📚┃suno-school

[polar spade]

The credit counter shows "0"? You would have to wait until today's refresh before getting a daily 50 credits.

Did you actually lose your credits innediately?

Free plans only get 50 credits per day.

[weary ruin]

I havew the pro plan and I bought more crdits

It stills says I am out of crdits and I need to upgrade

The counter showed 65 credits, I know I had more and I bought more and it is still saying I have no credits and need to upgrade

Is there a customer service number I can call?

[polar spade]

Did you also buy extra credits?

[weary ruin]

yes

[polar spade]

Oh. Thats odd then, if you only upgraded and never canceled first. You should always have your extra credits for any tier users.

[polar spade]

There are no telephone numbers for Suno. Correspondence is by email only.

[polar spade]

https://discord.com/channels/1069381916492562582/1373016771103166575. Help for support / billing issues.

[weary ruin]

The credits i purchashed show up but when I extend the song or try to create anything, it says I have no credits and need to upgrade

[polar spade]

Go over to https://discord.com/channels/1069381916492562582/1372994553233276970 That channel is to discuss and review your issue. You would still likely need to explain that your account it seems is not functioning...... since you have showing your extra credits but can't use them, shortly after you also upgraded to Pro.

[weary ruin]

Thank you

[sage urchin]

Hey

I'm looking for help, I want the AI to sing my voice but it doesn't sound close to me at all

Or like, how can I use the exact same voice I used before without it being changed?

I'll try this

Wait, what if the song is like, already there, can I just edit it and add my vocals?

Yeah, it still sounds nothing like me

Can I just record my voice and add AI to it or even add my voice to a song somehow-?

[half dragon]

Unfortunately it doesn't work like that. You can influence AI to generate as close as possible but it would never be exact.

You would have to extract the instrumental stems, then recording your voice singing it then layer on top of it offline in your DAW.

[sage urchin]

What's daw?

[half dragon]

Digital Audio Workstation such as Audacity... and others...

[sage urchin]

Ohh

Okay so I extract the instrumentals then sing over it?

[half dragon]

Yes, if you want your exact voice on the song.

[sage urchin]

I did that though 😭

How did you do that

Okay, how can I make it sound very close?

[half dragon]

Play with the audio influence slider until you got one that sounds closest.

[sage urchin]

So what audio do I put in? Cause there's a very specific song I want and I just want the song to sound the same just in my voice

Is that possible?

No, I know this

How do I say this is a simple way

I want the same song, exact same song, same vocals but in just my voice

It's already made

Is there a way to edit it or make a new one that sounds the exact same?

[half dragon]

Exact vocal? not quite
Exact melody? use Cover

[sage urchin]

But it's AI?

[sage urchin]

I'll try this

The lyrics are my lyrics

I just want the melody

No no

AI's melody

Suno xD

No, the lyrics are my own lyrics which I made

But the melody is from Suno

I just pasted the lyrics in

I have pro

Does that make a difference?

[half dragon]

If you're looking for a way to edit your already generated Suno song to replace or sound like your own voice - that's not possible on Suno - you can do it offline.
If you're looking for a way to tell Suno to generate a song with a vocal sounding close to your voice - upload an audio of your voice (singing) then generate a Cover based on that audio, type in your lyric and hit Remix

[sage urchin]

It was generated

And not published

The melody anyway xD

[sage urchin]

Trying now

Also, VR, nope

[half dragon]

Nope as in, not sounding like your voice? or nope as in not working?
If the former, then you would have to accept that AI would never sound exactly like you.

[sage urchin]

Not sounding like me

I haven't made any

I can take the instrumental and just cover it?

What I'm asking is, can I just use the instrumental?

And upload it if I want to?

But I'm not in the US

And you basically just made it sound like I could

The only thing that WASN'T mine was the melody which I'm not planning to use since you just told me 🙄

[brazen trail]

For me, it's still generating the same garbage songs.

[jovial widget]

How come no one's talking about this? Is everyone on the suno payroll

It's awful the drop off in quality is insane

[half dragon]

Maybe because not everyone has the same issue as you do? If anything, there's a feedback form you could submit to Suno.

[jovial widget]

are you kidding go on the suno facebook page people are leaving becasue of it

[half dragon]

Glad they found a better place.

[jovial widget]

you must be able to tell the difference in sound quality from when 4.5 came out and now. Its night and day.

[half dragon]

I just simply stated that not everyone has the same issue. Not the issue isn't there.

[jovial widget]

I'd agree with that, but there something not right with the library playback

[jovial widget]

are you saying your sound quality is as good as when 4.5 came out? Thats interesting.

[half dragon]

Still sounds good to my ears.... online and offline.
Some gens are bad but I don't use those.

[jovial widget]

you could hear every instrument now its altogether and dull

[half dragon]

Are you doing metal? I hear a lot of reports about metal is being dull...

[jovial widget]

I am running it through an amp and really good speakers so maybe thats why I can tell

[jovial widget]

no

its strange 3.5 sounded amazing then went rubbish 4.0 sounded amazing then went rubbish and now same again with 4.5. Could it be that they just want me to use all my credits so it take more spins to get a decent song?

[half dragon]

Again, this would be best to give feedback to Suno - #1379539060363759698 message

[hollow hull]

Hi, I don't know why but I paid my pro plan in Suno, may 31, I can no longer share my songs, and I'm being prompted to upgrade — even though my current billing period shouldn't have expired yet, I didn’t notice the issue until I had already created 6 or 7 songs, and when I tried to share one, I realized all of them were marked as v3.5, which shouldn't be the case under the Pro plan.

I didn’t notice the issue until I had already created 6 or 7 songs, and when I tried to share one, I realized all of them were marked as v3.5, which shouldn't be the case under the Pro plan. I updated my plan, from 3.5, I paid the pro plan and I was 4.5, I don't know why it changed and is asking me to upgrade it again if I already paid for the month, the month end June 30, why is this happening I already wrote to support suno, and billing, and they haven't solve this problem

I still have credits I don't know why is happening this

[vale owl]

I'm currently unable to log in via Google. When I try, I get the following error message:

"[your.email....] is not allowed to access this application"

However, I’m still logged in on the mobile app and can use it without any problems. Does anyone know this problem and know a solution?

[sleek stump]

Update your google email and password, that normally fixes that problem.

[shadow plaza]

anyone tried to do World Music ? im trying to do mine but im not sure if i structure it correctly . any advice about [ metatags] and structure is welcome

[polar spade]

I love the variety of styles found in World Music but haven't directly tried to prompt for it.

[shadow plaza]

The thing is i do have ready promts for the style , I needed help with the structure and maybe some tricks to make the track longer as turn out around 3 min 😄

[polar spade]

Anyone having issues uploading files with Android, here's a quick video (I did just now) to convert to wav file on order to upload. Otherwise, desktops should work OK.

[cinder gorge]

Did anyone else ever managed to get a "censor-peep" into their generation? Cuz' I just have randomly, and I didnt know it was a thing till' now!

[hollow hare]

I had it work with [bleep] before
https://suno.com/song/abe4d02e-f8a4-488e-b36b-d03988631a57 (happens at 00:28)

The other reliable sound effect I know is [laughing] (or variants of hahahaha)

This might help you, from the old Bark v0 documentation. Most probably won't work but a few are still retained https://suno-ai.notion.site/Using-Metatags-34944efe09ec41d693e314f13af44695

[hollow hare]

Ah

[fierce plaza]

Hi, I have a problem.I’ve been using V4.5 with “male vocals” prompt, but every output results in a female voice. Prompt included ‘male, heartfelt, natural’. How Could I do?

[patent sand]

New editor:
"Add more to your track by clicking [icon] on the right" for extensions
There is no button?

nvm I was stuck in lyric mode

[patent sand]

try "masculine"

[fierce plaza]

Thank you!

[main cairn]

So I used to be able to do this but can't anymore: I have an eight bar loop that I want to import and try to get Suno to use as a melodic hook or chorus idea, sort of make it build off my loop. Anyone who sees this, have you tried this before, and does it work better with cover clip, or extend clip since the new audio updates? Spending a lot of credits just experimenting.

[vivid wasp]

4.5 update has had some really fun new stuff to play around with, has anybody heard any hints of what they're working on towards the next one?

[opaque portal]

Since those new sliders (Weirdness and and Style influence) my song quality is very bad. They just don't sound as good as before. I played around with the sliders and burned thousands of credits already. Has anyone got any tips to get the same results like a couple of days/weeks ago ?

[unique stone]

Leave them at their default settings (there is a reset button to do this.) The only difference now is that you now see and can change these settings - they used to be hidden. Generally it's a good idea to use the defaults and then make small adjustments if you aren't getting the results you want.

[opaque portal]

I tried leaving them on default but all my results sound very different compared how they used to before those sliders even with the exact same prompts. It all just feels broken to me 😦 I have different music styles and they are all messed up since those sliders appeared 😦

[brazen trail]

Ever since the "creative sliders" were introduced on June 3rd (around 7:30 PM USA time), the quality of song generation has significantly deteriorated. Not only has the output become less impressive, but several previously stable features have also been negatively impacted. This update hasn’t enhanced creativity — it’s actually limiting what made Suno AI innovative and inspiring in the first place.

[glacial cobalt]

I thought it was only me that was being affected by this and I was doubting myself, but I definitely agree. Song creation has taken a dive since that feature was introduced. Even turning the Weirdness to 0 and style to 100 still produces inferior results as to prior to their introduction.

[tawdry goblet]

having the same issue, the editor is absolute garbage. i'm trying to replace the wording in one line of my lyrics, and have spent over 100 credits and its just feeding me garbage, glitchy mess. there should be at least a preview or something to let you discard it, coz theyre just stealing our money at thiis point

[pallid geyser]

Can anyone tell me why on Suno the number of listens to my song doesn't change when someone listens to it?

[polar spade]

But Cino mentioned, just after that, that they did not know how or where to use the Legacy UI. So that would be difficult to edit many things without use of both UIs at the moment. And many may find the Legacy UI to be simply easier to use to get most things done -- so being familiar with it would be helpful. Having a section with extensively more video tutorials, including work arounds would be helpful , I think.

[echo prawn]

What's wrong with this?

[hard beacon]

what did you upload?

[echo prawn]

I uploaded an existing song

I want to cover it with a female voice.

[hard beacon]

an exisisting song that was created by you?

[echo prawn]

No

[hard beacon]

so.. copyright reason

you are not allowed to upload songs by other artists

[echo prawn]

It was previously allowed

[hard beacon]

definitly not... maybe in that case the system doesn't recognized the track

[echo prawn]

I think if they insist on classifying this as a copyright issue, they might as well just cancel the local upload.

In the past, they detected copyrighted songs but just couldn't make them public.

[hard beacon]

well.. the idea behind the local upload is not using existing songs.. it's more about other things 😉

[bright egret]

Well for me Local Upload not at all working. It is showing Uploading clip. And after uploading clip completed it is still showing the same.
And upload Clip is not taking my Audio for creation for the song and extends the wordings of the lyrics sometimes.

[fiery basin]

Hey

I’m running into a serious issue: all of my custom personas keep generating very similar or nearly identical instrumentals, regardless of how much I vary the prompts. I’m not using the Remix feature, and this issue seems to persist across different genres and descriptions.

At the moment, I can only get fresh-sounding results by using NO persona at all, which kind of defeats the purpose of building unique artist identities. This is limiting my ability to use Suno creatively right now. (Using 4.5 and 4.0 Model).

Would love to know if there’s a fix or reset method for this, or if it’s a known issue.

Thanks in advance!

[echo prawn]

So how do you explain that they allowed non-original songs to be uploaded before, but they just couldn't make them public?

[unique stone]

Upload is for if you have sung something, played an instrument, recorded a sound, written some music, anything where you have created something yourself and want to use Suno to build upon it.

It's not for uploading other people's content and has always been against Terms of Service to do that. This is now being enforced more strictly.

[hard beacon]

@echo prawn : Maybe there was a court order? Or maybe people have somehow tricked the system? There could be a thousand reasons.

[echo prawn]

Well, for me, if I can't upload from my local computer, then I think this feature is completely useless.

[hard beacon]

for you as a non pro musician 🙂

a lot of people are recording their own instruments and upload them to create an unique track

that's what the upload is for 🙂

[echo prawn]

Well, if they have such a big legal problem, then they should open source it and allow some users to deploy Suno locally, and of course they still have to pay the money.

[hard beacon]

copyright is copyright.. This cannot be avoided with open source either.

[echo prawn]

Anyway, for someone like me who is not a professional and does not want to make a profit, this policy is very bad.

[hard beacon]

That is international intellectual property law.

[echo prawn]

If it is deployed locally, it is the user's personal behavior. If the music copyright owner really thinks there is infringement and wants to find trouble, it will be the user.

[echo prawn]

No, just the US

[hard beacon]

EU as well

I didn't make that up 🙂 You just can't change intellectual property... And it's logical that every company protects itself against this.

For my part, I am also ending this “discussion” now. After all, it's the law, regardless of your personal opinion 🙂

[echo prawn]

LOL If it is true as you said, then all the cover songs on YouTube will be removed

But the fact is that they don't

[hard beacon]

I am in the music industry since 2006.. i now what i am talking about 🙂 Youtube is working with an content ID System... any money collected goes to the original artists. And any rights holder could have the song deleted.

But hey... It's nice when extra money comes in. Nevertheless, it's actually not allowed according to YouTube guidelines. But that's what the Content ID system is for 🙂

By the way, it would be nothing new if creators of remixes that were not authorized were sued 😉

[echo prawn]

That's it, but at least the video can still be released if the copyright holder does not change their policy

Maybe I never planned to release the song publicly, I just did it for fun.

[polar spade]

Sliders
I must say, at the default settings , repeating the same prompts as an older song made before sliders, the results are not similar in scope of the overall sound and vibe. It just seems the "default" was adjusted before sliders came out or simply as part of the process of presenting the sliders..

Granted, the default levels for sound were likely changed and adjusted on a regular basis, but at minimum at least per every jump in version. So its of no surprise we can't always keep a consistent pace through the duration of versions. Personas help, but its cramming a square peg in a round hole and so we get close, but we lose bits here and there.

So, yes... its to my advantage to have the sliders so im not just slave to whatever the dev thinks is a sweet spot, but with a caveate...... I will have to keep good notes once I hit that sweet spot if im trying to emulate an older or previous vibe or sound.

[hard beacon]

It is simply not allowed to change original works without permission. And it's understandable that Suno (and other major providers) are protecting themselves against this. I would definitely like to remix songs for myself, but it's the way it is 🙂

[pallid geyser]

here @quick wedge

ok

[quick wedge]

I told her if she needed help she could ask

[polar spade]

Oh. Nevermind.... I get it... sorry. 😄

[polar spade]

I didn't follow the logic. Need that coffee first. My bad.

[quick wedge]

Try refreshing the page. Sometimes Suno goes through visual bugs and stuff clearing cache should help too

[echo prawn]

But they allowed it before, I'm sure. Suno used to detect existing non-user-created songs and just didn't allow them to be posted publicly on their website. But they could download and save them privately. It didn't say it was absolutely forbidden, but now I can't even upload it.

In my opinion, I don't think it should be banned absolutely. Or they should just make this site for professional music creators.

[quick wedge]

It’s okay here’s your cup though ☕️

[patent sand]

Litigation may have changed it, or prompted Suno to change it at least

[polar spade]

Mmmmm.... coffeeee...slurp slurp slurp. Nom nom nom.

Ok. Much better.

[patent sand]

Not worth the potential lawsuit

[polar spade]

I'd hope to maybe get Rick Beato on our side. Convince him there's a place in the world for a Sunonian to coincide.

[pallid geyser]

which cache do you mean?

[echo prawn]

Well, I think if they change this policy, I think it will discourage many non-professional music creators. Because we can't compose music. We don't know anything about music theory, and we do these things just for our own pleasure. As for profit, we don't want to do it at all.

[quick wedge]

The one from your browser there’s a section in settings where you can clear it.

[patent sand]

I don't make professional music on Suno and I can mass-produce decent stuff, regularly produce high-quality stuff. wyd

[hard beacon]

who is "we" ???? Maybe 70% procent or more are musicans and use suno for a creative process?? who knows?! 😉

[patent sand]

Most users just click the AI lyric button, AI style button, and Create

[hard beacon]

ok.. 70% is a little bit to high 😄

yapp.. but those who try to create unique stuff are inside the theory of music

[echo prawn]

At least not everyone is. My friends who use this software are all amateurs.

[patent sand]

Anyways my original point was that they don't want people uploading Taylor Swift music and trying to cause issues with her legal team, for example

It's just a blanket ban. I've seen some dumb stuff redacted to be fair

[pallid geyser]

ok

[echo prawn]

That's it, I don't think they should have such a one-size-fits-all policy

[polar spade]

In a way, im a bit happy using Suno, that I can catch up to my peers who are just stunning songwriters. Im a good songwriter, but less a musician. Either way, I always have to teach myself to play a F#7/11 chord or whatever i apparently composed, but Suno now is a timesaver for making demos. So thats my angle of creativity. I guess it doesnt affect me for whether someone likes my song or not.

[patent sand]

Ironically, half my 4.5 metal songs start with the Napalm Records intro

Yes I finally subscribed

[unique stone]

Something not being detected previously does not mean it was "allowed". It has always been against the Terms of Service to upload audio that you don't own the rights to.

If someone upgrades their security cameras to a higher quality, nobody can complain "but you didn't catch me before!" when they were doing something they shouldn't have been doing in the first place.

[hard beacon]

This whole copyright discussion is superfluous. It's just the way it is - whether you like it or not 🙂

[polar spade]

No one detected my hooking up to my neighbors power source, so thats allowed now then right? 🤣
No one detected my embezzling money from my company the past 2 years, so its all good right? We can keep doing it?

[echo prawn]

Forget it. I don't think any further debate will lead to anything.

There is no need to argue with diodes

In my opinion, the law is the law and people are people. The law is dead and people are alive.

[solemn zinc]

hello team, do you known how to make a one minute music ? 'cause i have to edit one by one for having it ^^"

[pallid geyser]

it works ty ^^

[polar spade]

To your point, the law is always a guide and not set in stone, in the real world, so if enough people feel differently sometimes those guides are moved to better suit current popular opinion. So you're right in that perspective.

[dusky salmon]

Does anyone confront the problem cannot download from Suno?

[polar spade]

In my best candor, my two attempts are to keep people from getting their accounts suspended and to keep things out of my outputs that can be traced to some previous public domain or copyrighted material. Stopping that from the front end is much easier and less costly than what it costs to research my creations for any unintended infringements.

[dusky salmon]

It seems the Desktop version has an issue. On the app, it is still smooth.

[polar spade]

Yea.? Ill keep that in mind.

[dusky salmon]

I cannot even play tracks on the Suno desktop...

[polar spade]

Desktops are fine for uploading but Android Suno , still having issue uploading mp3s and WAV files.

Do you use browsers on your phone or just the app?

[polar spade]

Oh wow. Thats not good.

[dusky salmon]

Just App.

[dusky salmon]

I know. It is weird

[patent sand]

You keep specifying a desktop app...? Are you using an android emulator or something?

[dusky salmon]

I am using the browser on the desktop and Android app on mobile. Now, i have an issue when using Suno through the browser on my desktop.

[echo prawn]

Perhaps some people don't know that DMCA is not so rigid. It allows fair use of existing copyrighted music in certain circumstances.

I think they should implement a local-only policy for such content, and that users should be responsible for any copyright disputes that arise. Of course, if some trolls continue to accuse from a moral high ground, I'm too lazy to keep up with them.

🤣

[neat nexus]

This is not about copyright law. Suno's Terms of Service are the contract by which you are allowed to access the service provided. You are free to not use the service if you dislike the terms, but the contract defines what Suno will allow with its service.

It is Suno's choice to not allow use of copyrighted material or public domain material in their service. Full stop. They have the right to enforce the terms of this contract with their users and users have the choice to find some other service. But as long as you wish to use Suno, you must not upload anything you do not have exclusive rights to.

[echo prawn]

But I think they could use a more balanced policy, rather than a one-size-fits-all approach.

[neat nexus]

They have chosen not to do so. Your choice is to abide by the terms of their contract.

[echo prawn]

Suno itself is a tool-based platform. If it is too strict, it will deviate from the original intention of "democratizing music production"

[neat nexus]

Trust me, as someone who loved to take lyrics to old public domain songs from the 1920s and earlier, I am not thrilled with this situation either, but those are the terms I agreed to when using the service.

It's especially annoying with traditional songs, hymns and holiday songs or patriotic songs...

[echo prawn]

But I am not satisfied with this policy, and it is my right to express my opinions.

[neat nexus]

If that's what you're doing the correct place is probably in a suggestions section. Try #💡┃suggestions-web

This is more for discussing how to improve our songs and techniques to do so.

[hard beacon]

That's your right, of course - but even after 100 non-legally valid arguments, it doesn't change anything 🙂

In that time you could have already created a cool song of your own, with your own vocals 🙂

[neat nexus]

And ReMi with some careful wording can pull off so much wonderful lyrics with and without editing them after the fact. I don't feel deprived.

[echo prawn]

My current need is to cover an existing Polish protest song in Japanese. I just want to use this content as a learning and communication. Of course, if you still insist on your point of view, then I respect you, and there is no need for us to continue discussing. After all, everyone's needs are different.

[hard beacon]

I understand what you're saying... I'd find a remix function cool for me personally too, but it doesn't change anything 🙂

I am only basing my statements on facts, not on my personal opinion

[polar spade]

@solemn zinc Im not looking at my prompts, but frankly just using appropriate prompts, such as "jingle, sound bite, memo, bumper, jingle bumper.... use the lyric box for some small direction tags, and when done, any over 1 minute, just fine the sweet spots to spice out enough to have the right length and a good stinger at the end. Remaster and cover for a variety of flavors.

Using common language in the prompt box may work as well, such as " a 45 second to 90 second jingle style instrumental, immediate start with no longer intro, fusion jazz vibe with a button stinger ending. "

I ran some and got lots of under 8 second stingers. Nice... and maybe i can build an instrumental before them... but ... I did get a couple around 1 minute. But its always better to run over time and splice back to 1 minute. Its almost impossible to, say, extend off of 45 seconds and try to hit 59 seconds exactly.

https://suno.com/s/ke05bOiCufsTxTg3

https://suno.com/s/5fDKcVkxnR2mpInw

[polar spade]

☝️

[graceful condor]

Does this not work anymore? [Instrumental Intro]

[polar spade]

You could try [intro] [instrumental]
Or [intro] [ interlude]

But typically an [introduction] infers its not vocals

Otherwise, if it doesnt work after 10 outputs then yea, we need to change something up.

[hard beacon]

We're going round in circles... I can't say any more about this than I already have

but i know what you are talking about 😉 😉

[wary raven]

Hey everyone! 👋
Does anyone know how to change the lyrics of a song you’ve made with Suno?
I clicked on “Extend,” but I can’t choose a starting point or edit the lyrics myself. It only lets me drag from the back to the front.

Anyone have experience with this or know what I might be doing wrong?
Thanks a ton in advance! 🙏

[polar spade]

Yea. Be careful. May be a few wu maos lurking.

[polar spade]

You're a paid subscriber?.

[wary raven]

Yes

[polar spade]

You can use Cover to make small changes that it might pick up on, depending on my much lyrics you already have.

You can also use Replace instead of extend to change lyrics within the lyrics.

If youre using extend in the Remix panel then its just about typing in the timemark you want. But if using the UI , youre right, extend is ONLY to add something after a point, we can not make a left and right side and choose a segment bracket for extend. That is Replace only, for that.

[wary raven]

Okay, that already helps me a lot—thank you! I’m working on my phone though, and I can’t seem to find the Replace option anywhere.

[polar spade]

In the new UI you simply drag a segment to start a replacement. Click New Lyrics on the left. Edit the highlighted lyrics in the new lyric box. CLICK regenerate and check your generated samples for any that work well for your goal.

You also can click Legscy , top left while in the new UI, and use the older version. There is a Replace tab at top.

[wary raven]

Does this also apply to the mobile version?

[hard beacon]

afaik: no

[polar spade]

Choose the 3-dot menu off of your banner lists of outputs . Then click edit , then Extend [or Cover] to get to the Remix panel. Click the extend tab to make sure its chosen. The extend box will be at top. And make sure all lyrics you need and prompts are in the appropriate boxes. (See video in new post below)

[hard beacon]

i just wonder how people want to create unique tracks via the app 😄

[wary raven]

“Aha, thank you so much for your patience and time—I understand now.”

[tranquil pike]

Hi

I made a accidental purchase the $96 membership by accident. I wanted to let you know. I only want the free one not membership

[patent sand]

Not sure how you "accidentally" made a purchase, but you should go here for help: #support-guide to see how to submit a ticket to billing

[nova hound]

Hi all

I would like to make only short song ( 45sec -1min 15' max )

How to setup please? ( without éditing after )

[solemn zinc]

thanks bro , i gonna try this tonight 😄

[nova hound]

Le goat 🐐

[tranquil pike]

Thank you

[polar spade]

Inadvertently?

[polar spade]

I just posted up a bit to another user so here's my same answer. #📚┃suno-school message

[nova hound]

Yeah thanks mate I saw it 🙏

[polar spade]

Remember, you can click 2 dozen times and 300 credits and try to find on that lands on exactly 1 minute, or find a few nice ones and remove a bit from the middle for free and usually faster. You're choice.

[main cairn]

So I used to be able to do this but can't anymore: I have an eight bar loop that I want to import and try to get Suno to use as a melodic hook or chorus idea, sort of make it build off my loop. Anyone who sees this, have you tried this before, and does it work better with cover clip, or extend clip since the new audio updates? Spending a lot of credits just experimenting.

[polar spade]

The new Edit UI has "duplicate" feature. Cover the initial upload if you wish, to see if it has a nicer sound, but then duplicate it all you want and edit between the duplicates where needed.

[main cairn]

Does any of that work on iOS mobile? I haven't seen duplicate on there.

[polar spade]

That's a good detail to lead with. 😁 the edit UI would be browser accessible only on phones

[main cairn]

Ok, lol sorry I should have said that. Thanks.

[solemn lintel]

I know I've asked this before, but never worked out a reliable strategy... does anyone have suggestions on how to get Suno to stop giving me the same really generic male voice? I almosr wonder if 4.5 is somehow leaning into less interesting vocalists or something...

[polar spade]

If any interesting voice from 3.5 or 4.0. Then turn them into 4.5 personas. Otherwise , depends on your genre whether you like whispery, rough, strained, raspy, growl, grumble, creaky, scream, dark, baritone, soprano, or whatever voices you're prompting for. Be creative. Hallucinations sometimes have better interesting voices you can make personas from as well.

[solemn lintel]

Yeah, I keep trying to prompt it for different voices, but just keep getting the same one over and over.

I've tried many of your keywords, in all versions. ^_^;

[polar spade]

Youve prompted these in version 3.5? Thats the better for creative voices, live performance, sound edfrct often.

And dont forget to borrow someone else's persona, rerun and layer it with more prompts and try to come up with something unique.

[solemn lintel]

Yeah, not even sure where to start in finding other people's personas...

I've done some searches for the sound I'm looking for, but didn't get much.

I'll work more with 3.5, though, if you think it handles the voices better! Good starting point at least.

[polar spade]

What the genre or comparator you want in the voice. Then search that style and find a song that you can at least run the same prompts.

@solemn lintel You can find songs you like, click on the profile and check their persona list to see if they have any publicly available personas.

@sweet leaf.
#archived-ticket-help message
Long song are still mostly for first run generations. Otherwise covers, extensions, and all dependent on what is incorporated into them, but no one factor that stands out to be used to precisely control a cover's length. . My average covers run about 4 minutes.for a 3-5 minute song.

[sweet leaf]

Is there any way I can force him to do 5 minutes?

[solemn lintel]

Interesting. I’m. It sure where to start with it but it sounds like a possible solution!

[late rain]

I need help. 😦

[patent sand]

With?

[late rain]

I looked in the past and found what I needed. I'll try it tomorrow. But I might need a bit more understanding of prompts. I'm trying to make a duet. And Suno just doesn't listen to me.

[patent sand]

Yep I have that exact problem right now too actually. Sometimes it would work, but today it is ignoring all male vocal input and giving me one female vocalist

The ones it worked on had a persona it worked on before

[late rain]

Plus I can't create Personas. But that's a different problem.

[tender eagle]

So to use the credits that I paid for, not monthly credits, I still have to pay another month? Even though I paid for these credits directly. Even though they don't reset. It won't let me use them even on a free account. In fact, I'm not getting any new credits on the free account. It just says I've used all my credits.

[empty coral]

anyone know how i can create a heartbeat effect?

i tried uploading heartbeat 15 seconds, but i can't seem to edit it to reuse as base of song

[empty coral]

Ok, ill try that!

[boreal briar]

does anyone know how to prompt for switching vocals like a duet ive been getting inconsistent results

[noble drift]

I haven't tried it with duets specifically, but I've replaced vocal sections of the song using edit and it worked out well

[boreal briar]

i was afraid of that thank you the advice.

[runic stream]

can I publish at once for many song?

[rigid pasture]

How do I find the songs in Suno Ai trash?

[polar spade]

You can put songs in a Playlist and publish the Playlist, if you want.

[polar spade]

Answered in need-help channel

[livid summit]

have you tried / with slash line this?

[livid summit]

nobody makes multiple tracks at once that are good enough to share. just don't

[solemn zinc]

works good, thanks bro

[zenith depot]

So like, I got this Epic vocal voice form a song, I would like more songs from 'him'. (on Udio)
But the way 'prompting' is today, this is not possible.

Can it be possible to make the singer into an artist so the AI knows to make it based on that AI artist's voice??
https://soundcloud.com/antdx316/83de5e28-97f1-42b2-b709-57fdd8d69643?in=antdx316/sets/my-music&si=ee9e3d665db14e13bfe5115faa39be27&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

[unique stone]

This is pretty much what "persona" is for.
Create a persona from that song, then select that persona when creating a new song.
If that doesn't work, try getting the vocal+instrumental stems from the menu of the original song, then create a persona from the vocal stem and use that one.

[zenith depot]

where

[unique stone]

"..." menu, then Create > Make Persona

Once you've created it, you can use the "+ Persona" button on the Create panel when you go to make a new song

[zenith depot]

I will see, it was from Udio

it will still work?

😟

[unique stone]

Ah. Didn't realise that.
No, it has to be a voice generated through Suno.

Alternatively you could try using "Get Stems" to split into vocal+instrumental and then for the vocal stem, use "..." > Remix/Edit > Extend.

That's the "old" way of working with uploaded audio. The new part should use elements of the upload, so if you separate the vocal part it should ignore everything else when using the voice as a reference.

It'll try to mimic it but, as Suno's models work differently to Udio's, it'll likely sound different to some extent.

[zenith depot]

the persona thing is good

its like a remix though

[zenith depot]

it doesn't work 😔

but the persona 'remix' is good

[frank ferry]

/chirp

[flint haven]

The /chirp command is no longer working

Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com

[polar spade]

Wouldn't, when we make an album for an "artist" or even ourselves, we think that they are good enough? We're always a bit biased to not know whether our songs are actually so-so.

You mean the outputs that are literally one after another? Those mostly are rough drafts, yea.

[bright cargo]

Is anyone else having issues with the app putting in the same 1-2 riffs in almost every single generation? Tried reaching out to support multiplle times, crickets.

[polar spade]

That would not be a Support viable concern. Not even a bug. But just a need to tweak, if it too common. Put that in suggestions channel if you wish.

Honestly, it did that up to 4.0, but less frequently since 4.5 for me. With often some great variations.

Similar items in outputs may indicate prompts are pretty concise and maybe more difficult to have as much variety.

But I DO still notice , if we run outputs back to back, many motifs still run over from one to the other, even after changing prompts or removing the seed audio from the Remix panel. And hearing the same or similar riff sequential outouts could be part of that.

[bright cargo]

I will post it there. Thanks for the feedback. Yeah I'm not exagerating when I say these 2 riffs it puts into at least one of the 2 generations for like 4000 credits 😦

[polar spade]

Not extensions, covers, ect. But out of the box new outputs? Do you click the clear all button before starting a new Create? Not that its some panacea, but it could help zero out any buffers or latent seeds, he says, as if he knows what hes talking about.

I do that as often possible in case I've left anything in the till.

[bright cargo]

I've made multiple different workspaces trying to get rid of it

I don't think ive specifically hit the "clear all"

Even when I "dislike" the generations it seems to do it more lol it's like it's being spiteful

Just did a clear all and redid everything... 1st gen gave me the riff lol

just listened to 2nd gen and it squeezed it in there too bwhahaha

[patent sand]

Does v4 have the weird BPM consistency, or just v4.5? For uploading Stems to DAW

[noble moss]

Good question! I’ve been wondering the same—when uploading stems into a DAW, the BPM shifts I saw might’ve started in 4.5. Not sure if v4 had it too. Anyone test both?

[true night]

small work around atm. make an remaster of it , then it , might give you an persona out of it, but i feel its an 50\50 if it work or not.

---

So a few days ago suno gave me an Description prompt of my upload track\song\sample or recorded well upload - and an somewhat accurate transcript of what was said\sang in the track\song if it had words in it.
now that is gone both the prompt and the transcript, why. anyone got any info on why ?
because i liked that quite alot.

[patent sand]

Turns out the Suno stem splitter is actively removing/ignoring some parts of the audio

So I've given up on that (50 credits/split anyway) and just going to use native FL stem split

[autumn rock]

I am beginning to believe that when creating new tracks that Suno gets "stuck in a thread" of using [one thing]
For example, could be a voice female over and over even if you ask for male voice. The only way out is to abandon that session clear browser etc and start over/fresh.

[coarse gulch]

is there a way to make the mix more consistent throughout? i've noticed that a lot of my songs's mixing starts getting "tinny" and "plastic" like by the end

[polar spade]

I've called it ghosting, since 3.0. But yea, I'll hear a riff of one recent song output show up in another within just a few outputs of each other. Or as you say, I could just put acoustic in style prompts and get a guitar. Change to jazz and still get the same acoustic (countryfied) guitar strumming, which should not happen. I have literally stopped, go out , log back in. I've even thought changing my IP address my stop it. Can't tell. 😅 Saves chugging through 20 outputs of ghosting other same riffs and styles until Suno finally catches up and changes its outputs.

Between every time it happens, it just feels there' some latency from it clearing out some buffer somewhere and making either a new random output from the same prompts, or a new output from changing prompts. That just the actual of doing anything else gives it enough time to make the reset. Whether vocals, instrument, rffs, motifs, etc. Even i line from some random recent output has snuck in before.

So everything I've done has a similar correlation: im doing something to delay the next output. So now I just let every output finish and give me a timemark as my "delay of choice" t stopping the NEXT output from ghosting old stuff. The jury.is still out.

I've always related it to changing gears on a 10-speed bike with a stretched cable. Takes about 5 good long second for it to eventually slide into another gear. In sunos case, seems about 2-3 minutes.

[azure hawk]

I am trying to change few words in the lyrics as originally the pronouciation is not correct. How cani achieve this ?

[patent sand]

Replace Section in the editor

[boreal briar]

try spelling the words phonetically

[patent sand]

you can edit the words and highlight the appropriate section

[azure hawk]

i tried multiple times, and lost 50 credits till now

[patent sand]

maybe change Fixed to Smart at the bottom-left

and use phonetics like Stars said

I lost ~100 credits last night trying to make the thing say "roses" instead of "rose" and had to end up using "rose-is" or something

[gleaming ferry]

Can someone explain me why when remastering an already v4.5 song they turn out to have v3.0 quality vocals

Quality drops are much worse now on v4.5

[polar spade]

Its unexplainable. Other than to say the fidelity changes with each combination. We hope most for qualities we want, but I get the 30% or so that just isnt up to my standards. Muddy sound, or tinny, too soft, hissy, grainy, changes in speed. After a couple of years and using both music and art/vid AI , anything 30% bad and under is par.

[gleaming ferry]

Omg yes

[cedar sentinel]

Hi everyone, is there a way to convert a song with vocals into an instrumental version? The vocals can be replaced by another instrument, but the background music must remain the same as the original. Thank you.

[sleek stump]

Can’t really convert it, but you can get the Stems for it ( the 2 stem option) which will take the vocals out of the instrumental leaving you with a clean instrumental track.

[polar spade]

You can't replace or overlay one distinguishable sound with another as you would tracks of instruments. Generative outputs are chiseled out all at once so that it appears to have instruments and vocals.

As Ginger mentions, we can separate some of those sounds into separate instruments, but for now we can't put them back together into a song again. Maybe that will be something they can look at doing later.

[cedar sentinel]

Thank you. I’ve tried separating the audio into two stems: vocals and instrumental.
However, since the vocals are quite loud, the instrumental stem only has clear bass sounds, and the accompanying instruments are barely audible.
What I want to do is convert the separated vocals into instrumental sounds, then use a DAW to mix these new instrumental sounds back into the original instrumental stem.
The issue is that I don’t know how to make the new instrumental sounds match the duration and tempo of the separated vocals, so that when I combine them in the DAW, they sync up well without needing much editing.
Can you suggest some solutions for me?

[north crane]

Hello, I've been using the 4.5 Pro and was wondering if there's a feature to make the songs longer, over the 5 min mark. I've mostly been getting random results so far getting the longer songs over 5 min mark once and blue moon. Also i notice randomly it will mix lyrics from older song from that persona which is cool sometimes, but also would like way to switch that off. wondering is the length of song a prompt thing? as well as the persona lyrics mixing in from old song is that prompt thing?

[lunar remnant]

yes. there is.

right click

doesn't work too well though

[polar spade]

For first runs, it does seems responsive to prompts and lyrics in that a few lyrics creates a shorter outputs, more lyrics tend to be a medium output, maybe 4 minutes, but longer lyrics then snap back and resist it from going further and a need to extend seems needed.

Putting g in a couple of metatags with no lyrics at all and I've gotten just some random over 7 minute outputs. I've got a 7:52 ambient noise that sound like a distant e-train after putting in just some train effect tags. No lyrics. No instruments .... just train references.

So where's the balance , I can't tell as yet.

[north crane]

ty ill try some of that out, and that's the thing with beta ive noticed, its kind of a playground for testing things out so i understand it. Wanted to get some more general idea of what other people have come up with. I have also noticed lyrics sometimes it works when i enter nots with example (type of vocals i am looking for for those lyrics or sound leading into them) sometiems it works sometimes its reads them as lyrics is there a proper symbol to use for this so it doesnt read as lyrics.

[polar spade]

Keep in mind, at all times, Suno reaches back and uses anything previous as a guide and seed to lay down the next bits. So covers, personas, extensions, all need a bit of that output time to allocate for seeding the next output. So if you have up to 2 minutes of an output and you cover it, it will cut into the resulting time of your next output. If you are adding in a persona made from a 4 minute song, you cut into the time allocated for the next output. [We suggest, crop an output down to 1 minute of the best bits before making into a persona] . If you extend a song, it will reach back up to 2 minutes of your origin output to match and lay down the next extension and so lose some of that time needed to result in a longer extension. A bit of give and take and tweaking is constantly going on.

After awhile, we find its best to just piecemeal each project s bit at a time. Like a sequencer.

[north crane]

So, maybe this sounds weird, but sometimes it adds in the chorus from previous songs I worked on with that persona—specifically the first ones I wrote lyrics for when I was still developing the character. I'm not sure if I'm explaining it clearly, but I can show you with some song examples if that helps. i dont think this is prompt related as much a persona related thing.

so say i worked on 50 songs with the one persona for example it will add the chorus in

[polar spade]

Making the 1 minute personas (and avoiding the chorus as your cropped bits you use go make a persona) does help in having Suno focus on the lyrics and prompts of the new output [and only the styles and musicality of the persona) than pull actual vocals and music from the persona.

Also, if your lyrics of your new Create are a good length, hmmm, say 1000 characters, this helps to "drown out" the persona also, from trying to fill in any extra "space" will old lyrics or hallucinations.

[north crane]

sorry im kinda understanding what ur saying but is there videos that cover this so i dont waste time with maybe stupid questions

[north crane]

then last thing is the lyrics prompt question and how to fix that kinda issue

[polar spade]

Nawww... mate. No stupid questions. Sorry if I might sound repugnant. Hahah... I just repeat things and hope it helps to "click" eventually. Im still learning every day and its been nearly 2 years of this.

[north crane]

not at all mate just learning as well and being new i dont want to come off as stupid haha

[polar spade]

The lyric prompts, those bracketed things in the lyrics are called tags or metatags, if you see those names pop up.

Honestly, the best means of understanding them is to click on many songs you like, or random ones and see all the ways people use bracketed tags to prompt or steer the lyrics to be sung in a way that is close to what we want.

Keeping in mind, suno is always nudged and nuanced, and rarely can be forced to do anything.

[polar spade]

One simple thing to know. Use Brackets to segment your sections, and Suno will do better to sing a verse as a verse, build to a chorus ( or Hook) and change chord progressions on things like bridges and pre-choruses or builds.

[Verse 1]
This session dkdkdn lala
Djdkekks
Ksiejenek
Ask

[Tag]
Thats the way it is.

[Prechorus]
So let's take to the top
As we all shout....

[Chorus]
Way to go
Way to go
This we know
Way to go

[north crane]

thank you very much brother i appreciate all the detail and help you have given and time. To me that what makes it so great right now, is it is in experimental stages and we just learning how to properly communicate word prompts etc.

[polar spade]

It may be some.confusing and vague readings, but there are some decent tidbits in the workflow channels that can give some clues as to what others find work best for them as well. https://discord.com/channels/1069381916492562582/1252267444773326848

[polar spade]

Takes time for sure. We have to have the mentality that using LOTS of credits is just part of the process. Haha.. but in the end , can be rewarding as well. Keep asking question in both channels. Lots of helpful users in here -- at different times of the day too, I might add.

[north crane]

I find that's the cost of being in the early stages of things tho, in the end the credits the cost you put in goes back into the business the tech and helps keep it going and perfect it. so its all a circle in the end and its returned rewarded in the end.

[polar spade]

For what its worth, a final decent publishable output takes me about 800-1000 credits, on average these days. The more skilled we get , the more we start tweaking and being more creative.

But heck, many are quite pleased just throwing in some lyrics, a couple of prompts and bam, 30 credits later, they have a song to play for grandma or their mates.

[north crane]

lol that's part of the fun of it lol. If your not going through 2000 credits in a day u did not have fun lol. I can get in one and just stay glued for hours messing around with different styles. at the end of the day unless it really flopped even duds can be salvaged i dont even want to call em duds maybe lyrics come out scuffed but still instrumental or value in flops IMO.

[north crane]

anyway you have a great one brother and best wishes

[graceful condor]

I guess once an image is set for a video it can't be changed? I've tried and Regenerate Video does nothing. When I Dl the video the image hasn't changed.

[glacial cobalt]

A problem I have lately is Suno not following the prompts to the end particularly with Death Metal songs. They never get to the end and always seem to end half to 3/4 of the way through. Don't even get me started on trying to extend songs as that is clearly broken at the moment with nothing more than 10 seconds of nothing being generated. In fact my latest attempt to extend a song to finish is up gave me 2 options 1 of 2 seconds and the other of 3 seconds. Something is clearly broken since adding in the advanced sliders options as things worked perfectly before that.

[polar spade]

Did you give it 20 minutes? It might need to catch up with the change.

[graceful condor]

No I did not, But I will wait now.

[polar spade]

I know it's a delay.for pretty much any title or artwork change. But honestly, i haven't tested in awhile. They [suno devs] can always make system changes.

[graceful condor]

It changed, I guess I was just too impatient.

[polar spade]

Booya! Great. Glad it worked. And thanks for the update. That helps for the next user.

[proud temple]

for what it's worth, here's how I use the 🪄 button for suno style

[polar spade]

And what consistent results do you get?

[proud temple]

it always give me text. so far.

[polar spade]

The output gives you new lyrics? What is "it" ?

[proud temple]

the 🪄 button expands the style description with the model v4.5. An LLM is used to generate a longer music description based on a short description. But it's an LLM so we can ask it slightly more complex things. (it will complain if you drift away too much. but if you expect a music style description, you'll get it)

[polar spade]

Right. I get that. You click the wand, it give you text..

But by then typing in "anti-music of...." youre saying it does something different? The opposite maybe? what results are you getting compared to if you click Create on the original prompts paragraph without "anti-music of...."

[proud temple]

it creates a variation of the same style description.

[ruby saffron]

the word "stand" and the word "decorative" .... i had to regenerate so many times trying to spell decorative in a way that they would actually say it instead of "de crrrrrrrrative" like... it felt very jarring to here hahah

[neat field]

has anyone done extensive v4.5 Advanced -> Weirdness scale experiments and able and willing to share notes/examples?

i don't wanna waste credits on those experiments if there are other examples ....

[fair mantle]

I have some Issue with my account I cant log in with my account anymore. But everything gots transfered like usual of my bank account so its paid whats wrong what can I do?
Does someone have an Idea how ti fix it? (I need the account immidiateley I have some songs for a bigger project saved on suno wich consinders recordings of movie scenes ect. Each day I cant work on this project costs me and my team arround 8300€. Im Thankful for each helping Idea. The Suno Support acting to slow.

@meager skiff @shadow pier @modern cairn @slender lodge @grizzled prairie Whoever can Help I would be very thankful for a fast solution.

[sleek stump]

None of the Team are working, its the weekend.

Also, you arent going to like hearing this, but your only recourse is to email support, which will take some days for a response

[fair mantle]

Oh great of course... I know it is weekend thats why I checked Suno on Friday. At this day everything worked well.

[sleek stump]

Are you trying to log in through Microsoft or Discord?

[fair mantle]

Microsoft

[sleek stump]

I may have a solution for you

[fair mantle]

I even checked if Google has took away permissions seems not to be

But im Thankful for every Idea

[sleek stump]

Got to your MS account and update your email adderess and password, that has been known to fix this exact issue

This is a somewhat common SSO issue with both MS and Discord

[fair mantle]

Ok sorry I am loged in with Google not with Microsoft. TBut the Email Adress is Ancient and was former a MS Email Adress. How can I fix this on my Google account?

Security Settings?

[sleek stump]

Same thing, go to your Google account and update your email address and also update your password.

Then log out and back into google, then try logging into Suno again

[fair mantle]

Thank you I will try

[neat field]

ok, just 10 credits low sample size into testing, "Weirdness" seems more like "Lyrical structure and vocal Improvisation scale"

best guesses
--higher Weirdness redistributes song lines in ways that align with internal model training better (ie if the model has training for some lyrics that should appear after a Chorus, song structure might move around)

--also there are less emotional adherence (alignment?) for some strange reason at lower Weirdness scale - the singing sounds less emotional even with specific emotional style tag

(i cannot believe Suno doesn't document this better)

this is the only announcement they made about weirdness
Creative Sliders: Control your output before you hit Create. Choose how weird, structured, or reference-driven your generations get—with new sliders that shape the result your way.

[sleek stump]

The entire platform is experimental still, and that sliders are one of the newest features. They have never explained what anything does in any detail, we are left to experiment and learn this stuff for ourselves.

[neat field]

sure sure, one hopes hopes they leave "startup" mode and enter into proper Enterprise mode with better documentation

they are already into 100s of millions of startup capital i think the news said sometime within the last year - startups typically have better experiences just before they become Unicorns

but maybe the funding dried up with all the legal stuff pending

regardless, weridness is an awesome feature (so far)

[sleek stump]

Keep in mind that just because a startup procures a lot of funding, that doesnt mean they have that now. It costs a lot to pay devs, lease office space, pay for administration and HR, and the costs for compute are significant. I am pretty sure they are still operating at a loss.

[neat field]

who knows, music subs are a vast market - i personally think they are occupying a great niche and hope Suno is around in a hundred years

they still beat open source, but Qwen music is in the rear view mirror

[sleek stump]

The legal battles also chew up a lot of cash... and just when it looks like Suno and Udio are coming to a deal with the distributors, a new class action suite is thrust upon them.

[proud temple]

in those cases, it's generally a good idea to break the word into phonemes to force the model to pronounce it as you want.

e.g. Instead of "decorative", you can try something like "deh-kor-uh-tiv" or "dek-er-uh-tiv" or whatever... "dec oh rative"

[neat field]

oooo that reminds me to test more dialects in v4.5 - i wonder if the training data has more South African English dialect somehow....

[kindred trench]

Does anyone know what the best way to tag and prompt to have a chorus repeat for infinity and fade out like an '80s song? using v3.5. I've tried lots of things none of them work.

[neat field]

paste the literal song chorus into the lyrics area like 20X times in an extension, then edit the song in the editor (or another editor) and manually fade out

[kindred trench]

Yeah, I have tried that. it still insists on doing breaks and other weird things most of the time. I can't get the effect I want

[ruby saffron]

"deckerative" worked xD but only once. and it was awkward but at that point i took it because it was the only one that was any different from other version

[hardy thunder]

With the Editor UI you can duplicate (ctrl-D) any section, like chorus section, as many times as you want. Would that suit your purposes for repeating chorus?

[kindred trench]

if it could be seamless. is that what the "heal edits" thing is supposed to do?

[hardy thunder]

I haven’t tried the healing yet as I haven’t had a need for it. Just try and share how it behaved in your case. If it does sound right after healing, maybe you could also try to remaster after the healing?

[kindred trench]

Yes the "heal edits" was the key. I had to heal edits on the first duplication, which made it seamless and realigned the sections. then I was able to duplicate the resulting chorus section and it's seamless. Thanks everyone!
I will fade it out with Logic.

[hardy thunder]

Nice to hear that it worked 👍

[solemn lintel]

This is really weird/frustrating... I finally managed to get a persona with the vocal sound I want; the persona is just based on vocal stems. Yesterday I ran a number of tests using the persona that (mostly) gave me that voice in some songs. But now I've run like 40 attempts (again, using a persona that worked yesterday) and that voice just never shows up. Anyone have any thoughts? Is this just another "try at a different time of day" thing?

[elder dune]

ok .. am I doing something wrong, or is the "Extend" feature completely broken? It used to be, I would pick the time from which I want the new extension to start, type the new stuff into the lyrics prompt, and generate away. Now, the AI is basically completely ignoring my wish to extend (I want to add an outro and some more time to the song) and simply generating different endings, completely ignoring my new outro lyrics, and any instructions, and even ignoring the style - most of the time I'm getting the default "piano" stuff .. (more coming)

I've also tried leaving the whole song lyrics and just changing the end with the new stuff, and it's only marginally better - I'm still not getting any meaningful extensions, just different endings.

is there some new "secret sauce" I'm missing to get this to work right, or is it just broken for the time being?

[solemn lintel]

Huh, I wonder if our issues are related... the song I'm trying to get with a persona is also extending another one. Can you try using 4.0 or 3.5? It won't let me switch.

[elder dune]

when I try using 4.0, it says "error tag is too long" or something like that

[solemn lintel]

Ah, yeah, you need to trim down your prompt.

See if you can just make the prompt less wordy and give it a try.

[elder dune]

oh, probably the "style description" part

[solemn lintel]

Yup.

[elder dune]

I'm trying to start farther back now .. maybe that's what I need to do

[solemn lintel]

I'm trying mine without extending another song... After maybe 50 generations, it sorta used the persona twice.

[elder dune]

ugh .. that didn't help either .. it just generated an appropriate length, but completely made up lyrics that were nothing like I actually entered

[solemn lintel]

Wow, great...

[elder dune]

yeah, this is ridiculous

I'm forced to conclude that extend is just f**d

which sucks, because this song could have been awesome if suno would actually let me finish it

[solemn lintel]

UGH. Well, I removed the extension and now it uses the persona voice just fine.

[solemn lintel]

Same.

I got a really great idea and the sound was amazing... until the wrong singer started singing.

But really, try an earlier version, at least as an experiment.

[elder dune]

trying v4 again

seems to be working a little better

but still not good

[solemn lintel]

Bleah.

Well, without extending, mine is definitely working better, but now it's ignoring a pretty important part of my style. 4.5 still has its problems, it would seem.

[polar spade]

Dont forget to monitor your title to not get past 50 characters or so, as well. Along with prompts it can affect output if there is too much of everything going on.

[elder dune]

I would note that extensions seemed to work find before the most recent editor updates, even with v4.5

[solemn lintel]

Hmm.

[polar spade]

There has been the issue the past month of the first extension returning those 2 to 5 second outputs with many users, for generally no reason, which theyve been looking at. If first extension are at least consistently going passed that time, then it could be some back room fixes going on that may be affecting other areas. Build a dam, flood the valley... its always give and take.

[elder dune]

I get the VERY short outputs when I forget to enter the "selecct time to extend your clip from" field, and it defaults to the very end of the song - it's done that all along since the 3.0 days .. but I used to be able to get good extensions - too good/long most of the time - starting 10-15 sec back from the end .. now I get maybe an extra 30 sec at best AND it completley ignores the lyrics

and most of the time it ignores the style of the song it's extending too

[solemn lintel]

Sigh. And now every song I generate starts with the exact same drum riff. Ugh. Maybe it's time to throw in the towel for today...

[solemn lintel]

I haven't run into that sort of problem myself. Of course, I also haven't done much extending since like v3.

[elder dune]

heh .. yeah

some days, this thing is spot on and gives you solid bangers

and some days it gives you garbage

... and now it's not even pronouncing my lyrics correctly

[solemn lintel]

I think it knows we're talking about it.

[elder dune]

heh

well, I decided to just go with the original version of the song and fade out the ending in my DAW software

[solemn lintel]

I've used up a thousand credits in 2 days trying to get this sound just right. I'm quitting for today. So sad.

[elder dune]

it's acceptable

[solemn lintel]

Better than the frustration?

[elder dune]

ye[p

I low-key (or not) kinda hate the new editor, btw, but I guess that's another soapbox

[polar spade]

@timid gazelle Posting for general channel.

[solemn lintel]

Yeah, I already posted a "complaint" about the interface...

[neat field]

does opening the Editor (Pro) sometimes (just opening it) eat credits ?

[polar spade]

No. If you can show it does, certainly post feedback to Suno.

[neat field]

maybe i am miscalculating, will have a sleep and look again - probably user error there

[polar spade]

Im saying, maybe we click for a few generations and before the credits debit out (30 seconds later maybe) we click on Edit UI. While in Edit UI, we glance and see the credits go down 40 credits (from the 4 generations we did earlier). Just have to watch. A new bug could always be just around the corner.

[vivid ore]

This is an age old trick but if you're having trouble getting the AI to sing what you're writing when extending, paste in the last two lines that were sung BEFORE the point of extension, OR the lines that the AI is currently singing at the point of extension. No tags, no metadata, just context for the AI and it works very consistently for me

[tender eagle]

@glacial scarab

See it is active free account.

[trim egret]

Anyone else having issues with extensions right now? I can’t get anything. Just get my credits refunded

lol I’m sorry…didn’t read back before asking

[vivid ore]

I'm extending quite extensively at the moment, working fine

[trim egret]

I guess I have bad juju

[vivid ore]

I've heard a simple restart of the browser should fix it

[trim egret]

Let’s give it a shot (although it looks like I’m not alone ^^)

[trim egret]

No dice, my aquatic friend

[polar spade]

Reuse should not focus on the melody and lyric arrangement, as Covers do, and for the most part end up creating something different. Though while the output is in place in the panel and the style slider slid far right, its going to pull more nuance, and that might include some melody lines or motifs/riffs from the output than if you click X and remove the song from up top of the Remix panel... In my opinion and observation.

[solemn lintel]

Yeah, "reuse" just reuses the same prompt. In theory, it ignores anything you generated previously. Though in practice...

[polar spade]

I ran another reuse styles on a J-pop funk and dance style song....and had these results for one gen each:

1. Output up top. Sliders to 0% both. Lyrics and prompts no change.
   result: Two nice and different, warm, jazzy pop kind of songs, as expected.
2. Output up top. Sliders : Weird 0% / Style 100% / Lyrics and Prompts same.
   Results: Gen 1: the 5-note riff from output the same and throughout song. All other bits different. more of a rock beat and strange chord progressions. Original is Jpop.
   Gen 2: Started out like Lady Gaga "Bad Romance"(strange)... first chorus different, but second chorus had nuances of original chorus. smoothed into a funk beat, but ended, with "Bad Romance" feel again.
3. No Output at top. Sliders: Weird 0% / Style 100%. Lyrics and prompts same.
   Gen 1: Got a disco song and rhythm guitar feel. No style or nuances from original.
   Gen 2: Actually, first line intro was similar to first lines of original , a James Brown, orchestral hit style. But then the rest was pretty much Back Street Boys.

[polar spade]

The sliders have changed the game. So, no, if you dont use sliders... yes/maybe if you do.

yea. Remix is anything now that isn't a first run creation. So its just a generic label

Fair point. Im having deja vu this was discussed before.

i cant swear that logic plays a part in every decision. haha.

Theres plenty of room to have the Reuse tab placed in the Create Panel. Why not, eh?

Well, just slide all sliders to 0% and leave the output at top or click X and remove the output at top. That should get you closer to a new output from just the lyrics and prompts (or new prompts)

[eternal geyser]

Hello there, im trying to finish a song I made a while ago. I just want to make minor lyrics adjustments, but replacing sections changes the music/flows and i cant get them to fit. And remixes are just worse than the original. Is there a way to do this?

https://suno.com/song/79289dbe-f852-4e43-875c-c92aadb69489

[sleek stump]

Change a few words at a time and only select a very short window that the words fall within. And make sure that there is part of the melody on either side of the replace window. Doing it in small parts forces Suno to maintain your melody and vocal styles.

I generally do a 6-12 second window, and never change more thna 3-4 words at one time.

[eternal geyser]

I was trying with 2-3 words, no luck

[sleek stump]

But what was your window, the whole verse?

[eternal geyser]

I've also tried whole verse, but mostly just the actually changed part. Is it better to keep a bit more?

[sleek stump]

The shortest window is recommended to be 6 seconds, I usually do a 6-12 second window and only change up to 3 words, and make sure the window falls in the middle of a verse. It has been a 100% success for me

Also, what model are you trying to do this with? v4?

[hollow hull]

I put male voice, male vocals, masculie, etc, and in exclude styles I wrote there female voices, female voice, etc... and suno still makes my songs with female voice, why???, (also in the song lyric I wrote there, male voice or masculine)

[sleek stump]

Its 90% a prompting problem. The other 10% is just Suno being the creative generator that is deaf 😛

[polar spade]

OK. Just dropping in to check out the room. The other room, a user thinks im a bot. Not sure how to take that. A compliment I guess. Haha

[north crane]

Lol, I got this result for the first time—it was pretty funny. One track played through normally, but the other made this long, buzzy noise like a guitar amp! I was experimenting with phrases like ‘long’—for example, ‘a long cinematic’ or 'a long slow building cinematic'—and trying to find ways to manipulate a 7-minute orchestral symphony. did get some decent results but weird it came like 4-6 3 min tracks and then popped up 6-7 min ones like 1 or 2 for every 5 3 min song so dont know if that helps others. also making the persona of the exact track that made results i wanted example 7 mins i wonder if that would help

[viral hearth]

I originally signed up with one account that gets significantly more engagement, but I activated my student plan on a second account. Is there a way to transfer the student discount to my original account so I can keep the benefits without losing momentum?

[sudden fog]

I was wondering if anyone has figured out how to make suno accurate, in the voices being sung? I can make great duets but I want to be a little more technical with this new song? I have tried to prompt it in many different ways and tried to structure the lyrics in different ways and its not getting it completely what am I doing wrong

[true night]

in what way or form , are you thinking of, sorry for askin dumb questions.

btw
anybody know why its an hit and miss with the audio transcript on suno? like in the last 100 uploads\recorded in added to suno last 3 days 5 of them gave me an style description such as (an eksample) :

"The song is a slow tempo blues-rock piece in a minor key, featuring a male vocalist with a deep, resonant baritone voice, The instrumentation is sparse, primarily consisting of a clean electric guitar playing arpeggiated chords and a bass guitar providing a walking bass line, The drums are minimal, with a soft kick drum and occasional cymbal swells, The vocal melody is melancholic and expressive, utilizing a blues scale with significant vibrato and occasional falsetto, The chord progression is a simple I-IV-V in a minor key, creating a somber and reflective mood, Production elements include a slight reverb on the vocals and guitar, giving the track an intimate and spacious feel, The song structure is verse-chorus, with instrumental breaks that allow the guitar to carry the melodic theme, The overall dynamic is subdued, emphasizing the vocal performance and the lyrical content."

and an somewhat accurate lyrical transcription of what i sang.\said .- but most of it its dont gimme anything.

5 days ago however it gave it to me every time . then it just went away 4 days ago until this 95\1 hit ratio

was it maybe an new feature they tested and took away. or was it never meant to be there (i miss it dearly) as i freestyle alot , takes eons writing down everything i said in an 8 min volley . so would be nice if its just bugged out for me and there is an fix for it.

thanks in advance for any reply's

[eternal geyser]

Thanks, I'll try that.

[polar spade]

I can't tell. It could be a text version of it hallucinating.

I give the example of my recording me counting to 10 and it giving me a similar long synopsis of my singing a blues song with instruments, chord progressions etc. Completely irrelevant and nonsensical, but funny at least.

@eternal geyser Note that replacing 20 words with 10 words give Suno lots of room to rearrange the meter and tempo of the 10 words. So it easier to replace with a smaller set that you start with, if possible.

Also, it can benefit to use the classic setting w/ custom context window adjustment setting . It combines the amount of beats to the left and right suno will engage to seed and blend in the new part you are changing. Less is more, for the most part, but you might find with one word change, you would need to reach out further to wiggle many of the adjacent words or phrases so that there is room to fit that one word in place. Experiment with the classic blend tool and adding more or less beats to make it fit.

Otherwise the "smart" and "fixed" settings try to determine that automatically.

.

[azure crane]

So frustrated that we can’t use the advanced settings through or mobile phones. That feature is crucial for creating the music I really want to do.

[azure crane]

What is this? V4.5, V4.0 choices gone in browser on IPhone… Really a huge bug…

[half dragon]

Report it in #📱┃bugs-ios

[azure crane]

Yeah I will but this is crucial… Hope this is just a temporary glitch.

[polar spade]

It may be about closing and restarting, and display settings. Its showing on my Android browser at least.

[azure crane]

Have tried several different browsers and the same issue on all…

[hard beacon]

tried to delete the cache and/or delete cookies ?

[azure crane]

Have done, nothing changed..

[hard beacon]

well.. it works on Dasvon and also me.. so i don't think it's a server based problem

maybe a script crashed on your pc... i think you have also reboot your system?

[azure crane]

It’s for the browsers for mobile phones…

[polar spade]

The app updated earlier this month, and you run the app on iPhone you say, in conjunction with browsers. Do your apps automatically update. If not updated, it may not being playing nice with the browsers.

[hard beacon]

oh.. i am sry bout that.. didn't notice that...

my fault 🙂

[azure crane]

App… As I said… it’s all about the browsers on mobile phones.. several people on Reddit have the same problem. Hope this is a temporary bug…

[polar spade]

Apps affect browsers. Just saying. Yea. Hope it's a bug and is resolved soon then. Good luck.

[cedar oxide]

My account is asking me to buy Pro to edit when I have the premier plan

[hard beacon]

This is the tech-prompt talk channel.. you should write problem down in #archived-ticket-help

[unique stone]

Seems broken for me on Android, when in portrait orientation.
Works when i rotate to landscape.

[spice birch]

Whats up with suno? It's asking me to upgrade and only showing 3.5V

And all my credits are gone? WTH!

[unique stone]

Seems to just be a cosmetic bug. I have the same thing, intermittently. Similar reports over in #🪲┃bugs-web

Are you using a phone? If so, rotate the phone to landscape mode and back again.

[spice birch]

Yes. If I'm using my phone I have to put it in desktop mode to make it work. But it's hard to deal with lol

[kindred trench]

Is there a way to get v4.5 extend to actually extend an instrumental track? I've tried everything and I cannot get it to go any longer than about 20 seconds... it just tries to make another ending to the song. The old v3.5 system would often try to do the same thing but you could eventually get another minute or more. I'm lucky if I can get 20 seconds more out of v4.5, usually it just refunds my credits, and it's not even adding that much as I have to extend from a point earlier in the song than is necessary to get it to do it at all.

[sweet heart]

Hello all, Nileppez, here. first time on discord and relatively new-ish to Suno. Love the fact I don't need a band to make my writing into songs! It's really awesome to hear my words come to life. Great way to "journal" my emotions. Lol. Question, i see people talking about "heal edits" what is that exactly and where do i find it? Also, I've written a letter to Suno because it seems 9 out of 10 attempts to extend or replace are met with gibberish lyrics or are 1 min or less in length (meaning most of what I wanted is ignored and not included). Sometimes both. Anyone else having these same frustrations?

[sweet heart]

Do you ever try using the old 4.0 system instead? Does that one work?

[kindred trench]

no I haven't tried 4.0

[sleek stump]

Extending in 4.5 does not work like it did in previous model. You need to give Suno something to work with. Select the start point in the middle of the last verse, then delete the entirety of the lyrics window, then add in the lyrics window a number of instruction tags.. [Instrumental], followed by a number of lines of punctuation, as an example: $$$!%%%^^^^!^!^<<<>>>>..,^%@@$&&&&
Add a number of these sections ans then an [Outro] and an [End]. That should get you a 1.5 or more minutes of extension.

[unreal quail]

I'll give that a try as well. Noticed 4.5 being horrible in the normal way of extending a track. Lucky to get anything beyond 15 seconds

Could also add new lyrics to an extension before 4.5 and it would use them. Can't figure out how to do that with 4.5 and new set-up that's going on.

[graceful condor]

Is this where talk about not normal happening on your account and how to solve it

[kindred trench]

Thank you I will try it. How ridiculous though lol.

[sleek stump]

That would be here https://discord.com/channels/1069381916492562582/1372994553233276970

[bronze pecan]

Remaster 3.5 -> v4 🙂
Remaster 3.5 -> v4.5 🤩
Remaster 4.5 -> v4 🤩
Remaster v4 -> 4.5 🤮

How comes that the latter is so terrible?

[spice birch]

Omg! This editor crap ain't worth nothing! When you try to replace sections, the preview won't even play your new lyrics so you can hear. really ain't even got it to take the new lyrics yet.

[sleek stump]

Cant hear anything that isnt generated yet.

[elder dune]

So I thought I'd return and mention that I (sort of) figured out how to get a proper extension with 4.5. Basically:

1. pick the time you want the extension to start and input that in the place for it.
2. Don't delete your lyrics to that point, just change them or whatever, adding your new stuff where it belongs.
3. This is the important part: change the style prompt to describe in vivid detail what you wnat the EXTENSION to sound like. for example, if you're adding an outro say something like "the outro begins, maintaining the style from the previous chorus, adding heavy guitars and bombastic drumming. A choir sings the outro lyrics, repeating them a few times before everything slowly fades to the end" or something like that.

It's still kinda hit and miss, but I did manage to get something I could use for my song. Here is the finished product [after some editing and re-arranging], and I hope you like Gregorian Doom Metal, haha: https://suno.com/s/Ug0Ex0TXhyA8QvHQ

[undone locust]

Is there a resource that explains all of the character logic and tags for suno? See below for an example:

ohmmm oo
ohmmm shhhh(->)(<-)
--..*'._'*---..*'._'*-```

Edit: I would ideally like to learn this stuff and not ask a billion questions.

[undone locust]

Okay hmmmm anyone willing to explain what all of that means by chance then?

[undone locust]

hahahaha

https://tenor.com/view/guy-long-hair-mustache-magic-shia-labeouf-gif-17741293

Okay... well then I guess the next logical question would be....Any helpful tips any of you all have learned from experience? I'm here to soak up the knowledge 🙂

[raven crystal]

I have an instrumental song that I want to add lyrics to. What's the best way to do this in Suno?

When I try to "replace" a section, adding lyrics. It generates pseudo/mumble nonsense that sounds nothing like the words I typed.

[tribal remnant]

Imagine the layers of processing that are involved between Suno relating waveforms to unknown training text descriptions, waveform components produced from interpretation of individual characters, and interpretation of waveform characteristics in relation to musical and vocal styles... then prompt 😂

[sleek stump]

Its adding instrumental content to the track.

[undone locust]

How does Suno interpret the characters? Does each one have a differet use/application?

I'd love more granular control of emotion behind certain aspects (i.e yelling, screaming, crying, stress, whisper, sultry, deceptive, erotic, raspy, breathy, and so on). So much of song building is about tension and release and having more control over where and how that tension is built and released would be awesome.

Also @polar spade AMAZING response!

[undone locust]

I have been but tend to lurk.

[wise kindle]

Do scales work well in prompts? For example, if you're going for a middle eastern feel could you say "phrygian dominant", or are you better off just saying middle eastern?

[sleek stump]

Try both ways?

[rose crane]

is there a surefire way to get two different voices in a single generation? i'm trying to do a rap battle and i'm struggling a bit

[paper gull]

What have you found so far... Which ones have worked for you ?
We have only been using the () for background vocals, which we started by doing *( )*

[tepid forum]

anyway i was about to say, that good sh*t meme came out really well with the emoticons and stuff... a lot of times emoticons change up the song a heck of a lot. srry for the french amigos

like stuff like this... 👌 💯 👌 👀 👀 👀 👌👌 the eyes make a wow sound i think

If my hypothesis is correct, would work good sculpting some wubs or general sound design

[tepid forum]

duet, rap battle, rap-off, battle rap, and the ("") stuff works alright for small stuff... ummm... you could attempt to put [Singer #1] [Singer #2 in response to Singer #1] and so on... havent tried that

I made a show tune musical with some back and forth banter but that's about it

if you want to prompt a particular singer's style ChatGPT has a bot for writing out prompts to sound like bands/people... its hit and miss, but got it to sound like dude from modest mouse pretty spot on so thats neat

[tepid forum]

GPT is ur friend. i just had the idea to rearrange the structure to match a dubstep kinda song

[round wind]

I created a persona from an instrumental track that I use to create background music for videos, but in 4.5 it just recreates the same song in every iteration!!!
Is this normal?

[livid summit]

yes, I got that too

[livid summit]

have you tried like this / with slash in the middle?

[unique basalt]

hello~~,I want to add an instrumental in the middle of a song, but all that gets generated is with some vocals. i used [bridge]
[Instrumental] [Guitar Solo] prompt, style is A bright pop track kicks off with crisp synths, a driving four-on-the-floor beat, and layered handclaps. Verses use tight vocal hooks over bouncy bass, building into a soaring, harmony-rich chorus.

[fiery basin]

Does mike from suno ever solve account problems? because I've mailed a few days ago and I still can't use 4.5 on my personas without the issue that the same instrumental gets generated.

[unique basalt]

hello，i just wanna add a section only instrumental, but all i generated always has some vocals in it.

@tribal remnant

is my prompt wrong or some thing else?

[elder dune]

I've had reasonably good luck specifying a key and a mode, ie: "key of B flat, Locrian mode" if you want it super dark like for black metal. I have done a lot of middle-eastern sounding stuff, and I find that adding "strong middle-eastern influence" to the prompt will do the trick most of the time. You can also specify the key like, "key of C minor" and get something decent.

just gotta be aware that for every banger, Suno will probably give you at least 15 or more total stinkers :-p

[sleek stump]

This isnt an account problem from what I have seen. "Some" users have had similar issues and they solved it by clearing their browser cache and cookies. Also, if you are emailing Support, it could take as much as 4 weeks to get a response.

[fiery basin]

Thanks for your response Ginger, I've tried that, but didn't solve my issue. I've tried different browsers besides brave and the problem remains for me.

[pseudo blade]

emoji free could work

[elder dune]

if you really wanna go nuts with the middle-eastern thing, you can tell it "use traditional arabic scales and instruments" or something like that. I've learned that Arab scales, in particular, are completely different than the traditional western stuff we all learn. phyrgian dominant and/or harmonic minor (aeoloian) are decent approximations, but locals can separate them from the "Real Thing(tm)"

example I just made: https://suno.com/s/eQajSS8IAR2rRhET

[crimson yoke]

Why is there still no option to edit the genres of songs? This feature should have been available a long time ago. I have old songs with genres that make no sense, and those inaccurate tags are still there. Of course, I don’t think people should be allowed to remove genres, on the contrary, editing should require assigning the correct genre. Every song in the world has a genre as part of its identity. Hiding that is a mistake, it erases part of the essence of the track.

[unique basalt]

hello,there.i have a question. can i control the Chord Progression, For example, may I make a song with all the chords repeated from C to D to E to F? how do i write the prompt?

[atomic plover]

Same question here!

HELP

Looking for support to recreate this kind of sound!

https://youtu.be/f1-VkegjhzE?si=9SVrZewKC5s1lSMy

and/or

https://youtu.be/qt9gT3JoLmc?si=TeIJeXRfTRnoNLdP

[lost holly]

Sounds like orchestral intro, or specify violins and cellos, maybe hebrew folk, if Suno knows that or Hebrew traditional, also i would add slow, relaxed or prayer, i dont speak hebrew so i dont know what they sing about.

[rose crane]

Awesome thank you. And yeah me and GPT are already best buddies 😂

[lost holly]

This sounds like something for the #💡┃suggestions-web channel, here the idea is to talk more about prompts and whats the best way doing them.

[lost holly]

You could try describing this either in the style field or within the lyrics to Suno, like [Chorus, Chords C to D to E to F] or similar.

[pseudo blade]

can someone explain this?

this doesnt load

and dont ask why I use weather network look im canadian okay

[serene delta]

Try to log out and log in again

[pseudo blade]

k

[serene delta]

Maybe problem on loading server

[pseudo blade]

unless genres is broken again

[fast bison]

cleared cookies CC?

nah i can see them

that looks cache

[pseudo blade]

cleared all stuff

[fast bison]

try a diff browser is that safari?

safari can be funky some updates

[pseudo blade]

dont have access ti Safari

[fast bison]

sorry i was just suggesting any

i thought IT looked like safari, must be a skin

[pseudo blade]

its google

[fast bison]

ok try Edge

or whatever

see if still broken

[pseudo blade]

I only have google 😭

[fast bison]

can u not dl something else? guess not, no access to that right

[pseudo blade]

😭 sadly not

[fast bison]

its deffo not a global thing tho

[frigid orbit]

Play with the sliders a little. Either turn up weirdness or turn it down, Either turn up audio influence or turn it down. I never had those issues with a persona, but I bet you that might solve it.

[frigid orbit]

Can you put that same link in chat?

Idk how to do those symbols to recreate on my end

[unique basalt]

@atomic ploverthanks1

[sleek stump]

Are you on a Windows or Mac machine?

[pseudo blade]

windows but I deleted edge

[remote kelp]

Anyone figured out a solid, repeatable way to do a BPM/tempo switchup in a Cover of an uploaded track?

[remote kelp]

Shalom Aleichem!!

Did not think I'd see the name "Avraham Fried" pop up on this Discord today!

[atomic plover]

🙂

[vernal magnet]

Personally for 4.5, to solve the issue of persona replicating the original track, I use audio influence 20%, and 50-70% style influence. Anything above 25% for audio influence seems to give me the same song. Great trick for extending songs though in the same style!

[pseudo blade]

its still not loading here guys:
https://suno.com/style/粵語

[raven crystal]

I figured out an approach that worked well. I used "cover" on the original instrumental track, and then set "Style Strength" to 20% and "Audio Strength" to 65% (I tried a few different settings until I started get songs that were pretty close and actually used my lyrics), then I pulled the crank a couple dozen times until I got something good. 🙂 Let me know if you want to hear the original instrumental and the cover (not sure if links are frowned on here).

[crude pier]

Has anyone noticed that the sound quality in version 4 has dropped?
I’m finding it really strange...

[eternal dirge]

Hi i have a bit of a strange question. I'm wondering how i can get my songs on "New Songs" so people can see them? I've used Suno for 3 months and made about 15 songs, and i'm paying full Premium Plan. I notice that some creators have several of their songs on "New Songs" at a time, and very often. Is there something special i have to do to make it there so my songs can be seen?

[lost holly]

Hi, in general its kind about getting interaction with your songs like likes and comments. Advertise your good songs, for example on Discord here and hope that people will like them.

[eternal dirge]

I think the "New Songs" section is to kickstart your song the moment you publish it if i'm not wrong. Right now i see one artist with 3 different songs(brand new) on the "New Songs" section. I don't think you need to promote to make it to "New Songs"? Or maybe I've got it all wrong 🤐

[lost holly]

I think youre right, but you have to keep in mind how many people on Suno publish songs, so they probably wont stay there long. For your songs to be visible, you have to think in social media terms, get likes, get comments. If this is your goal.

[eternal dirge]

Okay. I'll try to do that. Thanks for replying 🙏

[eternal dirge]

This can be it. Because my first song i made with no effort made it to "New Songs", and it's by far my worst song, but the only one made it to New Songs lol! The other i've spent a lot of time on, and they never get the same love.

[queen tartan]

Hi everyone 🙂
I have a question 😄
There is a song, that is in public domain, meaning that it doesn't have any copyright, since the authors died more than 70 years ago. If I want to sing that song and ask Suno to add background music, or to use the cover feature, it will tell me that there are copyright issues and can't proceed. Is there a way around this?
ChatGPT is telling me that I could use Suno's Bark and make my own workflow.
Does anyone have any advice or experience in this?

If I buy subscription, can I bypass this by any chance?

[solid moat]

Does anybody know what kind of audio file one can upload to suno? None of the audio files on my Android seem eligible. I just recorded some stuff and no go. I've got a couple of demos I'd love to run through it.

I've tried WAV, MP3, stereo, mono, different bitrates... Nothing's working. Android, Chrome

[scenic trellis]

Hey!

Just wondering if the Suno engineering team hang out here. I have a weird question... 😆

I'm working with a clinic in the UK, they have a programme called Audio Integration Therapy, which uses music to help with behavioural, cognitive, physiological and motor impairments that result from auditory processing deficits encountered by children.

I'm experimenting with using Suno to produce music we can then post-process to be suitable for therapy.

One of the key requirements for music suitable for use in this way is that it has a good distribution of frequencies across the human-audible hearing range.

I wonder if it's possible to prompt Suno to ensure a consistent spectral profile in its output?

Please let me know if this is something possible, I'd love to be able to generate a large library of tracks with Suno that could really help transform children's lives in partnership with the clinic!

[unborn star]

What's the problem in general?

You don't have access to your account? Or what's the matter?

[mental folio]

they falsely cancelled my subscription instead of deleting an account I had opened by mistake.
So now, I lost all the songs in v4.5 created with my premium account, have no access to my personas, and lost 9000 credits.

And as usual, support does not respond.

[lost holly]

Hi, public domain means copyright protected, since the public domain only applies in the us, so no, there is no way around this other then altering the song lyric enough so it wont be recognized anymore.

[lost holly]

Hi, that in general should be possible, although there wont be a way with a guaranteed outcome. What you can do, is describing the desired outcome to Suno in the style field and then check the results, then adapt the description/prompt from there to improve.

[lost holly]

Have you checked the trash folder to see if you by accident deleted the song? Go to library on the left and then top right the trash bin to get there.

[neat lake]

Hi, a blind person here, using a screenreader. I have the pro plan, but can't see the button to enhance prompt anywhere. Is it near to some other recognizable element on the screen and how does it work? I mean, if I write a prompt, will it make it more detailed and focused ofr me if I press the enhance button?

[main gulch]

It is not in the trash folder

[neat lake]

The song quality seems to vary from time to time and I've been observing it since the first models dropped up until now, however nobody from the staff seems to be acknowledging it or has any idea how to handle it. People who aren't sound engineers and have trained ears often think we're just making it up lol.

[lost holly]

The button is on the top right of the Style field, you need a subscription to use it, as far as i know. The button is named "Gives your prompt a creative boost (Beta)".
It makes the prompt more detailed and gives it a structure that usually applies to songs.

[neat lake]

Thanks, maybe I'll try on a different browser, because I couldn't seem to find it when using chrome, though I saw all the other options. Do you know if that element has text or is it only graphical? Because that could also be a problem for my screenreader to recognize?

[lost holly]

The element is an orange button with a pencil inside, so yes, your screenreader might not recognize it.

[neat lake]

Thanks, Suno forgot about us again lol

[polar spade]

If you hover over the magic wand it has a popup of " gives your prompt a creative boost (Beta)". Annoyingly, mine keeps the popup stuck open and so covers my prompts. 😁

[neat lake]

That could also be the case for me actually, because the options for song creation aren't always in the same place, at least in the browser I'm using. Thanks for your input!

[polar spade]

True. Im running three different browsers, so they act different depending on my display settings.

[polar spade]

For all concerned: Anyone having more than one account -- take time to follow yourselves. This gives an opportunity to retrieve or at least have some access to any public songs you may have on the other account, if the need arises.

[remote kelp]

What triggers Suno's explicit filter on lyrics?

[polar spade]

No one knows. Honestly, it seems pretty harsh but non-specific.

[polar spade]

Sometimes, just from rearranging the same lyrics, it can pass the filters. So, in understanding a phrase or words, the surround words may have an affect or even the position they have from each other. Just moving a smidgen to the left or right can make a difference.

[remote kelp]

I was able to copy paste lyrics word for word from a brand new hip hop song by a famous rapper that has N word and a bunch of expletives and it gave me no issues lol

[neat nexus]

Huh. Which one?

[polar spade]

The problem is, if the song is later reviewed and found in your account, you could have further issues. 🤕 I'd avoid too much "experimenting".

[remote kelp]

I'm not making it public lol

[scenic trellis]

Makes sense, what I’ve been doing so far is generating tracks, then doing a Fourier analysis on the downloaded MP3s and discarding those with more than 10 seconds of missing frequencies.

It’s a shame because I’ve ended up discarding some really cool tracks.

I’ll try prompting to make it ‘full spectrum’ and see if that helps!

[neon roost]

Hi

[quick wedge]

Hi. Welcome to the Suno community. Hope you’re doing well this morning 💬😊👋

[neon roost]

Thanks. I’m trying to figure out how to use suno/bark ai.

[quick wedge]

You can check the GitHub site search Suno Bark Ai and there should be instructions on how it’s used from there. I’ve always used the main site http://suno.com to create on there. Someone else might know more about Suno Bark though.

[autumn rock]

Apparently or unfortunately I have found a way to get mispronunciations consistently from female vocals. I must have hit the perfect structure/tempo. the words "weird" and "odd" (now add "Bold") cannot be vocalized in any way phonetically or not. I sent a link/bug thru the form but would love to hear back as this song style/effort is now officially dead. I have wasted enough credits on it and consider this a useful test case for dev.

[polar spade]

Yea. It would be difficult to work out when a combination of styles and lyrics and arrangement would consistently then produce an unwanted outcome. Those things are expected to happen now and then.

[autumn rock]

I do test cases for a living, so i am hoping they can genuinely use this song for testing. Hopefully debug it and see why this combo consistently fails. I could not help myself and tried rewriting the cadence, tempo and it did not let me down and this time replaced "weird' with sweet.

[polar spade]

Thats cool. Well. I hope its of benefit to them. They certainly have likely been spending much effort trying to find a sweet spot that works for all cases. But at time they need fo manually tweak it as well.

That kind of why 4.0 Remaster was brought back. 4.5 didn't seem to work the same in all genres and cases. That has seemed to already improve, though.

So any thing you can send them, im sure they appreciate.

[polar spade]

My one word was "calliope' that none could get right, consistently. Only recent tweaks have helped.

[sleek stump]

Key words here.. "brand new" ... Suno's filter is through ChatGPT, so anything it thinks is NSFW gets denied, as well with copyrighted or otherwise released content. If CGPT hasnt seen the new content yet, it will allow it.

[sleek stump]

A canceled subscription does not mean you lose your generated music,, you still have access to all the music you generated, even if the account is now set to free user. The only way you would lose access would be if they closed the entire account or banned it.

[lost yarrow]

Hello everyone, I'm from Russia and I'm using a translator to communicate. Please help me. I don't have a premium subscription and I can't afford to buy one, but I need to extract a minus from one song.

[mental folio]

Right.... what I meant is that you can no more edit your songs, remix or/and use personas. Of course you can still play the songs.

[lost yarrow]

The thing is, I need the Get steams feature to separate the voice and audio in the song, and I need a backing track to sing the song, but I don't have a Pro subscription or the option to purchase it because I'm from Russia.

[polar spade]

Ahhhh.... hmmm.... there are outside third party programs and websites, some have trials that can help split stems. I never write them down, but searching stem splitter apps or AI step splitter should bring up a few. It all depends on what Russia allows as access as well within your region.

Ахххх.... хм.... есть внешние сторонние программы и веб-сайты, некоторые из них имеют пробные версии, которые могут помочь разделить стемы. Я никогда их не записываю, но поиск приложений для разделения стемов или AI step splitter должен выдать несколько. Если все зависит от того, что Россия разрешает в качестве доступа, а также в пределах вашего региона.

[lost yarrow]

Can you tell me their names then?

[polar spade]

Let me see..

[hardy thunder]

I use bandlab.com for free stem separation (and also sometimes for remastering) and Audacity on my PC. Well Dasvon provided a huge list 🙂

[polar spade]

Audacity, AceStudio, Moises.ai, audiophile, BandLab, Voice.ai, singify, Kits.ai, Fadr

I never recommend one over another, really. Each person has their preference. And some sites dont offer much for free or free trials.

BandLab and Audacity are free downloads..... and FL Studios, I believe.

[lost yarrow]

Okay, I'll try it now.

[polar spade]

OK. Ill remove my video example. 😊

[sleek stump]

Don't forget Stemroller, a free splitter for PC.

The splitter you can get for Audicity is horrible..

[hoary burrow]

Where can I see a receipt of my Suno usage guys?

(pls)

[half cave]

How I can find a list of rumor ( Waves of sea, wind, bang) i can add on my music ?

[polar spade]

Suno doesnt do too well to consistently create sound effects exactly when you need them. There are a few such as clap and beep, that work fairly well.

Even if something works, such as thunder , it often doesnt come in at the right time.

Others may have a few suggestions though.

[half cave]

Ok, thk

[polar spade]

I never stop trying. Now if I get a chicken sound or something, then I start with the sound and try to build the song around it, rather the try to fit a sound somewhere into a song.

[sleek stump]

The bottom line is that Suno wasn’t trained on sound effects, which makes it difficult to get when needed. What we can get randomly most likely were part of songs that is was trained on, so it may throw them in here and there.

[worthy fog]

i was told to send my issue here. i am trying to regenerate lyrics to one of my songs, but it is stuck permanently loading. i have tried it on different computers in different geographic locations on different browsers

its just stuck on this

[sleek stump]

Try the Legacy editor

[shadow plaza]

Does suno writes better lyrics or GPT?

[sleek stump]

ReMi and classic and GPT are all sort of different, but all of them are obvious AI written.

[polar spade]

Suno constantly pings to assure functions are verified as coming from the same browser, same IP address. Using a VPN could create issues. To test, you might try logging in incognito to set aside cache files but still be functional and keep your address local.

[shadow plaza]

Ok which one ia better SUNO or GPT

[sleek stump]

That would be up to the user. Most of us that write our own lyrics can tell instantly they are AI written. Try ReMi and GPT and see what you get.

[shadow plaza]

If there others out there AI specialize in this?

Writing lyrics

[shadow plaza]

That sure i write my own but in my language.. My English isnt great for writing lyrics 😅

[sleek stump]

All any of us can recommend is that you try them all, see what you like. You can use GPT, or Grok, or a few other LLM, or try the 2 models in Suno (ReMi and Classic).

[shadow plaza]

So they are different Suno and GPT?

[sleek stump]

If you are trying to translate from your language, use GPT or Grok

[shadow plaza]

I tried but sound weird in english not enough words to describe

[sleek stump]

Suno doesnt do translations

[shadow plaza]

Its simple language

I meant GPT when translate

Is Suno writer based on GPT? Or its separate

[worthy fog]

ok so even though i wasn't connected my vpn, closing it did something. it still isn't generating, but at least it told me that it timed out. i'll wait a little bit more time but if it doesn't work again then i'll try incognito

[polar spade]

If depends on what you call better lyrics. But I've seen both gpt and remi gives some great hookie phrases.

The one advantage of chat GPT is its a learning AI. So if you feed enough of your own songs, then it will get better at writing in your style.

[shadow plaza]

Isnt the remi that give lyrics with no sense at all?

😅

[polar spade]

Great. Try what you think best. The best we can do is throw out some suggestions based on whatever detail you give.

I always include a basic one line lyric/one genre test. Do the bare minimum and see if that works. If it does then we can start looking at lyrics over 5000 character, plus too many prompts and a long title that can simply dam up the outputs from being created.

[polar spade]

You also can't have a conversation with ReMi to describe further where you think the lyrics need improvement.

[worthy fog]

my edit is 342 characters long lol

it is a very simple edit

imma go try the incognito now

[polar spade]

OK. Youre welcome. Just throwing out some thoughts that could help now or in the future. Good luck mate.

[worthy fog]

lol thank you, i appreciate it

ok its not working on incognito

once again i appreciate the input

ill fill a support ticket ig

[polar spade]

Did you say you tried Legacy as well?

I just tried a random song, using Opera browser on Android phone using current Edit UI Classic context with 32 beat context window on a full verse and changed words in every line. Got two replacement within 1:30 minutes. It replaced the word changes with no problem.

If you wish, Opera works well for keeping the display format within field of view, and for being lightweight and fast.

[main gulch]

Many important songs of mine are suddenly today no longer available in the library. Can someone help me please?

[half dragon]

Check the trash bin if they got there by accident?
https://suno.com/me/trash

[main gulch]

Thanks, they are not in the trash folder

[half dragon]

Strange.
Songs don't just disappear all of a sudden.
I always put my favorite / important songs in a playlist for easy access, without having scroll through milions of songs in Library 🙂

[sudden sedge]

hello, i try to upload my audio but it upload endlesly

[polar spade]

Check for turning off all filters in the main Workflow [especially the "hide clips from edit mode"] as we edit most thigs, to see if they then show up in the library.

[main gulch]

It's a good idea, tomorrow i check it out, it's now late in germany. Thank you!

[sudden sedge]

hello, i try to upload my audio but it upload endlesly

[main gulch]

It's a good idea, tomorrow i check it out, it's now late in germany. Thank you!

[half dragon]

Did you use pc or app to upload?
How long is the audio?
Did it happen on all uploads or just this file?
I uploaded mine and it was working fine, and I'm using pc browser.

[sudden sedge]

is pc, under 4 minutes, there is app for pc?

i tryed other recording of mine is the same

[half dragon]

No, just web browser

[daring lava]

S/O to @half dragon @sleek stump @polar spade for being super awesome and always helping people out! Very big deal! Can't begin to express how awesome it is to see you three helping so many people with things! Really does make a difference in so many ways! Celebrating you all big time!

Quick question... I'm making you all a song. And genre you'd like in particular?

[half dragon]

On browser, show the Developer console - then try upload again, see if there's any error that could help to figure out what was wrong. Could be the audio file itself...

[sleek stump]

Heh... uhhh... Classic 80s rock?

Era of the big hair bands 😄

[polar spade]

Deathmetal disco polka? Im kidding, of course. Oh god, what have I done.... 🤦 💀

[sleek stump]

And Heart 😉

[half dragon]

Latin Euro Dance?

[daring lava]

https://suno.com/s/JFgpU55zhkPXpZRu

[daring lava]

https://suno.com/s/3gPhfPj12HikjsFg

[daring lava]

https://suno.com/s/kihfewH7utmepwO1

😅

Seriously though, thank you all for doing what you do! You are very amazing people and Im very grateful to be in a community with people of your caliber

[sleek stump]

It helps to have positive people around, Thanks for being one @daring lava !

[naive ibex]

Did anyone notice another silent update of the v4.5 model drop? Regarding changed styles. I'm used one style a lot and it was calm melodies, yesterday it became almost 100% let's say I could describe instead of calm it became crazy-leg rapid flamenco....
💃 👯‍♀️

[remote kelp]

It's weird that it didn't flag the song I put as NSFW even though it definitely is.

Are you sure that Suno are using OpenAI under the hoot to vet lyrical content?

[sleek stump]

Yes, 100%, was mentioned by a Dev Team member about 1.5 months ago. I am not sure what the conditions are for NSFW content, some stuff gets intercepted and some does not.

[remote kelp]

Thanks 🙂

[cunning mulch]

What happens if you turn off "Hide Disliked"? I have a few songs that won't show up until I do that

and yes, I have 'em given thumbs-up

[olive meteor]

Hi, I'm new to Discord. Someone who can guide me with the following 1) When I create a song and then I realize that I want to change a word. SUNO doesn't let me edit the same initial melody. It gives me 2 other versions but with another melody, not the one I want to keep editing. This can be solved? a version that I really liked was cut at the end and I don't know how to complete the missing phrase. How can I fix it?
2) Why sometimes the song is cut at the end? I don't understand if it has a duration of less than 4 minutes which is what it indicates that it supports and my song lasts 3 minutes I don't understand why this happens.
I appreciate your guidance

Translated with DeepL.com (free version)

[sudden fog]

Hey hey i was wondering what is the prompt for specific voice placement suno is not following my instructions I'm wondering if I'm doing something wrong? I was wondering if anyone had advice or instructions on how to do that?

[marble wing]

Is there any problem with suno? There is a SERVER ERROR message on the web page

[glacial cobalt]

Getting the same error...

[wide karma]

ditto

[west otter]

Same, every time I try to access the website, it keeps giving me an error, unable to actually open the website.

[fading tulip]

@west otter Server is down. Don't panic.

[west otter]

temporarily I hope? 🙏

[twin flicker]

hmm... every time I use v4.5 to generate a song.. I notice something odd
even though it says the model supports up to 8 mins whenever I provide longer lyrics the resulting music almost always ends up with the vocals being rushed.. like it's trying to cram everything into under 4 mins..

It really feels like the model is compressing the song unnaturally fast just to hit that shorter runtime.. which kinda defeats the purpose of writing longer lyrics in the place

I'm wondering if anyone else has experienced the same issue
and if so, has anyone actually found a way around it? I'd really appreciate it if you could share any prompt strategies or insights that helped you solve this

[shadow plaza]

How you erite promts for the style in 4.5 model?

1. You explain it how sound should be (very long detailed description)
2. Use only keywords (Rock, electric guitars, heavydrums etc...)
3. Use only the name of the style you want?

[sudden fog]

I can't seem to upload music i want to use in my cover song, I finally got the lyrics how I want them but now I can't finish the song could I get some help. When I press upload audio it greys out everything and I can't select any files I can move between downloads and audio areas but not select any files have I not formated something right or is there something wrong with suno ai?

I figured out the problem it want a different audio format my mistake

[queen tartan]

Is there a good tutorial to Suno's Bark that I can build my own workflow?

[smoky crow]

Will AI generated music ever be able to maintain consistency in the generated music ? I never get a single straight instrumental techno track with a steady consistent beat and mixing. low compression in the beginning, ends up compressing everything up in the end especiallly when it's over 4 minutes and it's getting a mono signal at the end.

[half dragon]

I usually do step 2 and 3 that you listed there, then click the magic wan on v4.5 and it'd give me that on step 1, full long detailed description for the style.

[half dragon]

Have you seen this video? There's also a link to JonathanFly git repository as well - might be helpful
https://www.youtube.com/watch?v=8yzcHGzrBMQ

[sleek stump]

The genre or fusion of genre you detail plays a huge part in how long the songs can be. Some genre are easy to get 6-8 minute tracks, and very difficult for others. If you really want a long track in a genre that normally doesn't have longer tracks, you will likely need to use the extend feature to craft out what you want.

[polar spade]

This might also be some insight if it helps. https://discord.com/channels/1069381916492562582/1382451659434758276

[twin flicker]

yeah I get that the genre prompt can affect the song as you said
but in my case, even when I use the same style prompt, longer lyrics often make the ai vocal rush through the song like it's rapping
when I use shorter lyrics in the same style and structure, the vocal sounds way more natural and easier to listen to

the song sections are exactly the same and the difference is just 5~8 words per section

just to be clear, I'm not talking about the tempo of the music itself
it's not about the beat, it's the speed of the actual vocal delivery that becomes unnatural

[sleek stump]

What genre are you using and finding this behavior?

[sudden fog]

@polar spade thank you so much!

[polar spade]

Its a workaround until they work out why some files are greyed out on uploads.

[sudden fog]

Do you know anything about duet placement prompts by chance?

[twin flicker]

um.. I usually go for stuff like rock ballads, indie rock, or pop rock when I make songs
I've tried some other genres too but those are kinda my go-to

[sleek stump]

Theres a lot of complaints about Rock and sub types in general, most revolving around hard/fast/metal styles. Its pretty well known at this point that 4.5 needs some work there. That said, a lot of us have been successful in being able to get good first gens, but almost always we end up fixing stuff in post in a DAW.
You can also use extend to fix broken last sections, or some of the other newer tools. Unfortunately, its not cheap credit wise to do a lot of editing in teh Suno environment.

[polar spade]

@sudden fog I use a simplified form of just putting [male vocal] or [female vocal] before the segment. The rest is somewhat random, as genders are not genres and both male and female use similar medium range frequencies. So finding sucess to get the right gender isnt 60/40 or 50/50 , its more like 30/70 chance.

Most "controls" seem to be in the genres and lyrics used to indicate to Suno one gender or the other is likely needed , ot that this song lyrics cry for alternating singing voices -- so thats more of a learned process.

Using Legacy editor is not too bad to try to use prompts and change gender. I use lots of credits and just click 5 or 6 times until I guess the one I want by using replace . And then im happy.

OR now with new Edit UI, I can copy and paste in a female vocal-sung verse and then replace the copied lyrics with the new lyrics that should be there More than one way to do these things.

So I do my duets the lazy way I guess. 😄

[brazen yew]

Does anyone find that the simpler and snappier you make lyrics the better overall quality of the music?

[hardy cobalt]

Hello,

I would like to know if we can put several voices of the same genre or different genres in the same song, as if it were a Pop group. Kpop, etc?

If we can do it, what are the prompts to use?

[polar spade]

Yes.

[polar spade]

By definition, kpop , jpop will often have group or multilayer choruses.

In simple terms, getting separate voice can be about spacing out phrases from each other, indicate changes by new quotation marks, and even use [spoken] in places to sometimes change a voice even though it will sing instead of speak.

Prompts can be all the typical, kpop, jpop, bubblegum, hip hop, or whichever direction youre going, but you might add broadway, ensemble, opera, kids song, pr accompaniment to inspire additional voices.

[sudden fog]

Thank you so much unfortunately I am on a phone and I am no longer able to use the editor

But the information has definitely given me a few ideas to work around

[polar spade]

Other browsers such as Dolphin or Opera and maybe Brave can display the editor UI if you have those available.

[sudden fog]

Cool I'll have a look into it i have noticed that the function to slow the music down is not working for some reason its really helpful a more doom and vengeful feeling for my story telling

Are there server tweeks happening at the moment?

[round horizon]

question, has anyone managed to get Suno to do dialogue like a radio show? I've made a huge playlist of songs that'll be used in a video game that I'm working on and I'd like to create a radio station where the concept is kind of like V Rock in Vice City. The game is retro scifi so I think it'd be cool to do like a grungy/transatlantic accent for underground radio show

[lost holly]

Like this? 😄 https://suno.com/s/xCSjisXuy6AutTMv or do you mean it different?

[bright cargo]

"Aggressive post-hardcore energy drives the song with distorted guitar riffs weaving through raw punk drums and a pulsating bass, Verses lean into melodic rock, balancing the intensity, while choruses crash into metal"

Put this prompt in if you want to get the same 2 riffs in every generation. Support is completely ignoring it even though their AI is completely broken. Probably spent 8000 credits trying to get around these riffs. Terrible. No excuse for leaving me on read.

[round horizon]

oh interesting, thanks for sharing!

[lost holly]

https://suno.com/s/lPJARMDWiSC8Cm5T a bit better version with mostly 2 hosts, although they sing here and there it worked decently

[polar spade]

It was down last night for a bit. Maybe there's something brewing.

[sudden fog]

Maybe

[elder dune]

seems like the "change lyrics" inteface changed overnight, and isn't working as well as it used to

it either botches the replacement completely, or makes the replaced section like 3-6 db louder than the surrounding stuff, resulting in a choppy and obviously bad mix

[stone spruce]

I believe now that duets are a deliberate scheme by Suno to drain my tokens, I've tried so many different wants to tag a song to get male and female vocals to work on specific lines it's always just random chance if it recognizes some or ignores them. I've got so many songs that would be perfect if it just followed instructions, or songs that I've just had to suck up and accept one flubbed line because it's not too obvious. But in some songs if the wrong singer says a line it destroys the whole song narrative, it's incredibly frustrating

[lost holly]

Out of curiosity, what have you tried so far? Do you have an example to share and look at for us?

[stone spruce]

Oh I've tried male voice, female voice , male or female vocals, man's voice, woman's voice, I've tried all kinds of brackets, tagging literally every line, I've tried vocal switch tags, I've tried to encapsulated the whole line in a bracket or quotes

It would be easier for you to tell me something I haven't tried at this point

Like, the tagging works sometimes but never constantly and never for the whole song u less it's a very predictable back and forth style

Like if you just have one line handing back and forth, it'll Sometimes work but have a 50% of getting the order wrong

If you have an irregular style it never works

Like [male]
Line 1
Line 2
[Female response]
Line3
[Male reply]
Line 4

That kind of structure I've never had it work constantly at all with

[lost holly]

Something like this? https://suno.com/s/lPJARMDWiSC8Cm5T of course it doesnt works perfect but yeah

[stone spruce]

Now if you just did interlaced lines it can sometimes do that fine but that's just accidentally

Yeah I've tried that

Like I said, it'll sometimes work, but her for the whole song and it seems randomly

Like the very first line can explicitly say female voice or female vocals or female singer

And half the time it starts with male

So it really seems like I just ignores tags most of the time and just arbitrary decides which lines are male or Female

Like the duet is pretty reliable, rarely do I get only one voice or whatever

And it forget trying to reliably get both in the chorus, that's a complete crap shoot

[lost holly]

Understandable, i know most of the issues you talk about. What usually works best for me is a small number of styles like max 3 and () which dont switch every line, better works if it switches like very 4 lines or so. But yeah i dont know a perfect way

[stone spruce]

I just did like a bunch of generations, some better than others, none work all the way, I'm on my phone, I can probably make an extensively library of all the ways I've failed at when I get on my PC

[stone spruce]

Yeah I've had to work around the limit, and I've got some great songs

But this one specific song is more narrative and i cant just have it handing back and forth

And if we have to work around this, then it's really not following the prompt at all, we're just accidentally getting the right vocals by gaming the flawed system

I hope for 5.0 they can nail this because I do love songs with more than one voice but most of the time I don't even try because it's unreliably

The best I've gotten is the entire verse being a different voice

That is like , still unreliable, but more likely to be successful than a complex back and forth

[lost holly]

Agreed 😄 which song do you try to do duet for? Can you share a failed example, want to try myself 😄

[stone spruce]

I'll try to collect some examples when I get home

But recently I've tried just simple
[Male voice] lyrics
Or
[Male voice]
Lyrics

Or

[Male voice:
Lyrics
]

What's funny is even when encapsulating in square brackets and the voice prompt only being separated by a colon, it never sung out the voice prompt so it seems to be aware that's not lyrics, and it never skipped lyrics inside the square brackets

So on some level it's able to distinguish the lyrics from the tags

[lost holly]

Yes, Suno recognizes a few key words, if you go to far away from those the [] wont mean much though 😄

[stone spruce]

Yeah, I love this service and it makes some real bangers but it's never fun when you get a specific idea and generate 50 songs and none do what you want

[polar spade]

I've gotten more to letting Suno divvy up the singing as it feels best based on training. So I've put more focus on some prompts and massaging the lyrics to influence who sings, but otherwise, using less metatags to try to force anything. Concluding it only has about a 20% success rate, which means its kind of random anyway.

Heres an example of NOT using any tags , a modicum of prompts and just literally writing lyrics for the duet and letting Suno figure it out.

https://suno.com/s/YtszR42QErhHqQq2

[elder dune]

Just finished this - except replacing 1 single word in the lyrics that the stupid editor just won't do right (30+ attempts) - uploading the MP3 because I did some post-editing in audacity, too .. and because it's LONG. I thinks its sheer size was kinda choking the suno editor, too.

[round horizon]

check it out!

https://suno.com/s/PL9jvj92QsQjg4Db

thanks for your help

it does sometime accidentally switch voices but I should be able to cut it out of the radio show

[lost holly]

It does sound good, sadly we often have to try alot to get a good result 😄

[round horizon]

I think what I might have to do is one voice at a time, make a persona then output their tracks separately

[mossy trellis]

I need help placing sound affects in my music. Nothing i do seems to work

[lost holly]

Good luck, had cool results while experimenting with speeches, but it was alot of tries to get good results.

[lost holly]

What do you want to achieve?

[mossy trellis]

im a paramedic and making a song about ptsd. i want things like in the intro have the sound of sirens fading and a heart monitor flatlining

[lost holly]

Have you tried describing this in the style filed? Something like: "Sirens creating a tense atmosphere, fading into a flatlining beeping in the intro..."

[mossy trellis]

i have. ive also tried the ** and [] in lyrics

[lost holly]

All with Lyrics or do you aim for instrumentals?

[main gulch]

Hello Tiberious and friends from Suno, In create mode, all songs are available in my workspace, regardless of the filter setting. Only some are missing in the library. Are there perhaps too many songs in the library?

[lost holly]

I tried a few things myself and dont think that i will have a helpfull suggestion...

[wanton dune]

https://soundcloud.com/dom1nati0n/whack-a-doodle

You can do it reliably every time if you copy the original track you get the duet in, and just re-create it at about 20% original track strength.

I use the original track as the persona and rebuild it in v4.5 at 20% strength. I place both of my voices in the prompt field, And then I seperate the verse and chorus by genre not the singer.

I don't talk about the singer in the prompting.

You don't need to. Just seperate your music in the style prompt and let it pick one. I've got entire role-reversal tracks because I didn't really choose the first solid version it just wants to duet if there's more than one voice type in the field, so let it pick a couple and you can always make a new persona out of the next track that's more accurate. Just pick at it a little at a time it's fine.

[main gulch]

Hello friends from Suno, In create mode, all songs are available in my workspace, regardless of the filter setting. Only some are missing in the library. Are there perhaps too many songs in the library? Or why are some songs in library "hidden"?

[polar spade]

Firstly, you say they are shown in workspace "regardless of filter setting". If you click "hide stems" in workspace and the stems still show in the main workspace, then already, there is a problem.

[polar spade]

As Library has no filters it should archive everything you've done. There should not be a limit of songs we can have in the library.

In the past, its been noted that the workspace filters are affecting showing the same files in the Library. We almost need two browsers open to clear the filter in workspace and one browser to check for the files in the Library. I dont know if this has been rendered fixed or not.

[main gulch]

when i press clear all in “filters” nothing happens, filters remain on, but I can click them away individually

[half dragon]

Not sure if it was intentional, but Clear All only clears the lower parts of the filter; such as filters Upload, Full Songs, Covers...etc.. those with radio button, those with checkbox will not be cleared by Clear All.

[polar spade]

Is it Clear or reset? Hold on....

[half dragon]

It does say Clear All

[polar spade]

Yea. Clear All.... and it also has in parentheses how many to clear. But apparently not true. 😊

[half dragon]

Definitely a bug

[polar spade]

Reset to default, makes more sense if the default was to have two of them always checked. 😁

[half dragon]

Funny Hide Stems by default is on, but once off - it's no longer default on Clear All (or reset)

[polar spade]

Choose our own default. Cool. Some bugs I like.

[trim abyss]

Change lyrics works horrible. It's faster to generate 100 songs with hope of hit than struggle with this feature...

[thorn sparrow]

can anyone tell me name of this famous sample, i created a song via suno and this happend 😄 i know that melody, but cant tell the name, anyone else?

[half dragon]

I use Google to listen to the audio to id the song - came up nothing matched 100% - show me 3 songs with around 11-15% matched - I listrened to those and only first riff is sounding similar, the rest are not... so it might not be anything super famous?? just a familiar chord progression?

[thorn sparrow]

thanks i dont know but i remember this melody somehow

[round horizon]

okay another question, I don't suppose Suno separates the music from the lyrics as layers in the editor and I'm just missing the feature?

[half dragon]

Are you talking about Stems separation feature?

[abstract flower]

You perceive it as "famous" because it's just a generic rock chord progression with a generic motif of chord tones over it, which you'll find in thousands of songs. Some 20 of them, give or take, will be indeed famous... 🙂

[main gulch]

thanks, I'll try and get back to you later

[abstract flower]

Folks, what are the current rules w.r.t. uploading a song with a singing voice? And maybe your general advice. I recall that a year ago, it would be locked as private and could not be published, and any remix of it would not as well. Is it still the case?

The thing is, I promised someone to translate their Suno song to another language. I liked the song so much that I played it by myself, first on piano by ear, then all instruments, with real bass guitar, so I have my own stems of it, individual tracks in Cubase, everything except the sung line (syllables won't align in translation anyway). I want to preserve the feel of the vocals; the vocal phrases are short, vary in length, and each line breathes within the 2-bar space which it is allotted. I like my version better, I played the bass guitar and (sampled) piano with some embellishments that do it good; the original bass is just whole notes all the way. I cannot sing, so it's Suno who is going to sing. How would you approach it? I think of playing the vocal line to it, played on any synth, but how do I tell suno to replace this melody line with singing — did anyone do such a thing? Or should I leave the vocal melody line out entirely, upload the pre-mixed accompaniment only, let Suno just come up with its own melody to sing the lyrics, and try multiple times until it comes with something acceptable? I'll leave frequency space for voice (it's not damn opera, the voice range is quite narrow), but It should fall into the space I'd leave for vocals, so it wouldn't compete with instruments. There are occasional piano licks in a higher register than singing, so there has to be a tight fit, frequency-wise.

On a tangent, I could have Suno sing it, then pull the vocal stem and add it to my mix, and bypass Suno's "remix" — with its mixing and mastering — altogether, I do it easy and much better, but then it would become an upload with voice, so see the question one... And it has to end up a published Suno song.

[half dragon]

I remember seeing a mention somewhere that they changed the rule, upload with vocal can now be published; the only thing you can't do is to create Persona from it.
Though I haven't seen any official update or announcement of that; but its easy to test out.
Updated: I just tried and can confirm that I can publish my upload audio with vocal.

[abstract flower]

Wow! Thank you so much, it's so sweet of you to check it for me. 💖 This opens a whole lot of possibilities! 🤩

[shut forge]

Why is that sometimes Suno doesn't show all the lyrics in edit mode?

[abstract flower]

Aaahh! I know this progression!!! Pachelbel's canon, the very beginning https://www.youtube.com/watch?v=Ptk_1Dc2iPY

It's been reused in all modern genres, so you are certain to recognize it.

[sleek stump]

Sounds like Canon in D...

[abstract flower]

If you trimmed the song, or extended, got the whole song and extended again, the lyrics will fall out of alignment. Other actions may result in this, too. The old editor has the option "Align lyrics", and it was still available couple of weeks ago. I could not figure out how to use the new one, sorry. In the old one, it was on the hamburger menu at the very top left, a lone small hamburger too easy to miss, but it is there. Provided the old editor is still available. Always use the desktop site for the editor, even on a tablet or phone.

[abstract flower]

Our letters crossed. 😂

[sleek stump]

Legacy editor is still available, can select it from the top of the new editor

[clear thunder]

My issue is the style and timber of voices. If I want a amapiano (south African music) I would like the singer's voice to feel like they are from (South Africa) and currently there isn't enough diversity in the vocalists.

[magic shale]

the generations today have an awful sound !

[sleek stump]

Pretty sure Suno didnt get a ton of SA music in its training...

[clear thunder]

I don't know why ... it's very popular music...

[sleek stump]

And to be fair, it mispronounces stuff in every language, even English.

[sleek stump]

On teh world stage? Locally maybe..

[clear thunder]

no, afro-beat is world wide.

[sleek stump]

Afro beat is not the same as like Afrikaans

[clear thunder]

I didn't say Afrikaans. was said amapiano

[sleek stump]

Just saying, the vocals you are complaining about likely didnt get a lot of training.

[clear thunder]

I've had the vocals, i'm talking about in previous versions. NOT in 4.5

I get what you're saying.. but because they actually have the music of amapiano... there should be some consistency in the accent/style/of the vocalist who makes that music.

[sleek stump]

🤷‍♀️

We work with what we have, and provide feedback for future improvements

But older models are set, they arent being updated at all.

[clear thunder]

awesome. Then just consider this feedback.

[remote kelp]

I do this all the time, feel free to DM me

[short elm]

I found the perfect smart prompt for 2010's Progressive House.

[naive ibex]

And to be fair, it mispronounces stuff in every language, even English.

Do you mean singing mispronounces? I have might 10k+ generations. In my native language, I in 80% cases can not distinguish is it real person or not, and in 10% it near perfect, and mostly would never believe that it is generated if not created be myself. Except when they messed up with Jan mega shimmer and other defects (I would to say "features") that they drops randomly time to time..... Late last year it was some stess errors (as example Apple and applE) but for a while I even don't need to correct it.

[upbeat pollen]

I can not upload a song longer than 30 secs why?? Please help

[manic cove]

If it helps, I did something kinda like that https://suno.com/s/OatpyMabZhGFS7DN

[clear thunder]

That’s a good song but not quite. This vocal performance is nowhere in 4.5 https://suno.com/song/32bccf01-038e-4028-a6ce-e2cc4bb992a8

[sleek stump]

Every language there are words that are mispronounced and must be spelled out phonetically, quite a few actually in English. This is to be expected though, AI LLMs dont have actual reasoning or abstract thought.

[small bridge]

anyone else having issues today Its eating my credits and isn't making songs.

[cloud merlin]

yeah it's not working right now for some reason lol

[sudden fog]

Are the servers struggling at the moment?

[small bridge]

i think so

[polar spade]

If i understood your post, why dont you sing it, poorly if necessary, but hit the notes as best possible. Suno will hit most all the melody and follow your audio with the help of the posted lyrics. Just upload, chose cover, slide audio and styles to the right, but only post lyrics, styles box is blank. Hit Create.

[sudden fog]

I got this message

[naive ibex]

⚠️ eating credits and it not generate anything at the moment.... (it could be OK, if it reject "create" and return credits, but it just drain points with no errors)

[polar spade]

Yes? System may be a bit loaded?

[sudden fog]

I haven't gotten this message before and I normally create around this time

[polar spade]

OK. Other have had same error tonight. Apparently its a bit busy.

[sudden fog]

Ok with all the hype about suno it doesn't surprise me. I'm a pro account is it worth upgrading?

[sleek stump]

The only difference between Pro and Premier is the number of credits you get per month.

[sudden fog]

Ok but your still able to make music during busy times? It's completely block me out

[sleek stump]

Only free user accounts are throttled during heavy load times

[polar spade]

OK. Guess I should remove my post then to save any confusion. 😄

[sudden fog]

That's weird I paid for a year of pro

[polar spade]

I see nothing, i know nothing. 🙈 🙊

[sleek stump]

Pro is 2500 credits vs Premier 10000. All of the features are shared between the 2 tiers.

[sudden fog]

Ok but i shouldn't have that error right sorry for all the questions

[sleek stump]

I am a Pro account and I have not had any issues. I just did a number of gens for a track I am working on.

I had a problem similar a few weeks ago and it turned out to be a corrupted cookie. Log out of Suno, clear your cookies and browser cache and log back in... might fix your issue

[sudden fog]

Ok I'll give that a go thank you

[sudden fog]

Thank its fixed now

[main gulch]

That's great👍

[fair mantle]

Is there anyone today from the Support Team here?

[naive ibex]

I assume that it mostly and purely community driven...... (not official support even at day time, and you are asking at night 🤷‍♂️ )

[vale owl]

Unfortunately, that didn't help either.

[fair mantle]

Yeah but the Officially Support dosent react. I was losing a lot of money because I have no Accsess to my account since a week now. Fortunateley I could compense some if these lost money because I had open accsessed public some of the songs I created for a Webseries.An entire Team is depending on the songs which I not backuped yet and is still not public to find. My time runs out and the support literally dosent answer to any [e-mail] email.

[fair mantle]

Same here it dosent worked.

[spice cradle]

IS it possible to let it just generate 30seconds of audio rathter it deciding the length?

[naive ibex]

That's it man. Waaaaaaaaaaaait.... "Surprised" as well that someone cannot make normal communication....

I would aware you buddy, if you rely on it to do commercial work, they can change without notice 4.5 model as a random update drop by overwriting existing one, and style and personas voice could be changed (literally that bad as woman "persona" became a men).... So be awere about reliability and consistancy at least for 4.5 "beta".

[thorn sparrow]

that is not the melody i meant

[fair mantle]

Good to know thank you but this is not the problem the problem is that I dont can Login in my account anymore.

This is my problem. (normally I save all songs as an Update unfortunateley I created new ones wich are not saved yet. ANd I need them as soon as possible)

[woeful ingot]

Lately my songs have been either abruptly ending at the 5 minute mark or fading out, with more lyrics still left to go.

[sage jetty]

suno should create an option with time stamps in the lyrics so you can create longer songs;

[ember gyro]

the first wont work

[polar spade]

The chorus of Torn Between Two Lovers by Mary McGregor uses that chord progression (and well as hundreds of others) she's not quite prog rock as your example, but hey. 😁

[polar spade]

If you move your post to the need-help channel. https://discord.com/channels/1069381916492562582/1372994553233276970 you'll have a better chance of having a mod at least discuss the technical issue.

A mod is presently in the room as I type. This weekend , (which starts tonight) there will be no response until again Monday.

[polar spade]

All song lengths are affected by every bit we use: prompts, lyrics, tags, genre, and adding personas, cover, and now any attached audio.

The combinations of which can be thousands, the song lengths results can be thought of as random. I get my best 7-8 minutes outputs by throwing some random prompts and no lyrics, but without expecting any good results.

I'd say there is no good answer as to "why" you get what you get. Other than saying ,if it cuts off suddenly, that seems to indicate its processing a lot of information which it needs as seed info which leaves less "space" to create an output. Think of it as the output is being laid from beginning to end and the seed is being welded into the output from right to left. Where the two meet is where our output suddenly gets chopped off. 3:14 was the infamous timemark for many cover related outputs for 4.0 with a 4 minute limit.

[polar spade]

To this day, the work around has always been to do smaller increments of extensions until you have the length and complete song you wish, disregarding credit use as a priority.

[polar spade]

You would likely get more hallucinations in between where you wish Suno to fill in and the actual spaces Suno decides would be appropriate. No more than we can set one time length of a song, could we set micro limits within the song and it work flawlessly, is my thought.

[polar spade]

With experimenting, we can already get some cool transitions using extensions and set our own time marks.

[quick crest]

is this a recent change or something? i was able to audio as a persona earlier this month but now i cant do anything. trying my own instrumentals, nothing with vocals or copyrighted

[polar spade]

May be. Presently we're having elevated usage as well. So errors might be random.

[uncut pumice]

It's upsetting I can't use the service I pay for.

[fathom gulch]

how is this still broken

[mortal phoenix]

firstime in 5 months having elevated usage error cant create any songs when will it be fixed

Please

[polar spade]

Things seem to be already working better now on my end.

[polar spade]

I just uploaded an mp3, covered and added a persona and it created a 4.5 output..... however it didn't seem to pull from the persona did again, and the persona with Cover of an upload worked.

[quick crest]

still showing personas disabled here :c

[pseudo blade]

i cant see my latest song in the librarby now

[fathom gulch]

"You've used all your credits" No, you ate my credits... with no support or resolution in sight. How very disappointing

[main gulch]

oh coco, that's my problem too, I hope Suno finds the error, Mr. Dasvon?😁

[worn thistle]

hello, little question, it's possible to upload vocals and make them something like "ctrl+c" "ctrl+v" but add only beat?

[pseudo blade]

haha :00 i hope it gets fixed

[main gulch]

Hello Dasvon and VRKing, please can you help Coco and me?

[polar spade]

Its been fidgety the past couple of days for me too. I have ust click reload, or go out and back in. Eventually the library loads for me.

If not, check your workspace, see if it loads any faster.

[cinder rose]

Hello, sorry to interupt, but i have some questions about the distribution, wonder if you guys can help me out. Can i ask you for some personal opinions on how to easily distribute your songs made with suno? what distribution program do you use? or you do it manually? I recently payed for Distrokid license, but i heard there are some legal issues when posting music made with suno through Distro. Any better and legally OK method to easily post music on platforms like spotify etc.? thx a lot.

[main gulch]

In workspace it's all ok, only library has a problem

[polar spade]

Library and workspace are exactly same. But workspace filters stems and other files by default. That can be handy until Library works...... and Library may just be more laggy than workspace due to system busy or updates or patches or...... at the moment, I'll assume.

[main gulch]

ok, I can live with the broken library👍😁

[main gulch]

Can you see your song in workspace?

[pseudo blade]

i can t

[polar spade]

What knowledge I have is from somewhat static and hopefully reliable sources. So I read all the sites Terms of Service, I read sites from entertainment law offices specifically regarding distribution and rights/copyrights. Watch some YouTube video i can refer to again and again. Keep up with any news feeds. And ask the source, i ask Distrokid, I ask ASCAP and BMI, i check with songwriter association member and coordinators. I've done this over 20 years, so I know enough to get me in trouble if I give suggestions to others. Haha

Anyone here, I think, would be irresponsible to suggest anything in a chat channel that we could not back up with docs and proof and definitely could not continuing answering the mass of follow up questions that would ensue. It would pretty much be giving legal advice.

Spend several hours at minimum to ... let say, watch some videos... that would be a good start.... they are easily found by searching distrokid, copyrights, how do I monetize my song, how to distribute songs, just anything you wish to find an answer for.

That would be the best advice I could give.

[pseudo blade]

i still cant see my latest song which is this -> https://suno.com/s/0dlp577oHlcSmRFf
I already pinned it before so i can just see it via the main page

[polar spade]

Your latest song may also be found in your history.

[pseudo blade]

history doesnt show up for me either

[polar spade]

Youve check workspace list and turned off every filter?

[pseudo blade]

yes

[polar spade]

It exists, of course. So just need to find out where or why. Do you need to work on it for any outisde matter, you can copy it and keep working.

Or just want to make sure its "safe" because you have lots of likes and listens? In that case, just keep looking for it. You're now up to speed with anyone else's suggestions.

[unborn star]

Have anyone tried creating background music

[pseudo blade]

no, i just wanted to make sure it exists

[unborn star]

I made some tests and I got something like this, which is amazing. xD

Didn't know that Suno could create something like this

It could be straight from a game

Just investigate what's possible, anyone here with similiar results?

[polar spade]

The things that can be done are massive. So, generally the answer is yes. All subjective too. So, the things I like may not impress you. But everyone has their favourite things and ways to have fun with Suno.

[unborn star]

Want to listen to similiar pieces to take a look how people prompt music like that

Not really on this Level

[polar spade]

Me personally? Thanks, but not at the moment.

[unborn star]

You may misinterpret me

I just looking for people who tried to create game music.

with Suno

and how they prompt them

that's all

[polar spade]

No, mate. Im sure you got some banger stuff. Yeah. Look for gaming sync music creators . Im sure there are hundreds in here alone.

[unborn star]

You may work with [Morph] or [Fusion] Tag to create something you looking for, but there are some rules

I don't know of any tutorial, tried and learned it myself, so sry

[polar spade]

Those promos they showed? They may still be out there .... but start with the Announcement or update page and scroll up or search until you find those specific updates.