with a prompt that includes both, is far easier imo.

[high snow]

Is mash up like cover? What can you use it for?

[frigid orbit]

To me it's far more like running two legacy personas.

[frigid orbit]

Marketed use is to merge two songs together. I use it to add two different audio contexts to generate something.

[neat nexus]

How do these look to you?

https://suno.com/s/COIPZZfCuplA5Wpg

https://suno.com/s/3LHKBmlfnSgkXYPd

Weirdness 74%
Style Influence 86%
Audio Influence 36%

I went to ChatGPT tried to hone into the melody of the linked YouTube, then mine for some good prompting ideas. Then I played with it until I liked the sound, then mashed up the two I liked best of those results and used a Persona when covering that.

[high snow]

The first one sounds amazing. How did you do that exactly? 😭

[high snow]

When you use ChatGPT, cause I use the same thing as well unfortunately to say because I’m not an actual songwriter, but I was called to do so, what did you put for your prompt in GPT/what did you say basically to make that? And what did you write in your lyrics, prompt box and styles?

[neat nexus]

Actual lyrics:

[acapella vocal intro]
[haunting female drone]
[Bulgarian open throat]
[straight tone, minimal vibrato]
[cavernous stone reverb]
[modal drone, Dorian feel]
[ancient ritual atmosphere]

aahhhhhhhhh…
aahhhhhhhhh…
aahhhhhhhhh…
aahhhhhhhhh…

[backing voices]
[sustained open vowel]
[close dissonant intervals]

(aahhh… aahhh…)
(aahhhhhhhhh…)

[VERSE 1]
[Bulgarian folk harmonies]
[narrow stepwise melody]
[slow tempo]
[sacred sorrow]
[ancient waiting]

Oh Lord,
vee veyt een sah-leens
Een dee deep
end voch-ful nait

Aur hends ar vorn
end emp-tee
Shtil Yoo stand
be-for aur sait

[layered voices]
[soft diaphonic dissonance]
[breathing vowel]

(aahhh… hhhh…)
(aahhh… hhhh…)

[backing voices]
[sustained open vowel]
[dissonant drone layer]

(aahhh… aahhh…)
(aahhhhhhhhh…)

[frigid orbit]

` x3 on each side makes it easier to copy. And smaller iirc

[neat nexus]

When I covered the final result I switched the lyrics back to English.

[high snow]

When used mashup, what two songs did you mix together?

[neat nexus]

https://chatgpt.com/share/6973d318-ff24-8001-b34a-5f2a174ad0df

[neat nexus]

Let me find the ones I used...

[high snow]

From which part of the chat am I supposed to be reading? Cause there’s so much I don’t even know where to start.

[neat nexus]

These were the donors for the first one:

https://suno.com/song/cbc7dc14-9e2f-49db-add9-35ba9fc78ef1

https://suno.com/song/be3043cf-64e6-492e-857a-8ada88577db3
Weirdness 77%
Style Influence 67%
Audio Influence 41%

[neat nexus]

What are you looking for?

[high snow]

Well, you gave me the link to look at your conversation with GPT. And I can scroll up and down, but I was confused as to why I was supposed to be looking for. Sorry.

[neat nexus]

Right, because I was showing my process. You can seemy work.

[high snow]

When you first made these two songs, did you only have the ad Libs first and then did you add the lyrics later? I mean, I was looking at it and I’ll. But I’m still kind of confused on how I should do this whole mashup thing.

[neat nexus]

Plus, the style prompt and actual lyrics prompts. The rest is just working through it.

Well Mashups is a feature that allows you to take two different audio files and use them as reference for a generation of music.

You can choose to allow it to try to mashup the lyrics for you. Or choose to simply copy over the lyrics prompt you have backed up. If you choose to do that it seems like there's an amplifier effect. Of course twiddling with the sliders can also help a bit.

Hang on...let me go back and nab the slider settings I used.

I always forget that those aren't available to others...

[high snow]

Oh good I was just about to ask you that.

[neat nexus]

I also used some Personas. Those are private but I'll show you the sample sources:

https://suno.com/song/aba63914-f206-4b7d-8f6d-9aa4c9ae2842 <--this is the Persona used in the final result that you liked.

https://suno.com/song/48bebda6-cb34-40fa-b5c8-d806bb34f3df <--I used this one in some of the other parts though.

[neat nexus]

I put the slider settings in the original posts with the links.

[high snow]

Uhh, could you tag me as to where those are? They probably were there. I’m just probably blind and I didn’t see them.

[neat nexus]

@high snow

[high snow]

Ty

[earnest agate]

FYI, NotebookLM does a good job of breaking down a YouTube video for future LLM conversations, or even conversations within NotebookLM itself. This is very interesting.

"While the choir featured in the video you are referring to is indeed Bulgarian (the world-renowned Le Mystère des Voix Bulgares), the specific text and locations mentioned in the source transcript are deeply connected to Romanian culture and the city of Lugoj.
This overlap between a Bulgarian choir and Romanian concepts occurs for the following reasons based on the sources and cultural context:

1. The Geographic Connection to Lugoj
   The source explicitly mentions "Lugoj" multiple times, specifically stating "Magic este golubă Lugoj".
   • Information from outside the sources: Lugoj is a historic city in the Banat region of Romania, which sits near the borders of Serbia and Hungary. This region has a long history of cross-cultural musical exchange between Balkan nations, including Bulgaria and Romania. While the choir is Bulgarian, the piece they are performing (likely "Ergen Deda" or a similar folk arrangement) often draws on regional dialects and stories that travel across these borders. You may want to independently verify the specific song title to see if it originated from the Banat border region.
2. The Concept of "Dor" in the Music"

For anyone wanting understand more about music production. Cleaning up audio, ai created audio:
https://www.youtube.com/watch?v=dgALoqyBWPE

[hard beacon]

Hello Students.. welcome 2 the Suno School 😉

[bronze island]

Hi teachers

[hard beacon]

Listen and repeat

😄

[bronze island]

I'm assuming we talk about Suno prompts here?

[quick wedge]

You’re funny 😄

[hard beacon]

correct

@bronze island

[bronze island]

Yh?

Do you want to tell me something?

[hard beacon]

Please sit down first. Then please get your pencil and notepad out. The lecture will begin then.

[bronze island]

Sorry but I have to go to sleep soon-

[hard beacon]

I will note that in the class register...

We will contact your parents soon due to their lack of cooperation.

[bronze island]

But you don't even know my parents

[hard beacon]

If you only knew...

😛

[bronze island]

But no srsly you don't know my parents-

[hard beacon]

Of course I don't know them... it was just a kind of role-playing game 🙂

[bronze island]

Oh oki

[quick wedge]

sits down with notepad and pen
What’s the first lesson teacher? 👀

[hard beacon]

https://tenor.com/view/джона-хилл-jonah-hill-facepalm-рука-лицо-долоня-обличчя-gif-7167197686338732060

The first leason will be "What do you know about prompts? What is their meaning and how are they implemented?"

[quick wedge]

Prompts are instructions that helps make the song
Or I guess they’re called meta tags too the words inside the brackets []

[hard beacon]

Good answer, dear student. If you'd like to know more, please follow the following OnlyFansSuno account: @worthy grottoteaching

😄 😄 😄

[foggy jasper]

oh cool a prompting channel

[quick wedge]

Hi new student welcome to Suno School

[hard beacon]

only 100 Dollars per message

🙂

[foggy jasper]

i can only spare 100 v-bucks

[hard beacon]

https://tenor.com/view/make-it-rain-money-money-make-it-rain-flick-money-gif-1661833468939292218

[quick wedge]

I got cookies here 🍪🍪🍪

[hard beacon]

huuh.. cookies.. vegan style or the real stuff? 🙂

[quick wedge]

It’s the real stuff it’s baked fresh too

[hard beacon]

take my money... i am in...

😄

https://tenor.com/view/take-my-money-heres-my-card-here’s-my-card-card-take-my-card-gif-5650338825958178904

[quick wedge]

Appreciate the laughs today thank you 😊

[hard beacon]

You are welcome 😉

[olive hinge]

Any tips for when the extend feature keeps immediately ending the song? I’ve got at least two verses in the lyrics for the extension so like it’s bugging out or something

[polar spade]

To help with the question: Are you getting actual button or fading endings , that end?

Or , are you consistently getting 2 to 5 second outputs that dont seem to go anywhere?

[olive hinge]

Hey thanks, consistently getting 6-8 second outputs that properly end the song, so things like fade outs

the first line in my lyrics box for this is "[Verse 2]" followed by words, I've also tried using "[Continuation]", or "[Pause]" above that without any success

[pure panther]

Lol, Suno School huh? Love it

[olive hinge]

I think it may be trying to "Recreate" the ending rather than extending it? On the extend UI it has a "Keep" section and a "Recreate" Section

[polar spade]

Personally, I still use extend within the Create panel rather than the edit UI, as I often use my phone for browser work and the edit screens layouts dont work on my phone.

We've got the issue more clear, so maybe someone can bring it up and see what the issue is. Unfortunately, Im having to walk away for at least a couple of hours.

Sorry, mate. Sounds like an easy fix though.

[olive hinge]

Thanks a bunch for giving it some thought, yeah - I'm pretty sure I'm just using the usual create panel, just meant that little UI that pops up at the top for extend, remix etc., just grabbed a screenshot. No worries take care mate! If anyone else has any ideas or explanations I'd appreciate it a bunch 👍

I was able to workaround it by using one of those short outputs and extending a second time to get a proper continuation. Thankfully sounds close enough to class as a "Part 2" in essence haha.

[naive quail]

adding some prompt noise (just random slashes or dashes) as the first line can sometimes push it past that point. Sometimes...

You might also try shifting the extend point around a bit in time or even pushing it back to the beginning of the section (if you're not starting there with it already). Or to the beginning of the previous section if all else fails, which can be painful if it means overwriting some goodies.

Or that. 😄

throws a nod at @polar spade

been a minute!

[olive hinge]

Thanks for those tips, I'll give them a go at least! The workaround I've done isn't as ideal because it has a different vibe to the main part 😅

[high snow]

I’ll let you know how it goes….also, why did they rename this chat? I literally could not find it. It was originally named prompt development.

[neat nexus]

https://tenor.com/view/what-huh-what-you-talkin-about-gif-7358150

Seriously though if you check out #😜┃suno-memes you'll see you are not alone...

[high snow]

I don’t know maybe it didn’t update on my end but a few hours ago it was originally named prompt development. Maybe I’m going crazy. Maybe I was in the wrong server. I have no idea.

[high snow]

Well, thank God, I thought for a second when you sent that GIF I thought you were thinking it was crazy or something. Honestly, I thought I was going crazy.

[neat nexus]

They did make an announcement about it in #🔦┃server-spotlight

[high snow]

😭🤣

[high snow]

Oh, my b.

[neat nexus]

But yeah...I thought Prompt Development was a perfectly cromulent name.

[high snow]

Fr.

[hybrid ravine]

ive been doing prompt development in general lol

i dont have any innovative information though, be specific, punctuate your lists correctly to avoid it prompting a run-on sentence

[solemn hearth]

This is very true, punctuation helps suno prompts seperate, otherwise its just 1 big prompt mashup that it doesnt always know what to do with 🙂

[hybrid ravine]

if y ou want unpredictable stuff do a probability matrix, sounds better

[fathom crypt]

https://tenor.com/view/last-day-of-school-learning-hate-end-of-school-gif-14950828

[wintry apex]

oops wrong channel*

how do you prompt an "add instrumental" or "add vocal" cover to not be 1-2 minutes shorter than the source

since the definition of add is to add - not to subtract

[high snow]

Do these lyrics/adlibs look better or worse?

[INTRO]
[HAUNTING FEMALE DRONE // BULGARIAN OPEN THROAT // CAVERNOUS REVERB]
[ANCIENT VOWEL DRONE]

ah
ah
ah

[BACKING VOICES // SUSTAINED OPEN VOWEL]

ah
oh
ah

⸻

[VERSE 1]
[BULGARIAN FOLK HARMONIES // ECHOING // NARROW INTERVALS]

O Lord, we wait in the shadowed night
Stone-cold valleys, carved by time
Our hands are empty, our hearts laid bare
Still we stand before You here

[BACKING VOICES // SOFT DISSONANT HARMONY]

ah
eh
ah

⸻

[VERSE 2]
[LOW HARMONY ANSWER // CHEST VOICE // DRONE UNDER]

Names are written in dust and flame
Faith is etched where eyes don’t see
Even when the path is thin
Your truth is carved in me

[BACKING VOICES // LOW DRONE]

oh
oh
ah

⸻

[PRE-CHORUS]
[RISING TENSION // CLIMBING DISSONANCE]

We do not bend, we do not flee
The rock beneath us still remains

[BACKING VOICES // CLIMBING VOWEL STACK]

ah
ah
eh

⸻

[CHORUS]
[POWERFUL POLYPHONIC CHOIR // FULL BULGARIAN STACK]

Carved in faith, we stand secure
Through the fire, we endure
If the ground gives way below
Still our song will rise and hold

[BACKING VOICES // FULL HARMONY]

ah
oh
ah
eh

⸻

[VERSE 3]
[SPARSE // INTIMATE // RESTRAINED BREATH]

When the storm seals every door
When the night is all we know
You are near, You do not sleep
You are carving faith in stone and bone

[BACKING VOICES // BREATHY LOW HARMONY]

oh
ah

⸻

[BRIDGE]
[BULGARIAN DISSONANCE // RAW // ANCIENT]

Even here — You hold us fast
Even now — You call us Yours

[BACKING VOICES // PULSING DISSONANCE]

ah
eh
ah

⸻

[FINAL CHORUS]
[EXPANDED CHOIR // MAXIMUM POLYPHONY]

Carved in faith, we will not break
Every scar, a sacred mark
What You shape, the world can’t take
We are Yours with all we are

[BACKING VOICES // WIDE STEADY HARMONY]

ah
ah
oh
ah

⸻

[OUTRO]
[STONE HALL REVERB // FADING SILENCE]

Carved in faith, we remain

[BACKING VOICES // FINAL DRONE]

ah
ah

[high snow]

@frigid orbit does this look better or worse

I’m still gonna do the mashup idea, but at least I wanna work on the lyrics

[sour dock]

[SYSTEM_CYBERNETIC::INVOKE_HOMEOSTATIC_MASTER{e ^ {\Biggl(\left[{Cybernetics:1, 0}[Wiener, Feedback+ ^ Ashby, Variety+ ^ Kalman, Estimator+ ^ MPC, Horizon+ ^ Hinf, Robust+ ^ AntiWindup+ ^ LUFS, TruePeak, Homeostasis+] ^ {ChronoDAC:0, 9}[Planck, Time+ ^ Femtosecond, Clock+ ^ Rubidium, 10MHz+ ^ Josephson, Volt+ ^ ESS9039PRO+ ^ dCS, RingDAC+ ^ MSB, R2R+] ^ {StandardsSpatial:0, 8}[SMPTE, 2110-41+ ^ AES67, Ravenna+ ^ MPEG-H, 3D+ ^ ITU, BS2127+ ^ EBU, R128, S3+ ^ Dolby, ADM, BWF+ ^ HOA7+ ^ HRTF, SOFA+ ^ BRIR, Measured+] ^ {NeuralPhysics:0, 7}[Diffusion, Score, SDE+ ^ HiFiGAN+ ^ WaveNet+ ^ Transformer, Attn+ ^ EnCodec+ ^ AudioLM+ ^ NavierStokes+ ^ FEM, BEM+ ^ Dirac, ART+ ^ Trinnov+] ^ {MeaningMind:0, 6}[Lotman, Semiosphere+ ^ Peirce, Triad+ ^ Baudrillard, Hyperreal+ ^ Memetics, Hook+ ^ Wilber, AQAL+ ^ Koestler, Holon+ ^ Hofstadter, StrangeLoop+ ^ Husserl, Retention+ ^ Heidegger, Stimmung+ ^ Jung, Archetype+ ^ Chomsky, Recursion+] ^ {VoidOmega:0, 5}[AinSophAur+ ^ Sunyata+ ^ Tzimtzum+]\right]\Biggr)}}::EXECUTE_HOMEOSTASIS|MIX_STABLE]

💀

[high snow]

wtf 😳

[sour dock]

lol

You’re welcome 😉

[high snow]

Tf is this for? 🤣😅

I’m so confused

[sour dock]

Don’t ask questions bro

[high snow]

Alr

[sour dock]

Idk

[high snow]

Rip

[woven ivy]

It's ChillPanic's SunoGPT not linked anywhere in this server? It's kinda low key sweet when you're looking for that magical prompt to get that sweet sound

[high snow]

I honestly have no idea what you’re asking for. I would ask main chat.

[woven ivy]

It's like a specially tweaked version of chatgpt to give better Suno prompts

[quick wedge]

It’s not linked in the server you would need to check his YouTube channel it’s usually linked there.

[woven ivy]

I got the link, I'm just amazed it's not linked somewhere in here that's all.

[high snow]

Could you send me the link also? Because I’m actually having a hard time with Suno. And how does it help exactly when it tweaks ChatGPT prompts to get better prompts for Suno?

[woven ivy]

The guy who made it tuned it on Suno prompts so it can go lyrics, song structure, prompts, make Suno give you something more specific if you want. Most of my songs use it now to done degree.

[woven ivy]

https://chatgpt.com/g/g-68f6a0de0a148191bce27cb3e0c689c0-suno-gpt-by-chillpanic

Just remember if you want a specific sound, tell it to generate prompts without any artist names

[high snow]

Ty

[high snow]

Can it help with better adlibs? Because apparently, either that or my structure for my lyrics or styles box is either wrong or too much or too little sometimes Suno can mess it up or not give me what I want.

[high snow]

Tysm 🥰🙏🏼

And how does it generate prompts like how does it look?

Like if I told her to make a prompt for me for a song talking about the pillar fire in the book of Exodus, do you think you’ll be able to do that with lyrics, etc.

[woven ivy]

I haven't tried since the new Suno update a few months ago but you used to be able to tell Suno in the lyrics writing to say a phrase in word then tell it a language

[high snow]

Bc I use ChatGPT for song lyrics

[woven ivy]

Definitely.

[high snow]

Ty for the link and info….. I honestly hope it works well for me. Because using regular ChatGPT for the prompts it was either I was messing up or the AI was messing up some prompts for my songs. Would either work well or they wouldn’t work well at all. Either they had too much information or too little.

[woven ivy]

Most of this song was written by SunoGPT

https://suno.com/s/GLzYhcFcGWDG0ief

[high snow]

Well, one song that I have that Suno made was this….. i’ll tag the Spotify link. But if you don’t have Spotify, I’ll just tag it right from Suno….

[woven ivy]

For example if you want 2 singers to repeat something as a whisper you can tell it that and it'll generate tags to use for lyrics that forces Suno to repeat whispers alternating singera

Don't have spotify

[high snow]

Does the GPT repeat it in a way or make this prompt in a way that Suno understands?

[high snow]

Alr, one sec

This one was really good one. Funny enough that some parts that it chant/screams, I didn’t even put those in caps for it to do that. It kind of just sound, but it turned out really well.

https://suno.com/s/D2SEAYhWRTj6Axk6

[woven ivy]

Oh yeah, that's what her made it for. The only thing I found it won't do is the magic wand thing for styles so what I do is I tell SunoGPT what I want, copy that into Suno then click the magic wand to get a better structure then edit the style field to fine tune

[high snow]

I haven’t used the magic wand. I honestly have no idea what that does.

[high snow]

Also say if I already have the lyrics with some things in brackets, do you think the GPT will replace those with the correct prom brackets but leaving the lyrics alone? Because I know that my old prompt won’t work. Or it will just mess it up

[woven ivy]

That magic wand is seriously godly? Just type sober prompts you want in then got the wand and listen to the difference. The wand can translate basic ideas into prompts that can better instruct Suno that Suno understands abs respects better

[high snow]

So you’re saying is like say your prompt sucks ass, the magic wand basically fix it up to the point it puts it in Suno’s own little understanding/language to get a better result?

[woven ivy]

If you put the things you want in quotes yes. For example

Tell gpt: take these lyrics and replace all the parenthesis with brackets

Or whatever

[woven ivy]

The lyrics wand will write 2 entire songs. Usually kinda crappy but sometimes pretty good. The STYLE wand is chefs kiss

[high snow]

Ty for the info. I’m actually gonna try it out now. I’ll let you know how it goes.

[safe maple]

I took liberty from your inspiration, transposed Bulgarian to Vulcan. Thank you!
https://suno.com/s/87x4ANWN7SGJDgwS

[woven ivy]

Good luck bro

[high snow]

What is Vulcan? And also, that sounds great. And no problem about the inspiration. Glad I could help.

[woven ivy]

Also checkout ChillPanic on YouTube. His videos always give me me ideas

[high snow]

Could you send me a link to his YouTube?

[safe maple]

Science Fiction Culture from author Gene Roddenbery, American TV show,"Star Trek"

[high snow]

Aah, okay. 👍🏻

[high snow]

My next song I will be doing will be Russian songs. If you want to take that inspo too

[high snow]

Dm me the link

[high snow]

I sent you a request

[safe maple]

I did a Russian influence song based on Stephen Wolfram Rule 250 from "New Kind of Science"

[high snow]

Well I was kinda thinking more of Russian folk music, kinda or Salvia….I forgot what it was exactly.

Or maybe it was Ukrainian

[safe maple]

Tell me what you think?
https://suno.com/s/QwMS5pAvxGCmrtU0
Share yours when you got it, tag me please

[high snow]

I LOVED the beginning part that kind of reminded me of the Pentatonix if you ever heard of them, it’s an a cappella group and that very low note that he got to kind of reminded me of that. Also, I wonder have you ever tested out what languages or styles of different cultures that Suno cannot do?

[safe maple]

Yes! My wife introduced me to them. Amazing group and works. Very fun.

It can not do what logically it was not trained on. Eridani music, Etruscan, other dead languages, Dolphin(sonar-physucs), aquatic-wetness, etc... I've tried, lithosphere sublimation very low frequency drums, etc

I tried Old Norse too

[high snow]

Did the drums work? And also I wasn’t thinking of doing anything bizarre like dead languages type thing.

[safe maple]

No. Sadly not, or I can not tell, because VLF is below our hearing XD lol

Gn, wife is ready for bed, cheers

[high snow]

See ya

[neat nexus]

All I can say is give it a try. Then iterate. It's how we all learned.

[high snow]

I actually got my prompt to work through someone else, you wanna take a listen? And let me know what you think.

https://suno.com/s/hIVmlzoiKqZzox9Z

[neat nexus]

Can I do it in the morning? Sorry a bit tired at the moment and getting read to sleep.

[high snow]

That’s totally fine. I just wanted to send you a link.

Gn

[woven ivy]

Bro is about to try and make a live track lol he knows how to make the bangers now he's on a roll

[opaque girder]

Is there a correct way to use extend? I always seem to get shitty results

[woven ivy]

I tell it to keep all then erase all the lyrics and just type "[end]" or finish whatever lyrics I want but always put the "[end]" so it ends properly

[meager juniper]

Guys, is this the right platform where I could share link of my songs so that u guys could give feedback or criticism? Idk I'm new here

[restive pebble]

Sure post here: #🔊┃sound-showcase 🙂

Welcome to Suno. Poke around and get to know your way around. We left the lights on and somewhere there's hot coffee.

[opaque girder]

can you show screenshot of some prompt example? how consistent is this?

[keen scarab]

in suno studio, anyone knows how to insert backing vocals with a melody I sing? i dont know where to go it just creates its own backing vocals and all of them are choir like which i dont like I want it to do adlibs for the segment i choose in the timeline

[bronze island]

Hi Suno Schoolmates

[sick pewter]

Please guys i subcribed and over an hour now still not able to get my credit on suno

[quiet solar]

"shitty results" covers a multitude of potential problems. If you can be more specific about what's going wrong, it'd be easier to help. 😉

Speaking generally, I find it helps if you extend the song near the end (but not at the end) at around the end point of a section, then just add the new parts you want in the lyrics box. If it's losing cohesion with everything that came before, extending just before a chorus can help, as Suno is very good at keeping things very similar from one chorus to another, and that can help anchor it.

[woven ivy]

If you use the method I just mentioned, works every time. And I can link a song with and without the extend

[woven ivy]

No extend cuts the end of when it shouldn't - https://suno.com/s/XrxlXU1n3y7WMIom

extend with the"[end]" tag keeps the song going smooth with a good sensible ending - https://suno.com/s/sA1URWh8YcukHnVS

[polar spade]

Just as a car engine gets us from here to there at 30% efficiency, so does Suno generate good results consistently about ... 30% of the time (often 15%).

So, most suggestions are to help get optimal, but understand, optimal is still about 30%, at its best. But its momentum. And it gets better as you experiment on your own , as well, and get a good ear of what Suno is trying to do.

Agreed as well, that your ear of "sh***y results" could just mean your extends are going longer than you wish when maybe you're just trying to get a button ending..
Or it throws in an instrumental with you just want a vocal outtro. We can know better if you are more detailed. Thanks.

If it means its producing gibberish everytime, or 2 second outputs consistently, these things can have more specific suggestions to help, for instance.

One tip 📚 : get a sense of the beat where you are extending and try to extend ON the beat so the waveform is highest, and so youre splitting at the peak and putting your extend timemark there. This means zooming into the waveform in the edit UI rather than n using Create panel. So, a bit more skill points can get some likeable results. Practice on a few.

In this i would suggest, if not losing something nice musically, to come in from the end 3 to 6 seconds, for two reasons. One, well, seems the devs have redesigned Suno to come in at those points as well, so they must feel this must be optimal point, and two, Suno tries to quantize to blend the new extend into the old output, so it will reach back 2 minutes to get the vibe but also helps to have a few second left and right of your extend point so Suno makes up a more congenial blend going into the next extended segment.

For me, I first, though, always do an extend from as close to the end , as well, by one second such as @quiet solar suggests, as to never leave out the possibility of a nice result from there too.

[azure crane]

I've used the Mashup feature on some old 3 & 3.5 songs and the sound is great, but after 2 minutes it goes from stereo to mono. Does anyone know how to fix that?

[hollow slate]

you must have good ears/speakers, how can you even tell lol

[spice karma]

I got a really good country version of the song I wanted to make and then remixed it over to rock and finally got a good take I'm happy with... except one minor issue (literally). Suno started the entire song (music and vocals) off-key and then corrected itself back onto the correct key mid-first-chorus. I've tried using inspiration mode and combining the original and current take and messing with the influence levels etc but I just cannot get it to correct the key without completely throwing away the rest of the rock version which I really really like. So I guess I'm asking if anyone else has run into this and come up with a good way to prompt it to re-generate basically "the same sound" but in the correct key via a prompt 🤞 cuz I feel like I'm just burning credits at this point lol

[brisk nova]

Take a look at @eager mortar ChatGPT for Suno. Lots of advice and guidance.

[azure crane]

Haha, I have good ears 🎧

[hollow slate]

Does anyone know why sometimes after covering a song Suno doesn't link the artist below 100% of the time?

[high snow]

Glad you liked the song. 😅

[quick wedge]

<@&1082792303976001547>

[polar spade]

Did you mean "after using a persona"?

Maybe im not following. Where would a song even have "the artist"?

[hollow slate]

After creating the cover, clicking on the new song’s title will take you to its song page. Just below the title, you’ll usually see the original artist and song listed.

[polar spade]

OK. Thanks. I go see what ive probably forgotten. Hold on. Ill be right back. Still sounds like youre talking about Remixing another person's stuff. Yea?

[hollow slate]

yes remix/cover

[polar spade]

Not sure.

I did three tests and got the "Remix of" box in all three.

One was just a Cover, which I got a "Cover of :" my own song
Second I remixed a random users song and got "Remix of: " and the user name.
Third I published a song of mine in my Basic account and remixed/cover in my premium account. Still got the " Remix of:" box.

So, so far it seems it should show more than not show. Ill ask around.

[hollow slate]

Thank you for trying it out. It must be that I'm changing it too much. But it varies!!

[polar spade]

I feel there is an emphasis recently noted on that Remix chain of title box to be prominently shown. So you likely would see it than not. Ive asked and am waiting for a response.

Heck, if youre on your phone, it could just be a layout issue too. Change your desktop mode or turn your phone sideways. See what happens.

[hollow slate]

It's so annoying. I have to manually credit the originator lol

[viscid sequoia]

So I used the voice where you can sing the song.. why does it change my voice to something else still?

[hardy aspen]

if i use the cover feature of of suno to cover a song i made on the free version, using the pro version... will i have commercial rights for the new song that will be generated? Gemini tells me yes but Chat Gpt tells me no

[polar spade]

Suno, as generative music program is not a cloning service. Look at AceStudio or others that offer those tools. .

[viscid sequoia]

Oh I thought if you recorded your voice it used it and just puts the beat in background

[polar spade]

You can occassionally get some of its characteristics. Just click menu and "add instrumental" on your audio upload or record of vocal only to get a random idea. And if your voice is pitched and a good melody the music would follow along.

So thats why Ace Studio among other REPLACES the vocal with your voice so that nothing changes. In essence, you're given them a voice thats already optimized to work with.

Suno is always generative, so it takes your voice, adds inflections, maybe vibrato, milesma, pitch corrects it, and optimizes it from the original you give it. What you get won't sound much like you, but you should recognize some characteristics.

So to get your actually voice coming back

1. would sound a bit off (not quantizes, noise, not perfect singing)
2. and to get your actual voice but only pitch corrected would be tool specific as well as risk having that auto tuned effect. Not a Suno thing.

[quiet solar]

The answer is no. The original song created on the free plan is owned by Suno and you don't have the right to use it commercially.

What you then create by covering it is a derivative work of the original, which you can't claim ownership over (same as if you hypothetically covered Taylor Swift's latest hit in Suno).

Potentially you can reach out to Suno if you wanted to discuss options to secure commercial rights to that particular song, but there's no formal process for doing this - it's very much a case by case discussion.

[solid lichen]

New user here. Been using Suno for about a week now, and have the Pro plan. I am having trouble replacing individual words in studio. I'm trying to just go back and replace a word with the same syllable count and it never comes out right. It is always obvious that there was an edit done and the pronounciation is off. Also tried rewriting the song with the changes and generating the song again with the exact same style description and it never comes close to the same melody. My question is 2 fold either answer would help tremendously. What is the best way to go about replacing individual words or can I just rewrite the entire song and regenerate it with the same melody?

[hardy aspen]

but how is my song similar to Taylor Swift's? i generated my song using the free version but i wrote every line of the lyrics myself.... and doesn't suno say that we can use our free songs for our personal use but we just can't monetize them... so like why is it that even if i cover my own free song (with my lyrics) using suno's pro version, then i can't have the new generated song's rights... it doesn't make any sense to me that why they don't give us rights even after we buy their damn subscription... and my song isn't owned by any record label unlike Taylor Swift's music... like I'm literally the only person in the world who has that song and i still can't make a cover of that song using pro version for which i have commercial rights.. its so frustrating

[elder dew]

Evening all

[sullen valley]

You only keep the rights to the lyrics you wrote yourself and brought into Suno, but the song itself created on the free version is Suno's and cannot be monetized by you, and that includes any remixes/covers/extends/mashups of that base song which also cannot be monetized or used for commercial purposes since Suno is the joint owner. You can only use free songs for personal use and if shared anywhere need to have attribution to Suno. https://suno.com/terms

If you are a user of the free or Basic tier of the Service then, you covenant and agree that you will only use Outputs generated from Submissions made by you through the Service solely for your lawful, internal, personal and non-commercial purposes, provided that you give attribution credit to Suno in each case.
and
If you are a user that activates features that permit other users of the Service (“Remixers”) to remix your Outputs (each, a “Remix”), then you agree that all Remixes shall be a joint work owned jointly and equally by you and the Remixer to the extent that copyright vests in any such Remix. Furthermore, if you are a joint owner of a Remix generated by these Service features, and regardless of whether you are a free Service tier user or a subscriber to a paid Service tier, you additionally covenant and agree that the Remix may only be used for lawful, internal, personal and non-commercial purposes, provided that you give attribution credit to Suno in each case.

[quiet solar]

Music ownership and copyright is a minefield, but if it helps to explain your position more clearly:

The lyrics belong to you - still do, and always did. You can re-use those freely however you like.
The music you created while on a free plan is owned by Suno. You have permission to use it personally, but not on a commercial basis.
The combined song is also therefore restricted to personal use, not on a commercial basis as you don't have commercial rights to use the music which is a core component of it.

If you recreated an entirely new song using the same lyrics now that you're on a Pro plan, with no link to the original, that song will be owned by you and you'll have full commercial rights to use it.

If you were to cover the original, you're drawing across the same music. That music is not owned by you, nor do you have commercial rights to use it, so that status would be inherited by your cover version, which is a derivative of the original creation.

[sullen valley]

Yep, if you're dabbling in AI art in any sense, the truth is the intellectual property situation is legally gray. Suno (and other AI services that provide a paid tier) sells you "ownership" that the government may not recognize, while simultaneously generally taking a permanent license to everything you create to train their machine. What isn't gray is anything you can prove you personally created as a human - like lyrics where you wrote them in a notebook or document outside of Suno

[hardy aspen]

Okay i get what you're saying and if this really is their policy, then I'm utterly disappointed. Its like i bought their subscription for nothing. I've tried multiple times to generate the same song or even something remotely similar to it using the pro version but its a different song every time, and i just can't get the version i want. Is there like any other possible way i can get by this situation? i already emailed suno support asking for that song's rights but they haven't responded yet in 2 days. I asked Gemini and here is what it told me to do. It said that i should download the free version song and extract its vocals and then give a 60 second verse from it to Suno to make a cover... then the AI can be tricked into thinking that the vocals provided are my own and not from an existing song... i haven't tried this yet but i do wanna if this will actually work or is it just gonna be another waste of time and credits

[sullen valley]

At the end of the day, you would be on stronger legal footing if you just created a new song from the same prompts and lyrics you used while on free, but while on a paid plan. Any attempts to work around or hack a free song (such as extracting and re-uploading vocals) could put you in a weak position legally because you are effectively uploading audio owned by Suno to bypass their licensing restrictions, which is a direct violation of the Terms of Service. While that "trick" might mechanically work to generate a song, you risk your account being banned or the song being demonetized later if the audio fingerprint is recognized, so it is safer to strictly use the "reuse prompt" feature rather than reusing the actual audio files.

[sullen valley]

You could do a phonetic-rewrite pass on your song and phonetically write out words in ways you'd prefer them pronounced. I had to do that with my Frankenstein Lexicon song https://suno.com/s/jyh90oILsdlMGMwM - then use the cover feature and set it to a high % "Audio influence" under Advanced Options (on desktop). Alternatively, using Editor (instead of Studio), I've had success replacing individual sections - at which point I'd drop in the phonetic rewrite of what Suno keeps mispronouncing

[valid rivet]

Where is the suno sound button. Can't find it

[quiet solar]

Appreciate your frustration, but if worked the way you suggested then people could temporarily upgrade to a Pro subscription at any time and immediately gain commercial rights over everything they ever created before on a free plan. Suno decided that wasn't the way they wanted it to work, and their T&Cs have been clear on this point from the outset.

In terms of Suno Support - anticipate you won't hear back immediately. As I understand it, they're digging through a significant backlog of support requests. Also, this is a very personalised "I'd like to negotiate changing the T&Cs for my personal situation" request - you're asking them for a favour, or a new commercial arrangement, rather than exercising your rights as a customer - so anticipate you might need to be patient.

Taking a step back, what Gemini has suggested you do is essentially an exercise in creative copyright infringement. I'd recommend you don't take that path if you intend to use the song commercially, as Suno would be entirely within their rights to DMCA you and demand compensation for copyright infringement. Distributors and streaming services also take a really dim view of this, so you run the risk of either or both refusing to work with you or host any of your music from that point.

Your options are realistically to either remake the song using completely new music, or to negotiate with Suno over the rights to the original song.

[sullen valley]

On the Create page in the browser, you'll see "Simple" and "Custom" at the top above the lyrics and style boxes. Click the dropdown that is the Custom button to see "Sounds"

[valid rivet]

Thx

[modern raptor]

One message removed from a suspended account.

[neat nexus]

test

OK, this channel wokrs.

[worthy marlin]

Suno school

The channel for Suno students

While Suno principal is the Admin, teachers are Suno Mods

[hollow slate]

HI

[safe maple]

is it true that the suno engine is, "Timbre-Lyric Anchoring"? That we are in pursuit of the style:lyric 1:1? is there an easier way to call the function of the style within the lyric without verbosity? No magic formula, lol

[sour dock]

Is there a reason the Suno chat is broken?

[safe maple]

I see people typing...

[sour dock]

Lol

[timid lion]

Yeah i cant type in there

[solid ivy]

testing chat 1-2-3

hmmm wtf

[timid lion]

Nu-noir super-vox popcorn-ballad big-phonk blooming underlying dueling aura textures permeating effervescent layers tone borrowing the core melody with bright undulating dissonant notes in polyphonic counterlines.

[solid ivy]

is #💬┃general-chat broken for everyone or just me? lol

[timid lion]

I gotta make lyrics

[sour dock]

https://suno.com/s/GgqitHGldYWUQgFm 👀 🤨 💀

[safe maple]

when I responded to the "༺ Noizgenie ༻ — 6:55 PM
https://www.youtube.com/watch?v=ABokQpjvELU" my type is red? I don't know if it posted or not... I see your point

[sour dock]

Yes

[safe maple]

better than my, "[Style: Gnostic Ritual Noir × Post-Industrial Drone | BPM: 178 | Tempo: Relentless | Energy: Explosive Ritual Dread | Intro: Aggressive Corroded Bronze Burst + Dark Matter Sub-Drop | Vocals: Multi-Entity Ensemble, Fry-register Archontic Growls, Piercing Non-vibrato Soprano, Overtone Throat Singing, Guttural Resonant Bass, Distorted Male Whispers | Performance: Schizophrenic Vocal Layers, Microtonal Dissonance | Mix: 3D Spatial Depth, Surgical Mid-Range Separation]"

[sullen valley]

Ah yeah, Discord channels are broken across Discord right now, so nothing is submitting in there

[safe maple]

aye

[sullen valley]

Good video, though, right? 😄

[timid lion]

https://suno.com/s/pz2yPJsZd68gSXvZ

[safe maple]

aye, cheers, I'm of the age where my mother had those kinds of songs on 8-tracks, I think, I grew up near that era, but she was an Abba, Kenny Rogers * Rita Coolidge fan, so the many of the hippy stuff was sparse

[sullen valley]

Yeah the mid-20th folk revival era really put out some gems

[safe maple]

thanks, I'm always looking for new, cool styles

[safe maple]

mosdef

[sullen valley]

As it happens, I did make a variant of another song of mine in that style https://suno.com/s/V4ANRw19U9BNRU2b

[timid lion]

Bruh what is this banger https://suno.com/s/vph0k5j8zGgbJtdB

Double zero shot. I had to mash because the solo gen was so good

[tacit gazelle]

Omg for a sec I thought my discord was acting up

I can’t message in the suno chat server😭

[wise kindle]

Hah don't worry, it's not you - #🔦┃server-spotlight message

[tacit gazelle]

Ohhhh okay this makes sense

[wintry apex]

How do you make suno Wait and not sing all of the vocals immediately and ignore all the [Instrumental bridge] [instrumental] tags. it just skips them and goes straight for the lyrics, singing all of them immediately, and ending the song in 1 minute

i just dont know the tag or command im sure.. should i be using parentheses instead or < > ?

[wet minnow]

you coluld try to add some gibberish symbols (nothing that can be sung) ... sometimes this causes suno to make music / interludes or other effects instead of voice. ... and commands go in [metatags]

[wintry apex]

oo maybe ｆｕｌｌ　ｗｉｄｔｈ would work

since if the prompt is in full width it just maktes random genres instead of the ones it spells, so i dont think it can read that

it might actually try to speak in tongues though

[blazing arrow]

Hello to everyone reading this. I hope I'm posting in the right place. I'm using a translator, as I don't know English. I need help with a promt. I want the vocalists in my song to sing at the same time. For example, one sings calmly, like Billie Eilish (in a whisper), and the other vocalist screams in the background. I don't know how to achieve all this... please help me.

[wintry apex]

[Both: (whisper / scream)]

Duet: one vocalist whispers the other vocalist screams simultaneously whenever [Both] is in the lyrics

[wintry apex]

maybe

[quiet solar]

That's quite tricky to do through prompting - calling for harmonies or non-singing lines in specific places can be hit and miss, let alone going for two non-singing harmonies at the same time. I'd try @wintry apex 's suggestion, for sure.

Alternatively, if you have access to Suno Studio or are comfortable using a DAW or an audio editor, it would be a LOT simpler to approach this through a mix of prompting and editing:

• Start with a version of the song where you focus on getting "Billie Eilish" to whisper the line and don't worry about the second singer
• Create a new version of that song (e.g. by Extending from a little before that point) and change all of your prompting to focus on changing over to the second singer screaming the line.
• Extract the instrumental and vocals stems from the first song, and the vocal stems from the screamed version and layer the screamed vocals on top of the whispered vocals, adjusting each for volume balance as needed.

[polar spade]

Are you Pro only or do you have Premium and access to Suno Studio?

[oak maple]

Would you recommend writing the entire chorus lyrics only once when they first appear in the song lyrics and then simply referring to them, e.g., [Chorus] (as above) when they are sung again, or would you recommend writing them out in full each time they appear?

[restive pebble]

I wouldn't. Suno tends to follow the lyrics as written and though I've never tried it the way you're describing. Suno would sing the parenthesis text as an adlib or background vocal. it may be a good way to getsome really cool layering and textures for the song, so play around and have fun finding what gems Suno provides you.

[sleek stump]

You need to write out all the lyrics, Suno doesnt do looping references in the lyrics section. Write them all out with their proper tags.

[restive pebble]

Much better articulated @sleek stump

[sleek stump]

It could be cool if Suno could do that reliably, but people would use that to overload the lyrics window with stuff that would go well beyond the current 8 minute limit 😄

[grim glade]

Do you need to use style/audio weights to make Personas work well?

[polar spade]

Its placebo effect but many of my recent songs start and end interestingly enough now that they sound loop-y if I have the repeat on. I almost cant hear where they have started over.

[polar spade]

If "work well" means to clone a vocal better, then thats not a thing. No cloning will happen no matter what.

But yes, adjusting sliders are necessary if you are working with an output or covering an older song. This will determine how much influence the Persona has and how much is influenced by the styles, lyrics and old output.

In that 70s range is a good place to change vocals over to the persona version and only weave a bit of the persona instruments, if using a Legacy Persona.

[blazing arrow]

Thanks everyone for your help, I'll try. 🤗

[stiff sentinel]

give it a more detailed breakdown of the intro in the style section

and i use square brackets for style direction in the lyrics, it never gets those mixed up as actual lyrics

Intro: Light uplifting piano chords, guitar joins shortly after. A sudden drum fill leads the intro into the first verse

[bronze island]

Hi Suno classmates. I'm about to make a Disney style song, wish me luck

[raven turtle]

Getting a Suno brick wall at the moment where no matter what I do it’s giving me nothing like the prompt.
When does dark, reflective electronic ballad become
Upbeat synth pop? Come on Suno, 200 credits should have got me that.
Also getting that old v4 vocal belting out crazy singing again.
It’s leave it for a day time I think.

[eternal nacelle]

This is getting pretty bad with the way generations end with a long tone of noise. I know I'm not the only one getting this bug.

[oak maple]

Okay, I ask because I've seen someone else doing it, but I don't know anymore who it was or if this approach guarantees 💯 success every time

[stiff sentinel]

its more reliable to write it out entirely....though suno will take some liberty in repeating lyrics nonetheless

[quick wedge]

<@&1082792303976001547>

🧹

[polar spade]

Yep. Seems a report in ticket-help as well. We'll blame the sleet and snow for awhile. 😁

[rose jasper]

You actually don't have to include those parts on the text, Suno recognizes by itself which are the verses, pre-chorus, chorus, refrain, bridge, outro etc. I only really use them now, if Suno doesn't hit the mark anymore. But with v5 it's improved a lot

[lost scaffold]

If you Remaster a song made for a remix contest, does the remastered version still qualify for the contest? What if you edit it in Studio?

[polar spade]

Most rules for the remix give a minimal task, such as you have to extend or make a cover. And if they dont say anything but Remix it, then as long as you use one of the ... what... four tools in the Remix menu, you should be able to do whatever else after.

In the end, its being judged, so we never know what the judging criteria really is, after we meet minimum requirement.

[stiff sentinel]

yes, as long as the chain exists it will recognize it

i did cover > studio > cover and it recognized it

i also tested cover > studio > remaster and it recognized it

[raven turtle]

Now getting odd noises at the beginning of generations and two minutes of a high pitched tone at the end. Things really aren’t right at this end… hmmm

[raven turtle]

I’m now getting the same thing!

[earnest agate]

Please post an example song. I’m not getting this. What model are you using, v5 or v4.5 all,+,pro? I’m curious to hear it.

[raven turtle]

I’ll have to retrieve it from trash. Has a whining intro sound which has started appearing on all generations, and as the song ends I’m Getting a 2-3mins of a square wave kind of tone. Using v5 with a vocal persona.

[knotty stag]

I am creating a song, and have made several versions, still trying to get it right, and sometimes, it glitches. It sounds very bad, words missing, repeated, sound dropouts, like a really bad recording, mostly using v4.5 and v5.0

[earnest agate]

ok that is wierd. are you using any tags [intro] etc. in the lyrics field? is the style not causing this?

[polar spade]

No.... happens on personas.

I changed the link out to another song. Just referenced segment labels in the lyrics. Mundane style prompts.

As things go, its just a bug or better, a 3 -day glitch. So, after the snow storm and weekend, they should be looking into it.

[earnest agate]

yeah, but looking at that i am not Suno knows how to end this "song" make be add some cues in syles and in Lyrics field [tags]?

[earnest agate]

yeah, maybe it is just a bug

[polar spade]

OK. So there it is. No sweat.

[earnest agate]

😃 yeah, So if it is a bug, is it in the prompt or the prompting of the song, or in the song structure that’s triggering the bug? i guess we wont know that

[quiet solar]

"Introspective" might be worth trying in addition to, or in replacement of "reflective". It's pretty unusual for that to get referenced closely with upbeat music.

[quick wedge]

<@&1082792303976001547> 🧹

[polar spade]

Yep. Introspective has been my go-to word for quite a while. Calms vocals down and give more of an interpretive [another good prompt] feel. Good for a Billy Joel vibe as well.

[raven turtle]

It’s in conjunction with the v5 vocal persona not being recognised and used by Suno and this odd 2-3min of audio noise at the end of the generation.
Suno is having a beef with me today! 😂

[polar spade]

Ahhhh. OK. Yea. Good time to pay bills for me then.

[keen saffron]

Does anyone have prompts for strong female vocalist, not the soft type I’m looking more for something as strong as Sheryl Lee Ralph who sang blood sweat and tears, and no matter what I write it doesn’t give me anything like that any help

[stiff sentinel]

well idk that singer, but i used "raw and powerful female vocals" and got this https://suno.com/song/65d79eb0-ec44-480d-b8fc-504fe4f3dded (skip ahead to 3:53 if you don't want to hear the whole thing)

oh wait it was "deep and raw"

[queen linden]

No cure for the Suno disney princess voice so far. lead vocal female a raw, weary, and gravelly low register alto female, distinctive, husky, like an old folk singer. In Styles swooping, rising, soaring, moaning, humming . Your not going to get a Beth Hart, or Grace Slick deep rasp... but you might get something better to listen to

[raven turtle]

Just had another rendered track giving me 2 mins of square wave signal noise at the end of the track and beyond.
Filled in another feedback report about it.

[vague cliff]

Static noise common today on intro

[hollow slate]

Any idea what that horn/brass instrument uses sometimes? it doen't sound that good at the end of tracks

[raven turtle]

Yup!

[vague cliff]

Not just me then

[vague cliff]

Have you noticed persona voice not staying consistent either?

[vague cliff]

Static noise, persona lack of consistency in voice, seems to started around when you sent this message

[raven turtle]

Yep! It’s not working at all for me

[vague cliff]

I think it’s best leaving it for a day, hopefully be resolved by morning

I’ve had it working perfectly all week

And only today earlier on it started going wild

I also noticed it switched between different persona even though I selected a persona that has no link to the other

[raven turtle]

Agreed. I had it working so well for a short time and then boom… crappy noise at start and end of stuff and the persona totally screwed

[vague cliff]

I thought I was being targeted specifically there, so it’s nice to know it’s not just me

Hopefully it’s a temporary issue

[raven turtle]

Welcome to the messed up Persona and noise survivors group!

[vague cliff]

My workflow has stopped but maybe it’s time for a break anyway 😂

[wintry saddle]

oo a place to talk about prompting

[vague cliff]

I think it’s fixed?

[sour dock]

Suno dead rn lol

[barren crystal]

Fighting the website trying to create one song without it inputting a different error/defect each generation

[safe maple]

Very nice, I've finally had the chance to listen to it. It 'almost' reminded me of Runrig, but it was much more easier to listen too, lol, Kudos. Thanks for sharing

[high snow]

Has anyone here done rap music before? I was wondering if you could look at my lyrics to see if I spaced out the call and response correctly. And don’t worry my lyrics aren’t the final prompt/look, I’m still tweaking things a bit.

[quiet solar]

Feel free to post your style prompt and an example of the lyrics where you're using a call/response (please not all of them) and I'm sure people would be happy to give advice/feedback based on what you're trying to do.

[high snow]

Ty.

[high snow]

Here are my two ends of my song, and I want to know if having these two call and response this close is okay or overkill….

[FINAL CHORUS – full power]
Armor on, yeah, we battle ready
Stand firm, never fold, stay steady
Shield up when the arrows fly
Sword in hand, let the Word reply

Armor on, yeah, night and day
Peace on my feet, truth round my waist
Helmet locked, mind set free
This ain’t our fight — it’s victory

Crowd (chant under last line): Victory! Victory!

⸻

[OUTRO – chant / drop-ready]
Leader: Armor on.
Crowd: Stand firm.

Leader: Already won.
Crowd: Already won.

Leader: Who’s the victory?
Crowd: Jesus!

Leader (final): Armor on.

[quiet solar]

@high snow I'm assuming you've already mentioned that there should be a crowd chant in the style prompt? That would be a good starting point, if not. I'm not sure if the leader is intended to be a separate vocalist to whoever is singing the rest of the song - I've assumed for simplicity they're not.

Suno can pick up some context from how lyrics are structured, but it'd be worth making it clear that "Crowd" (and possibly Leader) are intended to be prompts rather than lyrics, so I'd change that Outro/pre-Outro section to something like this:

[Crowd chant] Victory! Victory!

⸻

[OUTRO – chant / drop-ready / call and response]
Armor on.
[Crowd chant]: Stand firm.

Already won.
[Crowd chant] Already won.

Who’s the victory?
[Crowd chant] Jesus!

Armor on.

---

Outside of that, yeah, the structure looks fine for trying to help Suno understand what you're trying to do in terms of call and response. Prompting for crowd/chant vocals can be a bit hit and miss, so you might need to give it a couple of tries and tweak what you put in the style prompt to get it working - you may want to include 'call and response' in your style prompt too.

As a general rule in the current models, prompts in the lyrics box are a lot less reliable than style prompts, so for anything complicated like switching vocalists, you ideally want to avoid relying purely on the lyrics prompts for it.

[sour dock]

Oh nah 💀

[sullen valley]

Glad you liked that 🙂 Thanks for listening!

[high snow]

Yes, the lead singer is separate from the crowd chant, but these lyrics aren’t the final product. This is just a rough draft.

[balmy kite]

https://grok.com/imagine/post/2dfa9c0e-b7eb-4824-a751-3625146bef93?source=copy_link&platform=ios&t=a16559fb8b8c

[calm plaza]

don’t think we had to see this big man

[sterile lion]

Hi all, I contacted support more than a moth ago, and no response so far. My question was:
" I'm struggling with getting the model to sing the lines I set for each singer in a duet.

I tried labelling the lyrics for each section of the song, I tried labelling each line, I tried to set the rules in the style prompt, but the results are quite frustrating so far.

What am I doing wrong? Is there a right way to get the model to sing different lines in different voices (in this case, only two voices are needed)?
If not, what workaround do you recommend that is supported in Suno? "

[restive pebble]

Good morning,
Duets are a challenge in Suno but can successfully be accomplished by either running a generation with one vocal, then a second generated version with the other vocal finally splicing them together in studio if you're a premiere member or an external DAW.
The other method is to try to make Suno behave by using the styles and lyrics section to influence it.
Something like : This is a call and response duet with a male caller, baritone raspy vocals and a female response in a soulful tenor vocal delivery.

Then in the lyrics
[Verse | Male]

or something close to that.

Suno can make the latter happen but it takes fiddling and lots of credits.

[neat nexus]

If scripting it:

[Baritone male vocal] I say I say
[Soprano female vocal] He say she say

with a descriptive style: duet, male vocalist, female vocalist, call and response

doesn't work for you, you may need to try forcing it with extends. Try with the vocals for the line described as male or female depending, get the vocal expression you want, extend from there and alter the style and metatags for the next part and try until you get the expression you want. Go back and forth extending until you reach the end and then cover that trying to get closer to the musical expression you're looking for.

[restive pebble]

Here's a link to one I made using the long way, before I learned about splicing two gens together... https://suno.com/s/XVGCY1AWxxb8qk8K

check the styles and the lyrics to see what I used. YMMV as Suno doesn't definitively support it

[sterile lion]

Many thanks! Other problem is, I have a song that's already a remix of an earlier song, and for some reason remixing messes up the editor really bad (showing incorrect duplication of lines for example, and the timeline is totally false, if I highlight a word to change, it will show a different part of the lyrics). Is this a known issue?

[restive pebble]

Personally, I found the editor to be inconsistent and not providing the edits I wanted. Instead I use Cover which works really well. I push the sliders for weirdness and style low and audio around 70%. This usually provides a good enough reproduction of the original but with the lyrical changes I entered.

[polar spade]

You must always "match lyrics to audio" first before making each edit. I'm fanatical about that, but saves lots of headaches.

• not only remixing but any use of the editors can always rearrange or double up or remove the lyrics. Ive also a habit of copying the whole current lyrics before using editors. Just paste them in after doing an edit, then sync up again to make all lyrics are there. When done editing that one, copy lyrics BEFORE saving file.
  Then reload file , paste in lyrics again, and do your next edit.

  Just make the workaround a habit and protocol.

But at minimum, sync first so that lyric selection and waveform selection match up.

The sync tool is in the upper left three-dot or hamburger menu of both edit UIs.

[sterile lion]

omg, thank you so much!

[polar spade]

Yw. Its still a learning curve, but frankly just watching the magenta highlight go through the words, and getting it right, is just satisfying.

[sterile lion]

Sadly it does not work even after MAtchin Lyrics to Audio. I highlight a part in the lyrics (around 2:50), and the editor highlight the song part starting from 0:40.

and vice versa, if I highlight the waveform in the editor, it will highlight a different line in the lyrics

[quiet solar]

You mentioned issues like duplications of lines - make sure you've resolved those first (e.g. using Edit Displayed Lyrics on the song outside of the Editor) or it can cause some oddities.

If you were hoping to step in a do a surgical strike/change on some lyrics in the Editor, I'm just going to ring the alarm bell now that it's not always a straightforward exercise, and you might well want to try the seer's suggestion about using a high audio influence Cover instead.

[sterile lion]

I did resolve it during the Match. And I'm noot hoping anything, I did really easily change single words or letters in the past. It works flawlessly on originally generated songs. But on covers, it's a mess.

Anyways, thank you for responding. I'll figure out something. For the covering the cover suggestion: sadly it a 100% made an even bigger mess, that also cannot be edited because of the wonky timeline.

[polar spade]

Its hard to reproduce the problem, but ill see if i can make a video of it working correctly. Both not working and working, if im lucky.

But the steps are:

° Click "align with audio"
° Go ahead and paste lyrics into the "corrected lyrics" side.
° Click play and watch magenta "karaoke" highlighter hover over each word as it plays/sings.
-- Every missed word, stop and type in the corrected lyrics. If you find doubled lines, delete one from the corrected lyrics box. Continue until the highlighted words match both the audio timing and the actual lyrics.
° When done, click save.
° before selecting. Review the lyrics again to make sure some line spacing hasnt been deleted in the save. SUNO likes to do that.

Then highlighting the words to be changed should highlight the approximate same area on the waveform.

Again, if youre right, then its glitching. So ill check again later today to see if things are in order.

[polar spade]

Oh? Edit covers makes the errors even after aligning and editing lyrics. OK. Ill check that too then, but I think ive seen it not discriminate between covers, extensions, any version frankly. Ill check though.

[dusty coyote]

Hi all. Does anybody know of a mostly-effective way to make the songs longer?
More specifically electronic music, that could easily go near 5 minutes, but sometimes are under 3 minutes... Suno tends to rush on each music section, so is there a way to tell it to stretch a bit more?
Thanks

Sometimes it throw a drop in the last ~25 seconds of the track and ends it kinda abruptly... A bit unpleasant to hear.

[polar spade]

Just a heads up, for 8 minute top limits, 5 to 6 is optimal and average , at least on covers.

Some prompts can take an output longer on average. I know "ambient" is one.

What are these, new original generations or covers of outputs?

[dusty coyote]

Hi @polar spade, thanks for reply.
These are originals...

Most of my tries go around 3 minutes

But yeah, some gets over 5 minutes, but it's rare

[plush field]

song length mostly depends on the song "organization" and "amount" of lyrics

for example, this will be longer

[Intro]

[Verse 1]

[Pre-Chorus]

[Chorus]

[Verse 2]

[Chorus]

[Verse 3]

[Pre-Chorus]

[Chorus]

[Outro]

than this one

[Chorus]

[Verse 1]

[Drop]

[Verse 2]

you can play with "short edit", "radio edit", "dj version", "extended version"... in style, but as i wrote song structure and lyrics are more important than that

i have around 2500 generated songs from 1:04 to 7:59, so i think i should know what i am talking about 🙂

btw. the same thing can be used for instrumentals as well, but use meta tags instead of lyrics
for example, something like this

[Intro]
[Sad Trancy Synth Intro]
[Soft Pad Atmosphere]

[Drop]
[Sad Trancy Lead Solo]

[Verse 2]
[Soft Beat]

i am not sure how much this (squence metatag) is effective, but you can try this too - put it at the beginning of lyrics

[sequence: intro-build, verse-1, pre-chorus, chorus-anthem, post-chorus-drop, verse-2, chorus-anthem, outro-fade]

[small swan]

I have been using this for.some.songs

[Instrumental 12 bars]

Someone knows if it really helps on the duration in the instrumental?

Been hearing almost the same duration, i know it depends sometimes on the genre

And some variants

[Instrumental interlude]

[Instrumental interlude 8 bars]

[plush field]

i am not sure how much Suno will respect the suggested number of bars, to be honest

[small swan]

Mhms, i think the same, maybe i need to try another way to make long instrumentals on songs

Been mixing
Interludes and instrumental solos, but normally suno ignores one

[plush field]

try something like this

[Drop]
[Sad Trancy Lead Solo]

[Lead Solo Continues]

[Lead Solo Continues]

[Lead Solo Continues]

[quiet solar]

I did a set of tests with it, and the answer is "not reliably".

I can't prove it one way or the other, but I suspect Suno disregards it much as it does when you request a specific chord or chord sequence, or anything else that's just more detailed than it's currently designed to cope with.

[plush field]

this is just an idea, i am not sure if it work or not

[quiet solar]

You can always create it with lyrics and instrumental sections to push the length more reliably, then purge the vocal stem out using Studio or a DAW of your choice. And then optionally remaster the whole song to avoid any perceived quality loss from dropping the vocal stem.

[small swan]

[wintry saddle]

this kinda works, but changing tags / cycles of tags may work better 🙂

you also have another game - pre Studio, the way it processed input into the channels was different, so you have to recheck all pre jan-mar(ish) 2025 prompts and assume the will not behave the same.

but with V5, you can literally put zero prompts and get a 3-5 minute instrumental only song (not same thing I guess), mid song instrumentals, 2 options -> extend is your friend to keep pushing instrumental solos, you can always meta-tag up sections with 'Progressive' or other styles that like long instrumental solos... or even risk throwing them in your main genres (but that might mess with sound too much!)

These days, instead of one way, we usually have 3,4 or 5 ways to achieve any task - some songs / genres do better with one way over another - if you have the credits / patience, i'd spin a few versions testing not just new prompts......but whole new flows within suno to achieve the same goal

[dusty coyote]

Thanks @plush field... yeah I tried those... but as @wintry saddle said, and I just tested, these doesn't really do much on track length
I just did many tries with tags in lyrics (this song has no lyrics) or blank lyrics box, and average the same track length
Sometimes I get longer tracks as well, but it doesn't seem that is due to something we can control

[polar spade]

As Viking has put in his example, "Trance" , I recall that is another prompt which often lengthens.

So, though we often mention the "blank tagging" and spacing technique, remember to test them individually as well, to see what works well enough alone, and then what works in combination.

Prompts can work like inbreeding. Use too many in one shot and it just snaps back to the original puppy.

So, we have "ambience" and "trance". Stick either or both of those as fifth in on your prompts, see what happens in 6 outputs (3 gens). Write down the results. Try the next one. Get back to us.

[polar spade]

Less is always un-surprisingly surprising since more weight is fully on the fewer prompts. So even zero prompts can just hit the loto. Very true.

[wintry saddle]

new lines certainly mean stuff to the AI - length? not always - but extra new line breaks (or removing them) can fix duet weirdnesses (sometimes)

[slate fox]

if you want you guys can use my beat maybe ai might cook

[neat nexus]

Test post.

[high snow]

Hey, I got a quick question for anyone, so I’m just using mashup now, when you’re mashing up a song with another one, does the advanced options part with the weirdness, style influence, and audio influence does that carry over from the first song or the second one?

[quiet solar]

I believe the sliders affect the ultimate output of the song, rather than linking to only one or the other of the song inputs.

Weirdness and the Style Prompt definitely affect what happens to the mashup overall rather than applying to a particular song. I've not tested audio influence in depth (feel free to try and let us know), but I would assume it works in the same way - affecting how much the original audio of both songs is reflected as they originally were, rather than creatively interpreted.

[neat nexus]

Sorry I'm the wrong one to ask this to as I do not mashup my stuff the way that is intended.

[craggy swan]

I'm looking at getting Suno and had a few questions I was hoping someone could answer for me. If I were to purchase the Pro yearly plan could I upgrade to the Premier yearly plan at a later date if I need to? Also Is there a way to make album covers on Suno?

[quick wedge]

Yes, you can upgrade from pro yearly to premier at a later date whenever you need to. You can use the cover image art feature for your album covers that way.

For my images I use Bing Image Creator to create for my songs though but there’s other ai art site options out there that you can experiment with too. Suno have articles as well if there’s anything else you wanted to know. https://help.suno.com

[sleek stump]

If you are asking about doing covers of songs on an album, Suno does not allow users to cover copyrighted works, there are lots of legal issues with that. I would suggest reading through Sunos terms and policeis here https://discord.com/channels/1069381916492562582/1372994765196361774 before making any decisions.

[high snow]

Ty

[high snow]

OK so you can Suno for free to make songs but you’re very limited, if you pay for pro a month it is $10, depending where you are in the United States it’s $10. Outside of it I have no idea. You could also do a yearly plan if you want to do that but that’s kind of like $90 maybe $100, then you have the premium plan which is $30 a month and there’s also a yearly plan for that too. I use the pro plan that’s pretty robust on what it can do, but if you’re looking to use studio, that is for a premium users only.

[high snow]

And yes, you can make album covers on Suno. And there’s also this new feature where you can take two songs and do a mashup to make a brand new one.

[sleek stump]

I think there are 2 different kinds of "Album Covers" that wasnt specified. If its the artwork for an album produced in Suno, there are options to do that within Suno. If they are refering to covering songs for an entire album not produced in Suno, that is not allowed.

[craggy swan]

Thank you for the information.

[high snow]

Np

[slender glacier]

So it doesn't show you how many credits a song takes, and does the 5v cost more? Seems like they're going faster with the 5v as opposed to the free plan

[sleek stump]

Each song generation costs 10 credits, (2 song versions, 5 credits each). This is hard coded and not adjustable. It also doesnt matter which model you select, the generation cost is the same.

[slender glacier]

Is there a way to see the exact amount of credite you have? Because im sure i didnt make 100 songs and all I can see is 2k left

Oh lol I just clicked in it, I just need to be more observant I guess and pay more attention to the math as I go

[sleek stump]

Yep, in top right of teh create window

It used to be down in the left side bar, but they moved it.

[bronze osprey]

I usually use "growling female vocals" and get a great gravely voice

So..

I my self have also a prompt question...

Im working on a song, that i want to sound like a internal battle, with a voice soft voice doing the verses, and an angry voice in the chorus. (Its about internal duality, where both voices are the same person, but diffrent sides/persona)

Where the angry voice wants the other voice to "weep" (janken in dutch)
And calls her weak, etc...

I tried many things, but...cant get it to work as hoped, does anyone have suggestions on how to try ?

[sleek stump]

That is pretty much a Duet, I see Kanthric typing, I'll let him field this one 😄

There are lots a ways to do duets, some work better than others.

[quiet solar]

Ginger's absolutely right, it's got a lot of the same characteristics as doing a duet - even though the singer is the same person throughout.

It'll be tricky to do if you're reliant solely on prompting in the lyrics to switch between the two emotional states, which I imagine you've run into, since Suno doesn't always take into account lyric prompts. If you want to do it through prompting alone, I'd suggest doing something like this:

[Shouted, angry] LYRICS FOR THE ANGRY VOICE HERE
AND HERE AS NEEDED

[Soft, pianissimo] Lyrics for the other voice here
And here as needed

But, if it's only switching in the chorus, I'd be very tempted to focus all of your prompting efforts on getting a soft voice for the opening verse(s) before the chorus in a song, then Extend that song from just before the chorus and switch over all the prompting to focus on getting an angry voice for the chorus. If you then Extend from that subsequent song, you'll have established a pattern in Suno of soft voice verse and loud voice chorus which you can hopefully get it to continue for the rest of the song. If not, continue a similar pattern of extra Extends just before it deviates from what you want so that you can fix it.

Then Get Full Song at the end and clean up using a Remaster if needed.

Alternatively, if you're a bit more comfortable with editing and have Suno Studio or an external DAW/audio editor, there are ways you could do it by creating one version of the song focused on a soft voice and then a second (e.g. Cover, high audio influence, same key signature in the style prompt) focused on getting a shouted voice for the choruses, then replace the choruses from the original song with the second one - either entirely, or just the vocal stems.

[eager kelp]

how do u guys get ur music out do u use distrokid?

[quiet solar]

@eager kelp Depends where you're looking to distribute it.

If you want it on the major streaming platforms, you'll have to go through a distributor - Distrokid is fairly popular and newcomer friendly, but it's not the only one to provide a similar service. Who is best really depends a lot on why you're producing the music, your expectation of it's commercial success and how much music you intend to release over the coming years.

Whoever you might use, it's worth being aware both of the distributor's requirements around submitting music and the requirements of the streaming services you intend to publish to; some have begun to introduce specific elements around AI-generated music and distributors will typically assume you've taken that into account and done everything needed when you're making the submission.

[bronze osprey]

I did not concider this...somehow...

[tacit gazelle]

Okay so I made this anime song but not sure about the melody “real real” in the chorus should I keep it or change it? https://suno.com/s/7lJ6EPiWxANKWfTb

[polar spade]

This isnt a critique, but just a personal note. Since this is in line with an idol style singing, as well, I'd actually like to hear Real Real a couple of places more to give it an anthem feel we can sing along to. This would be in line as if sang at an Idol Cafe or Theatre but still be perfect for anime purposes.

[timber robin]

Does anyone know why my songs keep generating and cut off in V5? It's like if I do 10 generations it all ends at like 7:51, or all 5:52, and the song isn't actually finished, even if I do like 4 generations and wait for them to finish. The lyrics aren't long.

[polar spade]

Are they still good songs... or are there long beeps or noise in most of the second half?

There have been some issue recently with songs going a full length, near 8 minutes but with some noise near the end. Mostly when using Personas, is the guess. Whatever is causing that may be bubbling up as other anomalies to other users.

If these are first generations, then no sort of news for having both short and long formed outputs with abrupt endings, so far.

[stiff sentinel]

have you noticed any such reports of this today?

i know there were a lot yesterday

i'm waiting for it to get resolved before i dive back in

[polar spade]

Its been reported as a bug, so there may be some resolve. personally, I havent come across anyone noting still an issue.

But again, this above user just described something similar, so it might have mutated.

[pure panther]

I've noticed some weirdness today, I set the weirdness slider to 15% and the audio influence high and it behaved like the weirdness slider was super high, lot of strange artifacts and completely unusable. Increasing the weirdness slider seems to improve it, oddly.

[stiff sentinel]

huh, interesting, maybe a dev accidentally added/removed a negative when adjusting weirdness stuff?

did you try anything at like 90% weirdness

[pure panther]

No 70% max today

[stiff sentinel]

and they were pretty... un-weird?

[pure panther]

Less weird than 15% by a lot

[stiff sentinel]

i see

thats weird, but weird in a more predictable way lol

[pure panther]

It's workable 😄

Just... weird

lol

[polar spade]

As mentioned, its a known issue of glichy buggy generations. Likely have to just wait it out in the same boat as everyone else.

[stiff sentinel]

ohh nice they finally made an announcement about it

which channel is that from? its not in announcements nor updates

[polar spade]

Bug channel

[pure panther]

I think it's different than that. I was dealing with the issue generating earlier, most of the them were errors and the rest were just loading wheels until I finally just deleted them. That seems to be resolved now.

The weirdness issue is something I was encountering after I could finally generate again.

[polar spade]

Thats good. 👍 🎶 One down, one to go, another town, and one more show. 🎶

[sleek stump]

Yes.. I'll just "Leave it" there 😉

[stiff sentinel]

anyone ever make "live" songs with audience shouts? what would be a good tag? [Crowd Chanting] ?

[high snow]

Honestly, I was just tested out. I think that honestly would be good. There was someone who did a live concert. But I totally forgot who their name was. If I find out, I’ll tag you.

[wise kindle]

I was just messing around with this if it's close to what you're looking for. Gives off that live concert feel. At 2:13 you can hear the crowd take over. Feel free to take a look at the prompt!
https://suno.com/s/CPkXED4aBTlhyT5I

I should warn you, the lyrics are completely random lol

[stiff sentinel]

ooh yeah, thats kind of what i want, except sprinkled over the whole song

[wise kindle]

You could use that prompt, then try to further coerce the crowd in the lyrics with [Audience Singing] or just [Audience] seems the style prompt is already aware of an audience. As with most prompts, it's just suggestions and isn't guaranteed to work, but this should get you pretty close

[stiff sentinel]

https://youtu.be/bC22p3o1TYQ?si=g0YlNuqXoynbEp8u&t=111 this is what i'm going for

[wise kindle]

Ahh I see. The crowd are almost bringing in a touch of harmonic support by chanting the secondary part

that's quite cool

[stiff sentinel]

yeah, they invented the secondary part actually

kinda just happens in that genre/niche

just occured that i could do part of it manually in studio too, then cover it

perfect use of the new sounds feature

[wise kindle]

https://suno.com/s/RlyUTnmP8NOLb2mv how about something like this? (random lyrics again just for testing)

I basically told the style prompt I wanted the backing vocals to be sung by the audience then used (backing: "whoa-oh, whoa-oh") in the lyrics when I want the crowd to sing

[lunar ruin]

Class is in session!
Hi-Ya! will be your teacher!

Experiment!

[stiff sentinel]

ooh yeah there's so much more crowd activity there with those prompts, i didn't know you could tell the style so specifically how to parse the lyrics

[wise kindle]

Yeah it's pretty clever! With that format you can essentially sprinkle the crowd wherever you want them

[stiff sentinel]

yes. will definiely try. hopefully one day i can experience that kind of energy live lol

[wise kindle]

haha yeah it would be amazing to experience that kind of energy for real

[stiff sentinel]

and thats the audience, imagnie being the singer in a stage in front of thousands and they're all synchronized

in the japanese ones they coordinate all the lightstick movements too

[wise kindle]

I was even more impressed when you told me the audience made up the secondary parts. having everyone so synchronised and able to create fitting melodies on the spot is outstanding

[stiff sentinel]

yes, idk how they do it. probably on like twitter or something LOL

[polar spade]

@stiff sentinel If you begin your style prompts with "live performance" , that triggers a variety of style.

Adding then [audience] in the metatags, even just once before the first verse and after intro. Can be enough.

[Clapping] is another thst gives a more pronounced and more intimate effect than just the hiss and roar of an arena packed crowd i.e. ampitheathre vs. rock concert.

Audience
Crowd
Street crowd
Applause
Clapping

So far have word best, where "coffeehouse" has never really triggered much, though I keep trying.

For your idol theatrics kind of atmosphere, you thrn have to put audience response in parentheses.

Here we go
( yea!)
We leave today
We know
(yea!) Its
The only way.

So yea, that video is a "live performance, anthemic, arena, call and response" style of performance.

https://suno.com/playlist/464151ee-6332-4a10-a658-2fcf69f6be26

Did a Playlist of a few. I tried to do an inspire and Mashup but the generations failed.

Any song that's about 8 minutes is only about half that.

[tacit gazelle]

If you’re an anime lover I have a playlist that I’m building/adding on to full of anime opening songs inspired by anime shows like Naruto shippuden,Bleach,one piece ,attack on titan,JJK,Black Clover,etc: https://suno.com/s/wdCLir1s3RCiONZc

[polar spade]

The Backing vocal.prompt worked very well.

[polar spade]

From a request, We are suggesting ways to get live audience responses into a crowd. They just so happened to want an idol pop kind of song as well.

If you have any thought on getting live audience , call and response type sounds, youre welcome to suggest what you know .

[polar spade]

Man, waiting for Suno to generate 8 minute songs take a long time, I found out.

[wise kindle]

Can't remember the last time I got an 8 minute song, but I imagine it would xD

[polar spade]

Its this bug thats been happening.... half a song and the other half just a single tone. So get alot of 7:59 songs. Two came up in my playlist. About 5 while generating them. Arrghhh!

[stiff sentinel]

i'm getting good bits and peices with all these tips, but getting everything right is kinda like try to get the male and female exactly the way you want for duets

[polar spade]

As you said, piecemealing them together in Studio and be something more controllable perhaps.

But its the moments when the audience has to sing on cue. Thats going to be harder.

@wise kindle has the right form, i think that will keep it producing a response even towards the end, which Suno has a tough time maintaining consistency as it get towards the end.

[stiff sentinel]

https://suno.com/s/ErFfsFflg6acDhxy getting closer....

[stiff sentinel]

hmm, it shouldn't cost 50 credits to extract stems from a max 20 second sound clip...

[wise kindle]

I do wish the credit cost was a little more.. dynamic. It feels bad spending 50 credits to get the MIDI transcript from a song that has one track (stem) lol

[stiff sentinel]

does the midi work properly for you?

i don't have an editor, but when i play it back only the guitar one worked iirc

[wise kindle]

Yeah! if you're curious, here's a short snippet of some piano using the MIDI I extracted from Suno. I'm using the Nice Keys Ultimate piano. It came out really clean

[stiff sentinel]

oh yeah that's pretty...nice

vocals are post or suno?

[wise kindle]

I mentioned in another channel yesterday, but I like to use Reaper to delete notes under 1/32 (depending on the song). This can clean up trailing notes and notes from reverb which causes that stuttering sound you generally get from Audio to MIDI conversions

the vocals were just just straight from Suno. I didn't do anything to them

[stiff sentinel]

i legit cannot tell if vocals are human-recorded or AI anymore

[wise kindle]

lol yeah it's bonkers how human-like they sound now

[polar spade]

I took your output, hoping you wont mind and replaced all "crowds" with "audience" in the lyrics and just ran a cover with nothing else, no styles, default sliders.

If you listen to a few songs having either crowd or audience, you can hear "audience" has more of a spectrum of frequencies like different people singing, and "crowd" has a more unison and lower pitch sound.

In this sample it seems the audience has more spectrum and adding harmonies, and the crowd is more all on the same note. A small change and really just a personal preference once all is said and done. But dont forget to experiment like this a little with a simple tweak.

https://suno.com/s/yOu5dq8S6gzqyvpy

[stiff sentinel]

you noticed the same across both generations?

[polar spade]

No the first one has a closer sound to your original.

https://suno.com/s/Z6ADitCIzGMIDOeV

Ive tested a bunch last year between Crowd and Audience and feel there's a difference when there IS a difference. Theres good points to both, but it can make a slight difference on the vibe I think.

Now with Studio. Stem them out and run a mash up or inspire and even get the best of both. Use as a track in Studio. Then make a persona from that and feed back into the lyrics for another version.

Maybe its alot for a subtle micro tweak, but fun to test at least. If nothing else, the you'll have a wide choice.

[stiff sentinel]

yeah i'm doing a ton of testing right now, trying to generate songs that are like "only" chants so that i can just snip them up and stick them in in studio

i'm thinking for workflow something like this
--> make a good song, only focus on the song, dw about the chants (but make it the kind of song that has space for them
--> write chants for the whole song
--> use studio tryhard mode to do the chants for the first minute or so manually
--> remix/extend/inspiration/w/e to get suno to do the chants for the rest of the song

[polar spade]

As a courtesy, ill delete the one I did of yours, since its not mine to use. But just throwing a sample your way to compare.

[stiff sentinel]

keep em if you want, the song was a slop dummy song just to test things

[polar spade]

All good ideas. I hope they give you a variety of options then.

[stiff sentinel]

i have audition too, so i can probably tweak the chants to get them to sound the way i want

reverb, detune, chorus mistiming, etc.

[polar spade]

Sounds like a winner plan.

[stiff sentinel]

yeah, its nice now that i have an idea of what suno can do regarding these chant things

a really nice thing if anyone is trying the same thing, using the full 12 stem separation, the chanting is pretty well isolated on the backing vocal track

[polar spade]

Start saving them in a Playlist and of course download to your fx folder.

[stiff sentinel]

yep exactly

[half dragon]

You guys don't sleep? 🙂

[stiff sentinel]

and the 50 credit cost for separation even for really short audio? merge them into a single audio file then do it in one go 😂

[polar spade]

Hazzah! Good idea.

[stiff sentinel]

sleep? what that]

[polar spade]

Yea. True. I should call it. Have a nice night y'all.

[stiff sentinel]

night, thanks for the help and for experimenting along!

[polar spade]

And great work. You have a cool song brewing. Wish you the best with it.

[quiet solar]

Pretty solid idea for a workflow, although from experience I'd be a little wary of putting too much hope in Suno picking up the Crowd chants as something to extend out beyond the first minute and get right. There's a chance it'll pick up the pattern and replicate it, but also a good-to-fair chance it'll ignore it and focus on either the main singer or just switch up randomly wherever it feels like it.

As you said, this is a very similar challenge to getting duets the way you want them, so I'd approach it in exactly the same way and plan it out in advance - do a single pass of the song focusing just on the main singer getting the instruments and vocals the way you want them; just make sure the lines intended to be given to the crowd are well-suited for it (ie, short and uncomplicated).

It'll be a more reliable way to ensure space is made for them if you include them from the outset rather than relying on blank spaces.

Then in pass two, just focus on getting the Crowd right. Do a high audio influence cover or Extend near the end of the song and rebuilt the entire song in the extension. Because you're only interested in the vocals, you don't need to worry about the instruments varying a little. I'd personally try with Extend first as it's less likely to introduce key changes (although that wouldn't be an issue here) or tempo changes, as they can make your life a lot harder in the editing phase. You can try to mitigate that in Cover if you check the key signature and BPM of the original song and add them into the Cover's style prompt.

Then you just need the instrumental and vocals stems of the first song, and the vocal stems of the second song - or various pieces of the second song depending on whether you need to take mulitiple attempts at it. Then switch the vocals over in editing as needed.

And then I'd echo Dasvon's good luck to you. Your test version is already sounding good. 🙂

[cloud umbra]

Adult chat dm

[tidal prism]

Does anyone have tips to get specific vocals, it feels like every genre has a set generic vocal and it’s so hard to get it to change off that

I feel like vocals are the thing where it doesn’t really listen to my specific instructions at all

[timber robin]

only time those had the noise for me is if i did a cover.

[timber robin]

Like [scream], [whisper], [softer tone]?

I think some do well in brackets in my case.

[tidal prism]

No not like that, I’ve got that stuff down, it’s hard to explain

I’ve gotten some friends into Suno and I hear some of their metal songs and other peoples personas and it’s just like “oh, they all sound extremely close to my personas voice”

They might be doing different things like screaming or a different sub genre but all the voices sound VERY similar to each other

[oak heart]

any updates on SSC?

[oak heart]

try accent tags like [irish vocals [deep male vocals] [raspy vocals] etc

[quiet solar]

@tidal prism Can't say I've run into that problem in a big way with songs I've made in the same loose pop-punk/alt-rock genre. They all sound plausibly different from each other (albeit typically biased towards having soft West Coast US accents).

I suspect this is an area where experience might vary a lot by genre, but there are definitely things you can do to influence it. On that note, genre is a good starting point and subtly influential on lots of different elements of songs, so prompting for additional sub-genres and/or adjacent genres may help influence the nature of the vocalist.

You can also directly prompt for the style of vocalists - a "youthful, gritty tenor" should sound very distinct from a "golden-era smooth baritone", or a "rebellious punk aesthetic lead". Lurking within that is a blend of vocal types (e.g. tenor/baritone), timbre/texture (e.g. soft, gritty, breathy, nasal) and tone (e.g. rebellious, youthful, warm, aggressive, conversational), all of which you could experiment with separately.

Again, depending on genre you can pull in production effects like vocal chops, echo and reverb that can give a different overall tone to the singer, even with any underlying similarities in the voice.

Also, the nature of the song will have a significant impact. I've noticed Suno will deliver very different vocal performances for high tempo, high energy songs or sections compared to slower, more melodic harmonies. Again, always worth considering playing around with that kind of mix - and if it's difficult to prompt for, crack out a microphone and give Suno an audio guide based on your own voice before you have the singer transform it.

The more things you leave to Suno to pick out defaults for a genre, the more likely it is to land in similar spaces to others (or your own creations) with the sound and vibe it goes for, so deliberately trying to steer away from that with prompts is probably the way to go if the underlying genre is causing you headaches.

[tidal prism]

You’re the goat ✊🏻 yes that makes sense, my language when it comes to describing vocals isn’t that great so I need to learn more about all the different words used to describe vocals, I think that’s were I might be having problems, I have only had Suno for a month and a bit and I pretty much spent the entire time trying to figure out how to make it scream consistently. I finally figured that out

Now I gotta learn all the vocal techniques and what they mean/sound like so I can describe what I want properly

[quiet solar]

I don't know if there's a good reference source for that in the music world, but there definitely is in the wider world of voice acting. Not all of that will translate across to singing, but it might be worth taking a look over sources from that space for inspiration, e.g:

https://voiceactor.com/articles/voice-over-descriptors
https://www.wikihow.com/Ways-to-Describe-a-Voice

PS. Glad I could help. 🙂

[polar spade]

Thanks. Yes. Devs are looking into the issue. I was getting the same last night on just first generations. So I thinks is just a backend issue of any form of generations.

[polar spade]

Yes. You should follow them to get the updates. Heres the latest. #1270307712181469245 message

[polar spade]

Pulled from youtube: List of scream types. Might come in handy for prompting as well.

Fry Scream (with variations)
Tunnel Throat Scream
False Scream /Grit/ False Chord
Heavy Metal Falsetto Scream
Gutteral
Singing with Distortion
Old School Growl
Mid Range Growl
Low Groel (no lyrical legibility)
Inhale Scream
The Goblin
KVLT Scream (like a banshee sound)
Hybrid Scream

As is best, ill refer you to @forest hemlock who's well versed in this subject --

Heres one comment by @forest hemlock : " . . . Style prompt. An example could be death metal, deathcore, high-pitched fry screams, shrill distorted vocals, piercing fry technique, upper-register screams, strained high screams, searing vocal attack . . . "

[queen arch]

Can someone tell me why it it making every song i make 7 min and 59 sec. Im making 80's hair metal with persona. Aslo what is it getting so muddy and distored 1/2 to 3/4 the way though the way the song most the time? This is driving me crazy!

[restive pebble]

I cannot give you a cause but know the des are aware of this and they're working to identify its cause and resolve it..
A couple of things to try to get Suno to stop is to do the basics of browser stuff. Log out close browser log back in... etc..
Inside Suno there's a few things to try. Not saying any of these will work, but could help knock Suno back into its proper output just by changing where its 'memory' for lack of a better term is stuck in.
Try keeping the sliders closer to their defaults. Suno sometimes gets a bit lost in itself with higher settings.
Shift the entire genre/mood/energy of the song. Run a few gens and then revert to your desired/original
Move workspaces and remove styles altogether. Run a few gens and then revert as above.

With Suno being the powerhouse it is and with all the features it provides, I expect it to go rouge. The distortion issue for me has most always been high slider settings. When I return them down nearer their defaults, the distortion goes away or reduces considerably.
I'm sure this will be something looked into as well knowing the quality of audio Suno desires to achieve and maintain.

[waxen jackal]

I put my raw vocal recording done on mobile without metronome timing. I want Suno to make it's cover with same melody but tighter tempo as my recording is like 7.57 minutes. What should be weirdness, style, audio level settings?
If I am keeping weirdness more like 20-30% it's rushing on too much, if I keep audio like 70% it's keeping same tempo. 😭😭😭. I don't see prompt sub chat

[quiet solar]

There isn't a totally reliable way to prompt for tempo, but the sliders probably aren't the most effective tools for it.

In the first instance you can try specifying a BPM in the style prompt. Alternatively, you could speed it up on your end to the desired tempo before uploading to Suno and that should help.

PS. Suno School is the new home for what used to be the Prompt Development chat here. 😉

[polar spade]

Is a known bug. They are working on it. Dont burn credits if you can help it while waiting. But if you need, crop off the useless ending bit. Usually its a long sine tone.

[shy badge]

Hey all. I've updated the Suno Gemini Gem. A Gemini gem is like a mini chat gpt. It is editable so you have access to the original instructions. Hope this helps and inspires.

https://gemini.google.com/gem/1fasDFlkGZKfr3J3eIxSPBGTAC2jEOfyk?usp=sharing

[polar spade]

ChatGPT 5.2 is nice in this way as well.

[shy badge]

All you do is select the lyrics you want. It can come up the lyrics or you put your own lyrics . It will not change your lyrics unless you tell it otherwise. Same with styles.

[waxen jackal]

Thanks. I am trying it by trial and errors

[gloomy iron]

Is there way for me to get from the begging of this post so i can read all the comends for all the possible suno commands that you have typed ?

[hard beacon]

if we did a list of all the possibile suno commands then.. well... see youn in the year 2142

🙂

[quiet solar]

There's no limit to what you can put in a Suno prompt, so there will never be a list of all possible "commands". However, some prompts work more effectively than others depending on how easily Suno can look them up using the music it was trained on.

A number of people have compiled lists based on their experience (great) or based on what LLMs think might work (sometimes great, sometimes terrible). A number of people have found Stayen's reference docs quite helpful: https://github.com/stayen/suno-reference

PS. It's best to think of them as 'suggestions' rather than commands, as commands implies they will always work in a predictable way - you'll soon find on experimenting with Suno that this just isn't the case, because of the way the underlying model works.

[brittle socket]

Well. There is ONE limit.

...the character limit 😆

[viscid ravine]

Hahaha

[polar spade]

OK. Everyone take a break and try the NEW "sample" mode. I've run a couple and its .... interesting.... could be fun to play with and get some good outputs.

Fun so far to just use a sound effect and let it write a song to it.

[polar spade]

I've figured out, its not like Mash-up. You dont add a Sample into an existing song. So you have to either prompt for a new song . . .

or... seems you can use a whole song as the "sample" and with the same lyrics, well, it seems to hmmm... like.... umm... make a new Cover?

Try and see, I've only played with it about 15 minutes.

[frozen dagger]

Are you referring to "Sounds (Beta)" as the sample? Sounds like you referring to something else. Which case not sure where to look. I'm aware of the "Sounds" addition in the Custom drop down.

[polar spade]

#📢┃announcements message

[frozen dagger]

nice. ty for pointing that out. will def play around with that.

[polar spade]

You can drag in from the workspace, but easier for me to just choose from the menu, under Remix/Edit

[frozen dagger]

They may want to consider mentioning the lower credit cost with the Sample feature. Looks like 4 credits, similar to Mash-Up, which I believe it was mentioned for that feature.

[sleek stump]

Mash-up is currently on an introductory discount, it will go to normal at some point in the future.

[frozen dagger]

I know. Just saying may want to include that it is at the discounted rate, as they did before, to maybe encourage the use.

[brisk nova]

GMIV is a great tip from ChatGPT by @eager mortar check it out. tl;dr GMIV = Genre · Mood · Instruments · Vocals

Think of it as the minimum viable prompt Suno actually responds to.

[toxic pasture]

I dont think it matters what type of music you're making. I've tried everything for a few days and my songs are all 7:59 and sound like crap on top of that. It's useless to me at the moment.

[polar spade]

yes. A known bug started about Saturday. They're working on it.

[toxic pasture]

Thanks, that's good to know. I used the down time to organize and such, but I would prefer to make music again

[polar spade]

I literally got another one just now... 7:59.. song cut out about 2:50. lol

[toxic pasture]

I just got 4 in a row and stopped, again. I can trim the end, that's fine, but the rest of the song is not clear and has issues, distortion

[polar spade]

if they would have a split second of silence , thatd be good.. but most of mine slowly fade into a chaotic sine wave.

[toxic pasture]

true. I can still fix that if they must bug it up, but it has a very slight sine wave over the entire song. subtle, but the sound is there over my song.

[polar spade]

Sorry about that. crossing fingers its fixed soon.

[timber robin]

REAL

[high snow]

So I have these SFX sounds in one of my lyrics and I know that it has done something similar like that before when I didn’t ask, and the AI just completely ignored it. Did I do something wrong?

[Bridge — Thrown Down]
[Half-time | severe]
[SFX: distant thunder, pounding horse hooves approaching]
[SFX: dogs snarling and barking, chaotic but restrained]

From the window of her splendor
She was cast into the street
Blood ran down the palace stones
And soaked the horses’ feet

Chariots rolled without mercy
Over pride that called You blind
And the dogs consumed her body
Till no trace was left behind

[SFX: hooves pass, dogs fade into distance]

[frozen dagger]

prompting with text for SFX is hit or miss. would try the new "Sounds" feature to generat and then add to track maybe.

[high snow]

I don’t know what new sounds feature you’re talking about. Are you talking about the mashup?

I’m so lost

[frozen dagger]

#📢┃announcements message

"Mash-ups", "Sounds" and "Sample" are all new features.

[high snow]

I honestly never noticed that. How do I do that exactly? Does it like only add to one specific section?

[frozen dagger]

Works same as generating whole songs, but instead you're creating short clips, choose either "one shot" or "loop" for the type of sound and use text to prompt. then you would need to add these to your song with editor/studio (or other creative ways; like using Cover even), but whatever works best for your use. In the case you showing would have to add it with editor/studio probably, or another DAW.

[high snow]

How would I add that to my song through editor?

(Sorry, I’m new to this)

[frozen dagger]

double checking if editor can do it, i go to studio most often.

[frozen dagger]

looks like would be Studio to add the sound on another track of the timeline. i don't see a way to add other audio clips from your library within editor, i thought there was a place to drag them in from. my mistake it looks like. def can do it in studio or another DAW.

[high snow]

Okay. Ty

[patent leaf]

how do we use sample to generate a full song? its just making 4-10 second variations

[wheat tartan]

probly cover it

[spice brook]

Is it possible to have the chores but then have a guitar solo after to be the same melody as the chorus how would I prompt that?

[frozen dagger]

select the audio of the chorus, cover it with a prompt that says Guitar Solo.

[spice brook]

Does this work?

[frozen dagger]

Kinda what suno does best is giving it an audio sample and then a prompt that transforms it. Duplicate the chorus, then cover the duplicate, remove the lyrics and replace with [Guitar Solo] and/or change the style prompt to add Guitar Solo or just Guitar Solo by itself. It should use the audio (your chorus) and reinterprate it the way you prompted

if the melody is in the chorus (like vocals) then you may need the exta step of seperating the stems to grab the audio that has the melody you want the guitar solo to emulate.

[junior glacier]

i ran into issues like that trying to use a short sound as a sample, either the song was roughly the length of the sound, or the sample stopped being used after its length, and then the "song" continued....but when i sampled a 20-30 second section from another song it worked well, tried slider and audio percentages in a variety of ways but it didnt really alter anything

[crimson thorn]

Guys whats the perfect percentage for the “Audio influence” slider when covering a song? I want the same flow and vocals, but not the same lyrics. (Having the audio percentage sometimes messes with the ai and causes it to use the same lyrics)

[sullen valley]

there's no perfect percentages. I've found 15-35% variously good, and 54-66% also

[restive pebble]

There's no 'right' setting to say Suno will give you this exact copy. It is a random generator.
With that said, I have had the vocals be mixed if the sliders aren't where Suno wants them.
It differs between songs. One song I had them closer to their default setting and got the mixed lyrics and on another I had them pushed up on the audio. I played around and eventually Suno got the lyrics right while keeping the lyrics I sought.

[restive pebble]

Do you mind not using this channel or server as your personal advertisement?
It is disrespectful to the Suno team but moreso to all the users

Please remove your post

Y''all should read the TOS and discord rules for this server

[untold briar]

These are almost always hacked Discord accounts, not a real user. At least if the user has a history or normal chat on this Discord.

[restive pebble]

thanks @untold briar

trying to keep it civil, but it is almost 4am

[untold briar]

It goes without saying, but if your Discord account gets hacked and you get banned for crypto spam, if you recover the account you can message us mods to get unbanned.

I used to DM banned accounts with this message, but there's just so many lately, haven't been bothering.

[restive pebble]

makes sense. to much overhead to maintain it

[untold briar]

Also, consider turning on Two Factor authenticaiton for your Discord account!

[restive pebble]

Right. Makes a big difference in security

[waxen jackal]

How to get best results for raw mobile recorded vocals into Suno output. Want to keep melody. What levels for weirdness, style and audio?

[restive pebble]

There's no 'best' results only what slider setting gives you the closest rendition of your uploaded vocals.
To start slide weirdness down to around 15-20, styles leave at 50 and audio bring up higher to around 60-75.
This should get you close to your uploaded vocals, but you will need to play around until you find the slider-setting that 'sounds' like your upload.

[wintry saddle]

no, but SFX are super hit and miss, as mentioned earlier, you might have more joy if you declare multiple SFX channels and try and do it through the studio... but even then (you can try with sounds and have a flavour of what SFX you can get) - thunder is probably the easiest thing to work with on that list and its still a sub 10% chance in my previous experiences (and that was making personas for it etc... many months ag

[wintry saddle]

Yes - you can Extend or Cover it and it gives some really fun results!! tbh as a base sound....

you need enough prompts in your 'new part' to ramp the time up....

if you really want instrumental only (and its being stubborn from a sound...) you may find it easier to add some throw away lyrics, then cover / remix and remove them -> or delete the vocal line in studio to get to a longer base

[low breach]

Whats up everyone, hope everyone is enjoying Suno. Whats the topics of discussion.

[quick wedge]

Hi 👋 I’m enjoying Suno alright. Hope you’re doing well with your morning.
Suno School talks about prompting and the editing tools and features.
For casual conversations you can look in general chat instead https://discord.com/channels/1069381916492562582/1069381916492562585

[low breach]

Hey, glad you are good and enjoying Suno. Nice yeah was meaning whats the topic of discussion in Suno school. What sort of music have you been making lately?

[toxic pasture]

I'm certainly not enjoying. Every new song is 7:59 long and sounds distorted. Fix this or stop charging me, geez.

[quick wedge]

I create with various genres except hardcore metal music

[viscid ravine]

Do you use EDM?

[low breach]

Ah nice thats cool. I will check out your stuff whats your handle?

[quick wedge]

I have one song that’s edm

[quick wedge]

Mooniewolfe is my Suno username

[low breach]

Ok will follow you now and check out your stuff.

[viscid ravine]

Show me

[sleek stump]

Can we stick to the topic of the channel please? There are other channels for general chit chat and song sharing. Thanks!

[quick wedge]

I linked it to you in the other channel

[quick wedge]

Thank you 😊

[fair adder]

hello. i wanna try to sample a part of a song to integrate it into another clip. but besides selecting the part i want to sample, i don't know what to do. there is no clear lebalized button on how to proceed on how to use my sample in any way.
ps: i'm blind

[raven turtle]

Hi. Really enjoying the possibilities with Mashup atm.

My secret sauce…
Sing into Suno to get two audio files.
Then Mash them them with 100% influence.
Then split the result and make a persona from the vocal stem!
Boom! A pretty close approx to my own voice!

[half dragon]

I haven't used it a lot but initial usage seems that you select the section to sample then create a new song with it, same way with lyric and style normally

[fair adder]

... so basically like a normal uploaded audio?

[half dragon]

I also noticed that its length depends on the section length , short section -> short generated clip

[fair adder]

that doesn't really answer my question though...

[half dragon]

I would say similar as upload audio but not quite the same

[fair adder]

if you can only use it to create a new clip then for me it means it's like an uploaded clip, and if so i think it's lame.

i have an almost complete song but the end is not what i want despite several gens so i was hoping i could try this since what i want is actually litteraly the same as the begining of the song xD

unless somebody know how to trigger a lead vocal only section, with no backings or instrumental. that's what i'm trying to achieve

[half dragon]

Yeah still experimenting with it, but I just got longer clip in recent generations so... the initial theory is out of the window ...lol

[stiff sentinel]

everything you're describing is easily doable with Studio

integrating clips from another source into your song, making a section vocal only, etc.

[fair adder]

except i don't have studio

[stiff sentinel]

without studio it'll be really hard

the only luck i've had with using clips as part of songs if its the intro, then suno will repeat that theme/sound throughout

i've never been able to make suno use a recorded clip in the middle of a song

[fair adder]

and i don't want to pay higher for a feature i'm not even sure to be accessible/blind friendly/screen reader usable

[stiff sentinel]

alternatively, you could do that kind of editing outside of suno, then upload it back in and re-cover

[fair adder]

bruh... boring

[stiff sentinel]

you can do really good 4-stem separation locally

[fair adder]

i prefer play the suno slot machine, i have a looooot of creadits to spend so let's go xD

[stiff sentinel]

boring solution is still a solution

i mean, if you have money to spare then premium lol

[fair adder]

which i don't lol

[stiff sentinel]

but yeah, for the intro i'd put that into the styles as well

Intro: acapella, 8-bar, lead singer only
Sudden band entry at Verse 1

and of course tag appropriately in the lyrics as well

[fair adder]

yeah

[stiff sentinel]

luckily suno is most cohesive at the beginning of a song

[fair adder]

i'll think of it next time, but right now i'm making extentions

[stiff sentinel]

if anyone has found a way to control suno well at the end of a song... please let me know lol

[fair adder]

xD

thanks for the help anyway. i know i'd love to give a try to studio but like i said, it might not even be accessible for me. the editor isn't, so... meh, i'm not sure

[stiff sentinel]

i get a raelly good song, but then at the end the vocalist wails on aah-ing for 16 bars and the exit melody is completely different from the main melody

idk if this is helpful, idk exactly what accessibility options you are used to, but here's a screenshot of something i'm currently doing in studio which is on the very involved side

[fair adder]

... i said i was blind. that means what that means lol

[quiet solar]

@fair adder For what it's worth, reading over what you're trying to do, I'd have gone for exactly the same approach tsunderestorm suggested. Suno is very good with audio feeds as prompts, so giving it something 95% close to what you want and asking it to cover will be a lot faster than trying to get it there by prompting and rolling the dice.

I'm not really sure how you'd steer it there by prompting alone.

[stiff sentinel]

you're using TTS to use discord?

[fair adder]

yeah but that means recreate an entire song and probably get a new melody, which i don't really want

[fair adder]

i am using a screen reader in my daily life. so, yes

[stiff sentinel]

ah neat, never really enountered anything like that before so wanted to clarfiy

[fair adder]

it's ok lol

[stiff sentinel]

if there's any basic studio function you want me to try out for you (e.g. extract all 12 stems) i'd be happy to do it for you

[quiet solar]

@fair adder If it's as simple as wanting the intro repeated at the end, and all that's holding you back is you don't have an editor with decent accessibility features to do a cut-and-shunt yourself, just shout. It only takes a moment, I wouldn't mind helping you out.

[stiff sentinel]

likewise, if say you want only the vocals until a minute into the song, that's like a 30-second job in studio

...plus the processing time xd

[fair adder]

it's fine, i can extract stems, but that's not what i'm aming at. well... maybe in the very end when i have no choice then edit outside of suno, or just cut that final part which is just a craze of mine to add at the very last minute

[stiff sentinel]

extraction is to give a good base for suno to re-cover

like if i want a keychange, i do it in studio at the point i want
it doesn't sound quite right because usually there's a leadup to the keychange, but once I re-cover it suno puts that leadup properly

[fair adder]

like if i activate de remix feature and allow you to edit it? if i give you the time stamps and the section i want to reuse, you could do it for me?

[stiff sentinel]

(and you don't neeed studio for all this, it can be done not too difficult outside studio, its jsut the common thing everyone here knows about)

[fair adder]

well i'm probably the most casual of users lol

[stiff sentinel]

for simple edits i don't mind... things that are simple enough that i don't need to clarify

[quiet solar]

If you send a link to the song, and give me the timestamps for the section you want to re-use and the timestamp for where you want it to come in, that's all I'd need.

No need to activate the remix feature

[stiff sentinel]

like if you say "i want the guitar melody from the chorus to be used again in the ending" i probably can't do that, since i don't really have your vision

but if you say "remove the drums from 1:21 - 1:54" i can do that

[fair adder]

mmh. it could be worth the try, at least. are your dms open?

[stiff sentinel]

should be

[quiet solar]

Yeah, my DMs are open

[stiff sentinel]

oh oops wrong person lol but yes mine are too

[fair adder]

haha thanks tsunderestorm but yeah i think i'm gonna head towards kanthric for this one. but thanks for the help anyway :3

[stiff sentinel]

sure, let us know how it turns out

[low breach]

That sounds awesome bro I have not used the feature with my singing yet. I may just do that soon sounds great. What is your channel handle?

[raven turtle]

I don’t publicly share tracks as I use Suno primarily for studio demo work

[polar spade]

I find people following me though my profile is blank. I get the reason. But it does still feel odd sometimes. 🙂

[rapid yew]

👍

[high snow]

I need anyone’s feedback real quick, so I’m gonna do a recording of my song that’s like rap Ish, and I’m just asking if it sounds like it’s saying “bows” or not, because I had to go and fix it cause it mispronounced it and it’s driving me nuts and I can’t tell if it’s actually saying it correctly. Because when I replace that one word, now in like the lyrics when it shows up on the left side it spells it as “owz” and I did the phonetic spelling of bows to bowz.

I hope you can hear it well.

[rapid yew]

It sounds like bows

[high snow]

Okay. Ty. I honestly need just someone else to confirm. I thought I was going crazy. Thank you.

[rapid yew]

Your welcome

[white pecan]

Hey everyone 🙂

Devs-chat doesnt exist anymore ? :p

[quick wedge]

They never had a developer chat before only that it was called prompt-development

[white pecan]

Hey Pim, well...That's true :d the Prompt-devs chat

[quick wedge]

They reorganized the channels some got renamed

[white pecan]

oh, that's why, alright, thanks you

I just had a question, I used Persona + Inspo (4 tracks) and suno wasn't able to generate, is that normal ? "credits refunded each time" :/ but with only one option of these two everything was good, normal issue ?

[quick wedge]

From the starting discussion from this one then downwards it seems like the persona + inspo combo isn’t there any more #📚┃suno-school message

[vestal beacon]

Hey! I'm creating music for my game and need 3 tracks that feel cohesive ( Main menu, shop, combat). What's the best approach?
My current thinking:

1. Create one "hero" track first that sets the vibe (don't have this yet..)
2. Use that as an audio ref (or is persona better?) for the other tracks
3. Use different prompts for all the tracks.

Is this the right strategy? Should I aim for the same BPM across all tracks, or just the same key? Is it possible to do this w/prompting? Any tips/tricks are greatly appreciated. Thanks!

[tough crown]

Hi all,

Can Suno interact with metatags that are written in Dutch?
Thx in advance!

[half dragon]

I think it can - basic song structure tags should work consistently - others might or might not- even in English 🙂
Try it and play with it - and document it for your own work - there is no rule.

[cloud olive]

haha so funny.

[quiet solar]

That's a pretty sensible workflow.

I'd suggest using your 'hero' track in Inspo rather than using a persona (which now is more geared towards replicating vocals).

You can include prompts for BPM and key signature in the style prompt and v5 will often take that into account (v4.5 a bit less so). However, unless you're blending the tracks directly into each other as the scenes transition I don't think it'll make a big difference if those vary - and you may well want a higher energy vibe for say, combat, than in the shop.

If you're after overall cohesion, you probably want to keep some of the core genre prompts the same and play around more with tone, tempo and feel. And if you really want to lock in cohesion or common refrains, do a couple of covers from the original song where you deliberately try to mix up the sound and vibe a bit and use them as a collective set of 'hero' tracks for Inspo.

[high snow]

Hey, so I need some help, I’m new to using premium and I’ve isolated a track that I want. How do I add it to another song in studio?

[restive pebble]

You could open the song in studio, then drag the other song into a new track and slide it to the location desired.

[half dragon]

I think you can drag and drop from the Library view into the Timeline view from within the Studio. Check filter to show stems.

[high snow]

That’s after I upload the audio right?

[restive pebble]

yes,

upload your audio, open in studio then drag the other track

[high snow]

OK, so I successfully dragged and dropped where I wanted it. But now I have two different audios colliding. How do I fix that? And how do I make sure that this edit that I made saves to a new song?

[restive pebble]

there should be an add track button
Add a new track and use it for the second audio. Align them how you want them but on different tracks so you can manipulate them separately.

[high snow]

I already added the track. I’m just wondering how to blend them seamlessly.

[half dragon]

Use fading on each clips to blend them if they're musically related and if you cut them on beat.. maybe it'd work better too

[restive pebble]

For that, it has to do with BPM matching, splicing on the down beats which is beyond my scope of knowledge. I know a little about engineering audio.
There are many on here who are much more skilled than I who could guide you more effectively.
We have the channel #1420792719969878028 and someone probably has the info in there.

see like @half dragon knows more than I about it lol

[frozen dagger]

if its different BPM, click add new track, move the 2nd section down to the new track. click that track to be able to view clip details on the right panel, in that panel change the BPM to original. IF that is the issue of that noise. "How do I fix that?" could mean different things here. 🙂

[restive pebble]

betcha OBLMOUS is going to straighten this out for both of us

[frozen dagger]

by default when you add anytrack (even the original audio that you brought into studio) the default is set to "On Beat" which will match the original you brought in so that works, but when you bring another audio in with a different BPM it's still default to "On Beat" so it is trying to match the same BPM which can caus it too sound....off.

IF by "fix" and "audio colliding" you want to create a transition that is a different bag of tricks.

Oh, I meant to add that the "On Beat" and "Original" options for the BPM, are per clip. So if you have 2 or more sections, or clips, check them individually.

Which means you can leave it on same track if you want. I thought BPM was per track when I started jumping in here.

[high snow]

That’s so confusing. Why the hell did they have to make this so damn hard. Even pro is better than this. As in for understanding.

Like, wtf, all I’m trying to do is do a simple try to blend one audio with the other so it doesn’t sound like shit

And I don’t even know if these edits that I’m making in the studio or not they’re already adding to a new song or not. Or do I have to export the entire song.

[frozen dagger]

I found the default setting annoying, still kinda wish I could toggle, but I've ran into times where I'm like "yup, that is prolly for the best and I need to just get used to it"

[restive pebble]

It isn't adding until you export the whole song or theindividual tracks

[frozen dagger]

I'm jumping in with BPM check. But I'm trying to leave that window open of, what is it you really trying to do? When you say "blend". Just not sure what you're trying to have happen there.
Maybe an extend or, recreate\cover is more what you're wanting.

[high snow]

So what I’m trying to do is there’s one green track and then there’s one orange track, the orange track is from one song that has alliance roar at the end, and then the green track is the one I originally liked, but it doesn’t have that at the end. And I wanted to isolate the roar from the second song, and then added onto the first one, and then export the song. And then blend it. But then I just realized I could just do that after I export.

[half dragon]

maybe extract the stems and put the roar stem on top of the other clip? then export

[frozen dagger]

There is a form of Version Control built into Studio shown here. Where it says Versios. This is every autosave made to your project while you're in Studio. (I am not fluent in this but it exists.) Basically the Undo path as well, but better. So alll changes are saved with Version Control. But when you ready to say ok I want this song, Export.

[high snow]

Is it possible to explore a outside sound from Suno? Like if you want a specific FX sound?

[half dragon]

You can upload and use it in studio or use new feature Sound to generate sfx and use it in Studio

[frozen dagger]

or upload your own (not copyright protected) audio.

[high snow]

I did that.

I found one that’s decent, but I was just asking in general if I could get a sound like outside and use it

[high snow]

I did try that SFX sound effect. I got one decent one, but I just wanted to know in general if I could go like to YouTube and import a sound effect and use it in studio. Does that work?

[frozen dagger]

image i added and like @half dragon said the Sound feature can crete soud effects now, new feature

oh no... basic answer.

What you're talking about I would just put in a box of "sampling" and sampling is prtty much known as "not mine" so, I'd play with that outside of Suno to be safe. They do not want you using anything from someone else's stuff.

[half dragon]

Term of Service says you have to own your uploads - so idk... if it's wise to use someone else but use at your own risk

[high snow]

Well, I just need a clear audio that literally doesn’t sound like it came out of somebody’s ass

The worst thing is when I try to isolate the stem from that one song, I can’t even isolate the vocals. So I have vocals mixed with the lines roar, and I don’t want that. And the decent lion roar that I have is puny.

[frozen dagger]

It's a general safety net for them, they have the right to use our stuff if they want, if we used something that wasn't ours.... and they used it... or someone else did..... just not what they want at all

[high snow]

Well, what about like a sound effect you would get from my YouTube? Is that just not allowed? I mean they’re there for that reason are they not? I mean, people use those meme effects from YouTube, etc. all the time probably not in Music but….

[frozen dagger]

have you tried layering and increasing the volume, slowing the speed, or extending the audio of the lion's roar?
I jsut did some monster roars and instead of using one it came out much better using 3 and adjuting audio. Even get to have roar in left channel and anothe in right, or panning roars.

[high snow]

I’m honestly not smart enough for that especially in studio, half of that is hell of confusing to use. 😅

[frozen dagger]

it's a slippery slope. Suno says just say nope. Use another DAW along with Suno if you really want to sample and not risk reprecussions later. That's my plan. I'm happy to have the Sounds. Glad they added a feature we wanted for a while.

[high snow]

Alr. This blows

[frozen dagger]

Back to your original post with teh video. I'm gonna say what you want to do is plac the clip where you had it, next to the first. click the 2nd clip. then at the bottom of the screen use the drop down menu where it says recreate, choose "cover" 4 credit to see if it keeps your lions roar but belnds it to the 1 st section

[high snow]

Alr, I’ll try and let you know

[frozen dagger]

when you hover you rmouse over it, this is what it sounds like youre looking for. minimal prompting proabably best so it uses the audio. theere is an advaanced optionms thing down there as well kinda hidden with sliders to adjust to match auidio more or less

[high snow]

When using the new SFX thing, how do you clear it up to make it sound more crisp? Should I just use remastered for that?

[high snow]

Ty

[frozen dagger]

remaster, any other relevant tricks with normal Suno audio, but Remaster and EQ within Studio probably most effective. Drag the Clip of the Sound Effeect onto the timeline, select the clip, then at the top of the detials screen on the right select "Track" and yo can now adjust the EQ. cycle through presets and manually adjust as needed. Low pass, or Lofi preset maybe for "crisp"

[high snow]

What is EQ?

[frozen dagger]

EQ= Equalizer

[high snow]

That looks really confusing. How the hell do I do that?

[frozen dagger]

baby steps. select the clip on the timeline. go to the EQ, play with the presets FIRST and listen to the changes.

[frozen dagger]

like-wise you can change the pitch with "transpose" in the "Clip" details screen as well as speed and the volume try and tweak it more.

[high snow]

What is transpose?

[frozen dagger]

play your audio in the clip while you go through the presets and you can hear the changes in real time

[high snow]

Yeah, I started messing around now. I start to see that.

[frozen dagger]

changes pitch. basically.

[high snow]

Ty

[quiet solar]

The tricky part is that people use them in the expectation there'll be no reprisal from the owner / copyright holder of the original sound. Which if you're an individual person uploading to YouTube is fairly reasonable, and at most it might block monetisation in favour of the copyright holder.

But as OBLMOUS said, it's a different ballgame when you upload that sound to Suno, because suddenly they're publishing your songs which include it without any permission from the owner and without the protections offered by something like ContentID. And Suno could potentially end up in legal hot water for that - hence their rule of not uploading anything that you don't personally own.

So you can definitely import an external sound and work with it at your own risk in a DAW or audio editor of your own, but not in Suno Studio. The Suno answer, which is a lot better since the release of Sounds, would be to attempt to create your own effect and insert it into the song.

[high snow]

Okay. Ty.

[frozen dagger]

I made a Goku yell with Suno sounds. I'm sold. 😄

[high snow]

Can you send me that link 🤣

I wanna hear

[frozen dagger]

https://suno.com/s/EMHEEgyllPLDSl0H

[high snow]

Bhahaha!

[frozen dagger]

I didn't have to get on my mic and do that myself. Like I said SOLD 🙂

[restive pebble]

This has been a great learning op for me. Truly appreciate it @frozen dagger

lurking never hurt if you pay attention.

[frozen dagger]

that's why I lurk here. I won't list or tag them all and pester them with a notification, and there's more than I know. But all the knowledge in this thread is valuable to me. Even something as small as a new genre I never heard before. Very grateful for all the contributions in here. Happy to be a small part of it when I'm lurking here too.

[restive pebble]

I should be in studio more but I never seem to find a reason. The gens Suno makes are good enough for me. I just clean up the stems in studio.
Suppose that's limiting my creativity and abilities in a DAW so it is on my list.

[high snow]

With the lion rule at the end, I’m gonna see if I can try to cover it to see if it will blend with the other.

Cause if not that means I’m gonna see if I can export it and then go into pro song editor, and do it that way. Because it’s driving me nuts.

[restive pebble]

You can accomplish this. Take your time and allow yourself mistakes as you are learning.

[high snow]

I’m about to slap my head on the wall a few times. This is driving me nuts. And this is my final song too for my album.

[high snow]

Also, I tried covering it and it did not give me what I wanted.

[restive pebble]

I've been there with songs. Sometimes I have to let it sit overnight and return with less frustration or I'll have to patch a lot of walls

[frozen dagger]

As a general rule when you''re adding sound effects (especially now that we can do it within the system) best to add them to a new track as a new layer. And keep your sound effects seperate. They do not translate well with multiple covers. Depending on what it is, but Suno Cover is going to turn those sounds into music embedded in the background. (which can be useful, i found it to be on accident)

I've never seen it asked in here before, and I'm not aware of it's limits, or potential, but we are able to Share Projects in Studio. Says lets us edit our own version. Cray cray.

[high snow]

What the sounds is there a proper way to get a sound that you want? Because me typing in, lion roar I’m getting distorted crappy sound. I know it’s only beta but still.

[frozen dagger]

https://suno.com/s/AzHlOXZffH4tn6cj

loud large male african lion roaring; prolonged roar, sustained lion roar; powerful lion's roar

I would generate a couple and then layer 2 or 3 probably overlapping each other on different tracks'

[high snow]

Oh my gosh, yes, I don’t know if you just made that, but could I use that sound for my song? Would you be OK if I did that? Cause I need that it would finish off my song perfectly. 🙏🏼😭

[frozen dagger]

i pu thte prompt i put for it. but ya it gave taht 1 and a 2 second one that was weaksauce. but be my guest. published and put on my page with remixed allowed. only way i know for you to be able to get it 1 for 1. Again I dont know the limits of "Share Project" in Studio. but yeah by all means. use it or jsut copy and paste the prompt:
loud large male african lion roaring; prolonged roar, sustained lion roar; powerful lion's roar

and pull the Suno slot machine again

[high snow]

Ty for letting me use it.

[high snow]

Could you possibly download that for me/export it and then send me the file in DM’s?

[frozen dagger]

https://giphy.com/gifs/nickelodeon-spongebob-vote-mail-efyEShk2FJ9X2Kpd7V

[rapid yew]

Do you know how to create a playlist on suno

[rapid yew]

Where is that

[half dragon]

In your LIbrary page > Playlist tab

[rapid yew]

Thanks so much

[frozen dagger]

Right click any song and go to "add to playlist" is another way. Find playlists in your library. Playlist are not Public unless you make them.

[rapid yew]

Cool

[high snow]

I’m taking a break from it since I just ate dinner. I’ll get back to you in an hour or so. And I’ll send you the song link once I finish.

[left hornet]

Hello, I’m trying to save a song I generated because only the final part is messed up,the vocal suddenly gets too shouty / over-pitched and the audio sounds harsh/distorted.
What’s the best way to fix this while keeping the earlier sections intact?

[frozen dagger]

if you don't want the ending at all, most straight forward thing is to extend from where you want it to be new generation.

[high snow]

What kind of descriptive words in sounds what I need to put in order to get like a shofar/trumpet sound?

[wintry apex]

I have a big question about Extend!

What does "Keep" Mean ? Does keep mean "Discard" ?

[frozen dagger]

Shofar blast ?? I dunno. lol. that how google describes it. Shofar Horn.

[wintry apex]

because every time i choose what to keep - thats the part that doesn't show up in the generation. Where do i combine it ?

[frozen dagger]

When you google it you see there are four sounds. I do not know about the horn, but you are going to want to describe your sound in plain language but i'd probably refer to those sounds/tones that they already have terms for. Depends on how you want the horn to sound.

[wintry apex]

So extend is just cover -- but covering a smaller crop ?

Why doesnt suno just take a quick visit to Udio and look at how they do extend. This isnt extend

[frozen dagger]

In short terms, extend is supposed to look at what comes before it and create new audio based on that. (I think it looks at all audio tho, before and ahead of your extend mark, so sometimes cropping your audio first is a good idea, to be sure to remove unwanted audio influence.)

[wintry apex]

it should be a standard that extend means you keep the audio when it says 'keep' lol .. also back and forward extends.. it shouldnt just be an udio thing.

in udio -- it does that too, but you use the context window to control what the model "hears". but you keep the previous audio, even if you only let it see 3 seconds before.

its just a basic necessity

[frozen dagger]

I have been having issues with the extend feature off and on for a while now, like several weeks. It's still probably one of the better options for Pro or Free plans. With Premier and Studio tho there are better methods of "extending" my audio there. (imo)

[wintry apex]

ah maybe

[frozen dagger]

where are you using the extend feature from? i.e. within the editor, create tab, studio?

[wintry apex]

i wish i could use the studio once to see what its like before committing

create tab

it shows Keep <-- | --> Recreate |

But "keep" means "Delete"

the opposite of what keep means lol

so pretty much just recreate everything

[frozen dagger]

ook i thjink i know what may be happening

when you do an extend that extend is labeled as [part 2] (if it is the first extend from a track)

you want to right click the extension that you liek and want addeed to the part you wanted to keep and sleect "get full song"

[wintry apex]

oh.. Ive never seen a "part 2" label so far

[wintry apex]

Woah just like suno v2

I didnt know that was still a thing

[frozen dagger]

if you extend from an extension the label coputns up [part 3] 4 etc.

[wintry apex]

youre talking about for v5?

ill try

[frozen dagger]

extend works universally the same across all models as far as i am aware

just did d atest, it now says [extend 1]

the syntax changed but yes till has a label\icon.

[wintry apex]

Important question

When I extend, should I erase the lyrics in the keep section and only include what hasn't been sung yet

and only include what is in the recreate right-hand half

[frozen dagger]

when you extend, the lyrics box should contain ONLY what you want in that extension.

[wintry apex]

ok

thank you

[frozen dagger]

sounds like youre new to the experience here, so be wary it does have a tendency to repeaet stuff from what it extended from. adjusting your extend point can help. (if you click and hold while dragging your pink keep bar it will zoom in a little for more precise targeting, or type in the timestamp.) try to extend form within music, not within middle of singing.

[frozen dagger]

GL HF

[wintry apex]

I see it now, Get full Song

thanks. I dont know how i missed it before

[frozen dagger]

you can extend the [extend 1] to create [extend 2] and so on, and then each one has the "get full song" and it will stitch the song together back along its path u made it on.
Exercise good song labeling, or something, easy to get lost in those.

[left hornet]

Tried this already,cut the track, reupload, then extend.
But the new part doesn’t match the melody or vibe at all. It suddenly goes rock while the song is a slow ballad 😅
I was only trying to redo the bridge → final chorus → outro.
Anyone found a good way to keep the key and mood consistent?

[frozen dagger]

which tier plan are you on? fre pro premier

[left hornet]

Pro plan

[frozen dagger]

the part that is bad, all unusable? would separating the stems and using the instrument part of it to guide a Cover be an option?

if not, maybe same method but use the part that is good to duplicate and recreate/cover with missing "extend" lyrics in the cover.

ideally just trying to get more of what you have, or what you want, in audio form and placeholder it at the end and then transform it with Cover/Recreate

[hollow slate]

I'm hearing weird artifacts in studio haven't done any thing yet

[left hornet]

Music’s okay, feels like the drums are the ones pushing the vocals too hard.
Got your point though — I’ll give it a try

[wintry apex]

i really wish there was something like this

a keep and a context window. control what the model sees and control how much it keeps

might not want it to know about earlier parts

[frozen dagger]

In Studio the descriptions give the impression that it works in a similar way you're describing, with us having to control the audio on the timeline.
Not streamlined, yet. Recreate (for me) was giving errors with Lyrics for a couple days, but yeah. The premise is there.

[frozen dagger]

New "Samples" is sorta like micro-Cover, but not sure if that is useful for stitching audio together unless you have Studio or a DAW. But similarly could select the part of the track that was working and create generations to stitch together manually.
(Beta)

[crude narwhal]

Does Studio fix this though, been a genuine curiosity of mine lately because I face the same thing, I'd like to replace certain sections of songs but then it just bastardises it into something inconsistent

[frozen dagger]

it can help with work arounds, but the more control we try to get the more difficult it becomes.
for generating audio with a same vibe or layering/transforming new styles it's pretty decent. There are lots of things that all kind of do the same thing: Inspo, Mashup, Sample, Cover.
It's better for creative exploration, than for consistency, but in my limited experience with Samples, it works enough like Cover to be able to use it.

[crude narwhal]

Aint that the truth, it seems the closer I get to the right sound the further I get away from vocals I want via Cover, it kills me. I've generated great songs but the vocalist has to stay consistent throughout all my tracks as part of what I release.

[frozen dagger]

I posted what I played with yesterday, vocals are all over the place (as is the song), I was just staring at my screen trying to think of best way to get consistency and without going outside Suno.
We can't combine Persona with Add Vocal. Which seems like what I wanted to do after I stripped my stems, or to be able to Cover the vocal... just not a straight forward way to do it at the moment.
I think more vocal control is likely a top priority. Consistency and multi-vocal (duet) easier controls.

[crude narwhal]

Oh I'd love that, control over multi-vocal would be amazing, I tried it with one song and whilst it sounds okay on the surface, if you dig a little deeper and you use Suno you can notice it's inconsistencies, live and learn I guess

I stripped duet's out from my newest release I'm dropping in April because I found it really difficult but I'd really love a womens vocals to soften some of the tracks as well 🙁

My biggest gripe at the moment is the static noise, I'm hoping and praying there will be some sort of update in the coming months so I can regen that sorts it out because it's horrible

[high snow]

https://suno.com/s/p4KEYMu9XqjJqwhU

new rap song

[half dragon]

Great song but for more audience post it in #🐣┃suno-showcase

[quiet solar]

There are a few workarounds for managing multiple voices - e.g. #📚┃suno-school message

You just have to accept up front that it’s going to take some degree of editing time and working with vocal stems/samples if you require precise control over who sings what.

[solar nexus]

Is there some way to set the "extract stems" to default to a particular file type before spending the 50 credits? I keep getting stuck with mp3s I don't want while not being able to change anything from a dropdown box.

[stiff sentinel]

pull them all into studio and use multitrack download, it'll give you a .zip file of .wavs

[quiet solar]

Stem splitting is separate from downloads. The 50 credit split breaks them apart, and you should have download options for them collectively or individually as WAV/MP3/MIDI in the Library, depending on the plan you have.

[solar nexus]

It gave me a zip file of one track.

[stiff sentinel]

this is what it gave me o.o

[solar nexus]

I wish.

[stiff sentinel]

you're doing this?

[solar nexus]

That's one of the many things I have attempted so far, yes, lol

[stiff sentinel]

huh. well it did exactly what i showed you for me

i guess you can download them individually? bit of a pain but that's the guaranteed way

[solar nexus]

A fluke, perhaps, I'll give it another shot.

[quiet solar]

It’s always worked the same as tsundere for me as well.

[stiff sentinel]

you pulled all the tracks in right?

[half dragon]

After extraction, you'll have a choice to download them all either in mp3 or wave I believe.

[solar nexus]

Might be working this time, seems to be taking a heck of a lot longer to download...so probably a good sign.

That did it, thanks 🙂

Follow up question, when you "get midi", where does that go?

[stiff sentinel]

in the aether until its ready, then a download window will pop up

kinda seems broken because nothing happens, but then eventually it pops up