#📚┃suno-school
1 messages · Page 98 of 1
If I put [sfx:horn] in my lyrics, would that work? Does that format even work?
experiment
Yes, I tend to walk the song to the fx and include them in the style
You feel knocked down for suggesting that I am doing something wrong. Nice.
You said, "Some of us actually know what we are doing." This denotes that those not included in US do not know what we are doing.
You’re not really supposed to upload other people’s content to Suno 🤷♂️
[Machine Sound FX: (whiiiirrr-chk-chk-chk)]
[Intro] [Tempo Change] [Atmospheric]
System boot, the kernel wakes
Root permissions, no mistakes
Voltage streams as data ticks
Activate spine: (pssshhhtttt_sszzt_cssk_pssshhhsss)
(hssssshhhhhhhhh)
Well is that just gonna be vocalized?
Did you listen to the song?
@naive quail has done a lot of sound effect experiments — I know he’s said “beep” is reliably promotable, but the vast majority of others seem to be hit or miss.
not yet
I don't have headphones rn. At work
The song is where I used those lyrics
You can hear it for yourself. It is all about inference. You have to lead the AI to where it would make such a noise. It is weird but it works oddly well in v3.5 and I believe even better in 3
I know [gunshots] works decently well
[Extra Verse] [Tempo Change] [Syncopated Rhythm]
[Female Vocals]
Machines drum on, click_click_clack, Electric hums, power on...CRRRACK!
Sudo command password input: (chk_ka_tck_ah_tic_chuck_cack)
Since airhorns are a relatively common staple in EDM, I wouldn’t be surprised if it’s easier to get an airhorn than it is to get a dog or something (which is still possible)
Anyone know how circumvent the clicking sound in tracks
Both ways work, it is about getting it to know that sound effects should be there. If that makes sense.
What did you use to get a deep voice before. I've never flat out gotten a deep baritone or bass by directly prompting for it. That's cool if you got before. What did you use as a tag?
barking dogs is tough. Took me like a dozen gens to get one.
I second that. The bass level of a quartet alone I have not gotten before
I just get lazy and make the AI vocalize it while also doing [barking]
either Deep Male Vocals, or Baritone Vocals. I'll add descriptors before that, but today I'm getting ladies with that.
"Yo... like, Ruff Ruff!"
I've pulled out the vocal and lowered the pitch then after extending it slowly raised the pitch back up
Yeah, not something that I would expect to be SUPER prevalent in the training data, while maybe nature stuff or some hip hop songs might include it.
The fact that you could get it through enough dedication should say something about the possibility of getting an airhorn.
Mind you, the only barking I’ve ever generated was in a song ABOUT a dog, and it wasn’t prompted for 😛
More like
[Barking]
[Barking]
(ruff ruff]
Like [Barking: (grrr...ruff...ruff..ruff)]
Weirdly, Bass Vocals has never worked for me.
It hasn't worked for anyone I have heard
I used
[*dog bark*]
[*bark*]```
But most gens failed. This one got it (first few seconds): <https://suno.com/song/8a3a40b7-05e1-4655-9670-d1732a55b5bb>
Barbershop quartet is a start but I can't get the bass vocals out, even a Cappella
Speaking of bass vocals, anyone try turning dubstep into vocalizations?
You mean like the growls?
You've been lucky then. I , in 8 months, haven't gotten a baritone unless it just appears in a song somewhere as in African chant or barbershop. I thought maybe a special prompt. Yea, I've tried deep and Baritone, with no results, darn it. Ok thanks for the response.
A capella dubstep
That's what I'm talking about. Jacksfilms dubstep solves everything
Same. Never, not without uploading audio even then to hear it pitch back up
My new prompt one is "isicathamiya" . Not just an enjoyable sound, but once I get a good bass out of it, I'm gonna persona it and hope to capture a nice sound. I wait.....
Oh... isicathamiya is the fun African chant style similar to Lady Smith Black Mamba
What in the hell is that? Lol
The word is beautiful...haha. it read like it is fun
Like, how do you even write out dubstep as lyrics?
Call and response African chant (often a cappella)
Lol
Cool... I need a drop bass in lyric style. Go!
How do I stop it from giving me 7-second replaces? Waste of damned credits.
I shall try. Why not...lol.
i can't finish songs on 4.0. If it's a first generation and it finishes before 4 minutes, it's ok. If I have to extend to change something or add the ending, that's the end. It never finishes for me. I always have 2 minutes of endless music. What the hell do I do? [end] doesn't work
Just funin'.... if boom is a bass.. drop bass is... hmm... ooooomb!
Does [Fade Out] work?
No problem. If I can help, I will.
asking once more, as I can't seem to get rid of the sound: Anyone know how circumvent the clicking sound in tracks?
Sometimes. Depends on your luck
It tend to work better followed by [End]
Or [Instrumental Finish] [Fade-out] [End]
If you're talking about the "shimmering" effect, as the community has been referring to it, the devs are aware of it & are working on it. Expect updates in the coming weeks #📢┃announcements message
It seems like the issue is caused by something in the encoding process, so separating the stems and reassembling the song externally seems to remove some of the v4-accentuated shimmer, but to my ears, still seems to persist. I'm looking forward to whatever the team cooks up for us
I managed to get one nice baritone vocal, instantly made it a persona to use for later songs 😂 It's like one in a million chance getting a decent deep vocal. That's why I think the "Persona" feature is so underrated. Proper life (or song) saver 🙏
@polar spade DM me if you want to talk about this song direction. Because this seems like an actual direction of stuff.
Oh, I'm bout to make magic happen...lol.
If i make another account of Suno and get a membership an admin can transfer my credits to the new account? because my old account have the Fake App problem but the same doesn't happen on a new Suno account
Hi, guys! First: I was here with another account, but I add this to the main one, not to switch every time. I'm having a problem I don't understand: I should be able to upload a file of my voice or another voice to make a persona or a song with a similar voice (that's what I knew), but it doesn't work. It tells me "You can't create a persona from audio files at this time". What does it means? I red in the FAQ that it was possible. Did someone have the same problem?
Thanks with anticipation for helping me.
If you are. Please stop pulling me out of your hat. I have better places to be besides in that musty hat
I'll try but let's just agree that you are a classic...
True. Also I get to chill out with the doves.
Your old account was also a Suno account made on Suno itself?
maybe try deleting your cookies and sign back in
Yes, both are original suno website accounts
Only Suno support will be able to do anything with your credits. As long as they are both on Suno I don't see why they couldn't make the transfer. I never read the agreement tbh
Yes, thats why im trying to get because i paid for the membership and i cant use it because my account have 2 weeks or more with that problem
Ok we are users like you, support@suno.com is your best bet
Does V4 cut off att 3:14 for anyone else?
It does at times
I am not having that issue atm. I just made two songs that went to 3:58 and 4 minutes
One-shot
Heya, not used suno in a little bit, i want to ask if i'm able to set BPM, and if i can set it with a C key and a minor mode while setting time signatures or is that still a bit too out there right now
Not quite something that can be reliably prompted for, but using an audio upload with those qualities would help ensure that any extensions from that clip follow suit -- easy way to dictate the key, at least
Hi. I want to create similar songs to an already created one/uploaded song.
I don't want to sound exactly like that, but I want the style to be remembered. For example when I upload a rap instrumental and I want to create similar instrumentals to that, but not with the exact same beats and sounds. Is there a way to do that? Thanks in advance.
Oh I've never actually used the audio upload, just cause I'm not sure what it actually does
I have a song thats in the style, key bpm and time signature that I want would i just clip that and upload it?
do we know if the sub genre ''void of phonk'' is there?
If you record / upload a clip, it gets added to your library, then you can interact with it like other Suno generations (some restrictions apply if vocals are detected)
I'm just putting together a short example video to demonstrate 🙂
A video would be great.
Exactly -- with the caveat that we're only allowed to upload content that we've made ourselves.
If there's a strong enough musical context, Suno can "hook onto it" while extending and continue the track, but I'm putting together a demo of what can be done with a simple clip and the cover feature 🙂
this is what i do to get inspiration for what im working on, export 16 bars and the generate ideas to help continue building out the track
ahh right yeah I didn't make the song I just want the bpm time sig and key xD I'll see what i can do with prompts in the genre then
prompts dont work any more correctly in 3.5v
then i'll use v3
** they work differently
???
The FAQs might be vague. It states you can use a persona when creating a Custom song.( add in lyrics, use you're own styles, etc) As [Upload Audio] is on thee same create page, regardless of using Custome mode, instrumental or other, that seems to imply you can use it with persona.
If it's written differently, post it up so they can look into it.
But sadly, no. Likely as a precaution to avoid that urge by some to upload Trent Reznors voice instead of our own voice.
Pro tip: Anytime you upload an audio it goes into that slot as a new song. So though you can also see [Audio Upload] on an extension page as well, doing so will just start a new song.
It seems odd. That the prompt, that I've created months ago has always worked! Till lately.
#🪲┃bugs-web message It's a different model -- v3.5 was based on v3 so there might have been a bit more similarity between the two, but even then you had to change the way you prompted to get the best results.
I have always ever been using 3.5V. is what I mean. But since 4V has come. Its been messed.
I'm relatively comfortable with prompts relating to indie math, so I understand some of it just being intuitively felt, but I'm personally not experiencing anything outside of the usual variations that I'm used to getting.
Not trying to be dismissive of what you're experiencing, just throwing my personal opinion into the void just the same
I did a gothic/dark fantasy 80s sound prompt. (I dont want to share it as its my own custom) and its been working fine!!!
till 4V..came. and am free user all I have is 3.5
see this the sound I made...I love it..but one used for these dont work anymore.
Anyone else having issues with v4 extending songs? All my extended parts are finishes at exactly 2 minutes and ends abruptly... All of them.
The max limit for extensions is 2m, unless it's the first extension off an audio upload.
In order to make longer songs, you click ... > Create > Get Whole Song from the extension to add the complete song to your library
Well yeah, but there is no in betwee, they all get 2 min, not less and end abruptly. I can't keep extending songs 2 mins without a proper end..
guys can anybody help with this?
To clarify, it worked properly in 3.5, but v4 makes all the extensions 2 minutes even when using tags like [end] and [outro] or any minimal lyric.
I gotcha -- I'm getting similar results, using simple style prompts without anything else to influence style, end, fade-out, outro, etc.
14 different attempts, and beyond a 33s track, a 1m20s track, the rest were 2m (with a couple that managed 2m03
)
If you're on a paid membership, you could create a persona from this track -- that's what the feature is intended to do.
If it's an uploaded track that contains vocals, you won't be able to make a persona out of it, though.
I have had luck with (Hashtag) (Hashtag) (Hashtag) Final Flourish:
(Use the hashtag symbol, you can't put it here, or else Woody bites your head off. ^^)
I think it might just be hashtags with trailing whitespace that Woodhouse gets angry at? (#)
what does that do?
It ends the song at this point when all the lyrics have been processed.
This is one of the extensions where I used FF. It ended at 01:21 as needed. 🙂
https://suno.com/song/0a572fca-f0be-4f5d-894d-d858b6e44cf5
Don't know. I've tried several times but Woody kept deleting the post. ^^
Is there a good (and ethical) way to get Suno to capture the style of a specific song (not trying to "copy" the song, just try and get a certain feel)? I understand for copyright reasons it won't let me include band names or song titles in the Style field.
Hi I´m pro member and Suno won't let me download .WAV file in v.4.0 , Anyone else with this problem? how to fix this?
Seems to work for me. Have you tried a different browser?
Yes. I tried in Chrome, Mozilla, Safari, Windows and Mac. 😫
Weird, then I don't know, sorry.
Have you tried another device?
yes, I tried in other pc´s and iphone app, and nothing, I can't get it
When you did it on your phone were you connected to the same network or actual cell data?
Use your phone and the browser on your phone not connected to the same network.
yes, I tried in same wifi net, and others mobile data
And you are able to download mp3s is that correct?
yes, mp3 its Ok, I can download it.
Yea someone else had that exact problem yesterday or the day before.
I do not know what came of it but it is in this channel
I had same. 3 hashtags and you're out.
Here is one #📚┃suno-school message
ok, I'll look in the search box to see what I find, thank you
me too
is there a preferred way to end a v4 song that doesn't want to end by itself? I used to use [song ends] or [end], but those don't seem to be working anymore
Have this been asked? What are the difference between Persona and Cover?
Sorry for late answering (work), I understand, but I don't understand HOW can I integrate my voice with the songs, as I read here: "While currently, we cannot create personas from uploaded audio directly, users can upload their vocal samples and integrate them into their projects.". I explain: if I upload an audio with my voice, the song that comes out is EXACTLY what is in the file, it doesn't let me create a song with other lyrics, integrating that audio part or something. I see that my friends can, if they use an instrumental base, on the contrary. Even if I just wanted to "arrange" better a complete song I made, it doesn't take my file and... I don't know, integrates with instruments and other, the song that comes out is just the exact copy of the file. O_o
Try this one. 🙂
I have had luck with (Hashtag) (Hashtag) (Hashtag) Final Flourish:
(Use the hashtag symbol, you can't put it here, or else Woody bites your head off. ^^)
we don't want to rouse the old man 😉
I'll try that, thanks
I am a bit surprised that moderation let that track go trough https://suno.com/song/cc73a5e7-6deb-4706-be9a-9fbf29b55aea any idea why? Anyway was just a test but did turned out quite good !! What do you do usually if you want to use some words that are not allowed, do you just try to use words that sound as ... ?
trap, aggressive and raw, heavy basslines, sharp hi-hats, intense and direct flow, gritty lyrics with street energy, unapologetic delivery song. Listen and make your own with Suno.
is it just me or do other people notice this, i feel since i heard the shimmer once i hear it more often now
I'm sad. Adding [No Shimmer] seemed to increase shimmer.
I dont know don't seems to have shimmer in mine but maybe i just don'T notice ?
Enjoy some shimmer.
https://suno.com/song/835d6b70-2191-49e8-88c0-923464914bc1
Tempo: 80bpm
Experimental hybrid of crushing death metal riffs, and clean Baroque elements
Acoustic guitar break
Egyptian violins
Fiend Choir
Massive, dominating basslines
Aggressive Deathcore Rhythms song. Listen and make your own with Suno.
I see yeah
Also in the track i just linked i got a producer tag at the beginning, i get Frank Patrick. What do you guys think? Is that like with training that this tag went in?
I feel like I am ordering a burger and it doesn't matter if I say [No Pickles]. They're gonna cram it full of shimmer. Er, pickles.
I don't see the producer tag.
Look my link up it's at the start Frank Patrick
I mean in the audio not in the lyrics
there is a producer tag added by SUno
sorry if i not explained correctly at first
Ah, I went right over it.
did look online don't find any Frank Patrick beatmaker 😉
Persona is a style/vocalist that you can apply to any new song you make. Cover is having Suno make another version of an existing song.
That’s not how Suno works — “Exclude Styles” is how Suno understands negative prompting.
But it doesn’t recognize the shimmering as an instrument it’s trying to add, this is an encoding issue.
The shimmer remains even when I download the Wav Stem for the Instrumental track.
I'm just gonna have to wait until it's fixed. Throwing away credits as it is.
does the music we put on public are non commerciable ?
( on suno )
If you're a subscriber, you have the rights. As long as your wrote your own lyrics.
quick question, I thought pro subscribers get their first 200 songs made with Personas without being charged credits. I've been using a Persona and still been charged 65 credits so far, but most should have been free... has this problem happened to anyone else?
Seems like the 200 was a time limited trial during the early beta
hm the message is dated Oct 31st 2024 so 25 days ago... doesn't seem long enough ago to be over already
I did email support directly several hours ago so I'll wait a day or two for an answer
It's not a problem with your account -- they've removed the trial persona and cover songs.
any advice yall can give for keywords to get good beats for songs?
like, for example, im trying to make something in metal, what should be some keyword prompts i should put in?
oh... that was super fast haha not even 1 month beta period
maybe they'll change the message on the website and make it info more current
😅 yeah, it wasn't communicated anywhere, and the #📢┃announcements page still mention that the first 200 are free. I agree it's misleading
me too
Trying one at a time? It chokes when you do multiple.
Try in a another browser. Firefox should work fine
I tried firefox, chrome, opera, .. windows, mac, iphone app...no way 😦
What’s up with the Error Code 422 🤦🏼♂️🤦🏼♂️🤦🏼♂️
so put them public nobody can steal it ?
It's the same as any music, private or commercial. It can always be stolen.
there are instances where I hear it, but int he song it's in it makes sense
iOS 15 default
I went on a hunt for it, I knew it was a thing I heard a ton lol
all my co-workers got iphones lol
I went after the basics. it was my first guess lol
Please team just explain
This is the point
hmmm 😢
I mean that's my voice
do you have some insight about why they disabled ?
replace section is a mess I need like 10 tries until suno picks up that I want to change something as simple as to take out male vocals and replace it by a female voice.
Why are all default songs in v.4 set in a country style? I'm even having a difficult time to hear the song I'm writing in other styles. They all end up sounding like a country tune.
I'm still using 3.5 you can use that and remaster it if you want, but the shimmer is hideous in v4 currently
Entirely speculative, but my guess is just that it checks off some kind of box that they can point to in order to protect themselves legally.
It's easiest to assume it's related to music industry or copyright stuff, but I legitimately have no idea what kind of legal issues a vocal-cloning company like ElevenLab needs to contend with in comparison to a text-to-music company does -- we've heard the horror stories about vocal cloning before it was something that was available to consumers, so it might be a limitation to protect themselves from bad press or bad actors
I did switch to 3.5, and not getting too many varieties. Lots of country
what styles are you promting?
Furthermore, since people own the rights to the music they create when subscribed, we the creators are liable. anyone who ever had to deal with DCMA from vevo or so knows, this industry is a shark tank.
try meta tags i dont even use the style box anymore:
[Meta Tags]
[Genre: Club, Electro House]
[Style: Emotional, High-Energy, Raw]
[Mood: Hypnotic]
[Tempo: 128 BPM]
I found a way to get multiple voices to work as long as you don't re-use them within the same song
For like duets?
no I have a different narrator for each verse
Ohhh, I'm doing something similar with character POV songs. Any tips are greeted appreciated
I've been using the persona and it works sometimes lol
I sent ya a DM to my latest song, I'm going to test it on another to verify it works, but.. that worked this time properly
Perfect, ty
I'm going to be removing the prompts in a moment so look at it now if you want to see
So this goes in the Lyrics area, right?
yep on the top, like in html
Anyone know how to reliably get deep voiced male singer, rather than generic dude singer?
there is no default, you have to prompt it if that's what you want
Persona?
https://suno.com/song/29cd2976-07cc-4e84-94d2-0c314d6e9d0b something like this voice?
dark grunge distorted, clear male vocals, multiple male voices song. Listen and make your own with Suno.
Well I'm doing different styles, and country always seems to come back.
try giving a random title
Has anyone cracked the code for prompts which will tell Suno to build energy to a song? I'm struggling to find any prompts that Suno will recognize in this regard. Thanks!
https://suno.com/song/4247ef37-0c26-4fe0-bf01-ef948cf00eaa I was messing with covers and this came out ok in EDM. I don't know exactly what you're prompting though, so.. I don't know exactly how to help you (the original is a grunge song, by the way)
Changed the title, no change. I'm going through the list of genres. I'm on "Christmas Carol" and getting more country songs.
ae you tying to create chistmas music-type thing?
In 3.5/4.0, iirc, you need to repeat it a few times (with commas) for the increased weight.
Works pretty well from the get-go with 3.0 if you want to start off as an intro instrumental piece
I tried to get my voice learned on ace studio
damn that thing works
If only sumo and ace studio was like the exact same app natively with a plugin like splice, they would skyrocket overnight
i cropped a song, for 2 bars, i want to create a new song with the beat of that bars, it does work but how do i tell suno to extend from 0 from the intro and not try to snap it to the cropped extension?
You can't extend from time 0:00, but I'm not sure if that's what you mean by snap to cropped extension.
ok, so basically theres a bunch of songs that have good beats but only one can win, so I want to reuse the left overs in ne songs with new lyrics 😄 whats the best approach for that? splitting stems leaves traces of vocals inside, cropping doesnt start from 0 when you extend, im a bit lost 😅
you won't keep the beat in that case, in an way, unless you MAYBE persona it. you're required to start from at least 0:01 I thought about doing that to keep a style, but it just butchered it completely. so what I ended up doing, was extending it from after a bit and then now if I wanted to, I could crop it away. (I just intend to redo the song at this point)
Have a great song but lots of Hisssssss in the background. Any way to fix?
v4?
no way to fix that. read announcements
I would A: make it a persona, B: reuse the prompt and C: set it to 3.5 and HOPE it comes out the same
was mentioned.. Is that the Hissssss
That is what I have been trying
Alwasy seems to "Switch it up"
They really need a "Keep the S**t the same option 🙂
for musicality...
Yep, that sound
tire going flat sound
yea, try what I suggested above, I'm not able to promise anythign though
Will do... Gues I can also hope for the fix
Ahh, I think I have a guess at the problem. You mean "I find a nice beat and I crop it, but the song is still a Part 3 or whatever, I want the crop to be Part 0"? If that is the problem, the answer is actually easy. Instead of cropping the "part 3" crop the "get whole song" instead, of the same section.
UPDATE: Nope, that is not how crop works, all crops are Part 0. I could have sword that used to be different but I may just be sleepy.
As a general answer, you could Cover a song or a Crop of beats, you could Extend or Replace to generate new verses and keep the beats consistent, if you need a longer sample using a stem or remastered stem might help, there isn't a single answer.
I have a song that I want to extend, but It's in v3.5. Suno is not allowing me to extend it because I'm using the v4. I've remastered it but it's not sounding as loud and clear and catchy. Does anyone know how can I extend from the 3.5 version? Thank you in advance.
that should be an option now?
Is it actually preventing you from extending using another model, or just giving a warning label that you should use the same model it was made with to ensure the best quality?
I used what we talked about today and made a song that goes with the "shimmer" it is a one-shot generation. V4 listened well. I was able to direct each section. V4 would take it and give me versions of each section I instructed. I had it doing all kinds of stuff today. I had no idea...lol. v3.5 had the same abilities, I just never knew, but v4 follows chord progressions better from what I can tell.
https://suno.com/song/51782f1f-705d-47f6-84c2-4b2ee2a2b88a - it is a dubstep/darkstep Instrumental for those that dont want to waste their time...lol
Yeah, It's just a warning label. Actually I can extend it. I used a cover I made with that song that is sounding better to the subject. but thank you, guys!
Guys there is a big problem with acoustic music. There is fking noise that looks like the "stall" alert in planes wtff
tictictictictic driving me crazy
with v4
Has anybody tried a [voiceover] tag before? Does it work? At work thinking about a song that could use that tag if it works. But of course don't have time to type out the song to try.
Does Suno have a developer API?
Anyone found a way to more consistently keep "Replace section" at the same lenght as the part you select?
The "Keep same length" does not infact keep the length.
Hello, I'm making a Cover from a Full version song, 03:25.
I'm using v3.5. Half of the Outro is being cut off at 3:18.
How can I fix this?
So I’m trying to remaster one of my old songs and it’s blocking it due to one of the lyrics having “producer name”…… bloody how???
best you could do it email support and ask about it
It says "Personas from audio uploads are disable at this time" and I used my voice for a sample...why is this happening?
I dreaded to come here because I never could get a reply or any assistance. Happy Holidays.
So far. First the cover not changing the tracks, now the "Extend" feature, changes completely the voice of the singer. WTF are they doing?!
You extend from 15 seconds before it cuts off is one way to do it
even the Persona feature, worked for the first days, now it changes totally the voice wtf.
I mean, I can achieve better results with open source MangioRVC. are we serious? LOL
I believe it's a glitch.
bro, when you pay for something, glitch should not happen. even after reporting them in all possible ways.
But seriously, why are my uploads are disabled? It's my voice and my music years ago before AI music existed. Are they disabled for everyone on upload music?
bugs, on bugs, on glitches, on bugs, no idea.
nothing really works LOL
like you put "punk rock" and it pulls out ballads.
I'm not a developer of suno. It's just my guess.
this will be my last month of subscription
next goal, buying x4 4090 and coding something custom, and definitely better.
Yeah. I kinda getting turn off that my audio uploads are disabled to use as personas
the whole Persona and Cover feature aren't working properly. LOL
not to be surprised
you set a female voice, specificy that in the prompt, and BOOM, song with male singer. LMFAO
And no answers. No announces about it. No information whenever if it happens. They just keep us in the dark.
It’s because there’s vocals detected in the audio.
Yeah, that's what touches the nerves
When there's is any... Interesting. Thank for help.
basically getting a music generating model to generate with the same vocal footprint and retarget the vocal with mangiorvc won't be a harder task
I've found out what to do once I'll finish my fork of unreal engine. LOL
This is a user-run forum, and you gave all of 30 minutes for a response — maybe there weren’t any users on at the time who were able to help?
My powers been out for the last few hours, and I wanted to pop in to see if I could offer any help before bed, but it needs to be noted that if you’re not satisfied with the level of support the community is able to provide, you can email support@suno.com for assistance through their official support channel.
Seasons Greetings
Anyone having issue with like undesired backgound soun i feel like in all the song ive made on v4 i have a kind of clinging noise cant describe but its when guitar plays faster for some reason
It's probably what is being called "shimmering". Devs are aware and say they will fix the problem.
Basically all people have talked about recently — the team is working on it.
Okay thank you 🙂
anyone figure out a reliable way to get v4 to extend well? Seems it struggles a lot to sync an extension from where the prior section leaves off. It mangles the transition every single time
I plan to distribute the track on platforms like Spotify, YouTube, and other music services, as well as promote it for public listening so how i get lisence for this ?
v3.5 is no longer using my voice when I sing into the app or upload my vocals, and they're going nuts with the autotune. I have no idea how this has degraded so badly since the v4 release.
Unusable!
And extend no longer extends your song. It just makes two new songs with different vocals LOL.
Has any of you ever used itranslator, the translation bot for messages on this server? I, for my part, have rejected one."
Bug report: I made a soul song with V4, eveything fine. No edits, no extends etc. Then wanted to create cover versions in different genres - no way! Whatever I try, it's always a new soul version, not the chosen genre. Tried workarounds by remastering first, then covering, or using V3 for the cover. All that didn't work, too. When I wrote the original genres into into "Exclude Styles", it worked ... sometimes! That's definitley a bug, the Cover feature doesn't like the taste of my song: https://suno.com/song/46c896a7-3b23-4b5c-9a7c-cebd79e77d93
funk, Soul, R&B, male vocals song. Listen and make your own with Suno.
Anyone figured out some tricks to stop it from layering vocals, and applying reverb to them? Consistently
Is it possible to change a word in the lyrics of a song and otherwise regenerate the same track with only the one word changed? I tried cover but that didn't seem to do it.
You can!
Make a personality of the original song
Make a completely unrelated song using that personality
then turn THAT SONG into ANOTHER personality
and use the new personality to edit the original song
@torn spoke thanks I'll try
This so bad!
yeah, the longer the song goes on, the more reverb/overdub it makes
it's really annoying
It'll start off fine
and then 1 verse in it has completely ignored your instructions
what do you mean by using the new persona to edit the original song?
Remake the original song, keep everything the same, but use the persona you made from song 2
ah
It should stop it from fighting your attempts to change the lyrics
When remasterimg or making new songs with v4 i'm Hearing a strange sort of Vibrating sound in the background.
Is there any way to change just the chorus of a song without changing the rest of it?
It's a known bug... It has been called "shimmering effect"... Suno is aware of this and trying to fix it
Thank you very much
only gets, login, logout, what to do? For hours
希望大家會喜歡,請多多支持,點擊訂閱,點讚支持,大家的鼓勵是創作最大的動力。虛創夢想 - https://www.youtube.com/@cmtworks --稍微控制住了~再來要練動畫了!
Wong place
go to #🐣┃suno-showcase c:
Also, anybody else have an issue where the replace section tool is ignoring the "Make same length as selection" option? It's creating the entire song from my starting point and onward, regardless of the setting. :/
I can't wait until we get midi's from suno
I doubt that will happen
have hope

I remember someone once said the same about negative prompts and then suno actually added them
Suno only allow us to create personas from music generated with Suno. Did anyone managed to import some music and use it to create a persona after remastering, extending or covering it? Any ideas on how to avoid this limitation?
I started producing songs with version 4.0, the first pieces both regenerated and new were satisfactory, yesterday suddenly started generating songs with croaking defects, as if I were playing old ruined vinyl. I was surprised I tried to change style, regenerate, but the defect remained. This note of mine to warn you, maybe you already knew. i wrote to Suno this...
Nice, but I wanted the vocals, specifically 🥲
Better maybe never. Just learn to sign or use the voices Suno already offers.
*because the issues if someone cloned a voice without permission
HEHEHE
Hummm Commands block seems to not working very well. Lots of generations where the instructions where sang 😦
"command block"? Do you have an example?
Electronic artifacts
Everyone seems to agree – the artifact problem is real and impossible to ignore. That strange electronic clicking, scraping or "swish" sound is everywhere and seems directly tied to the track's complexity and BPM:
Higher BPM = More artifacts: fast songs or complex instrumental layers? To forget. The artifacts are so bad that they can make the track almost inaudible.
Slower tempo tracks: The problem is less pronounced here, but is still noticeable during remasters, especially if there is any complexity in the arrangement.
Yup. It sucks. They've been aware of it for a while - hopefully we'll get a fix soon... and one that we can retroactively apply to our existing shimmery tracks to "deshimmer" them.
the last bit is wishful thinking on my part. 
It's really annoying, I hope it gets fixed as soon as I can, because I won't stay a subscriber if they don't fix this bug
I have a feeling they're referring to [metatags]
or maybe that whole <song_details> thing
Why is Hebrew much worse in Suno 4, even when I try to transfer a song from version 3.5 to version 4? In version 3.5, there were very few issues compared to version 4.
https://www.suno.wiki/faq/metatags/
Metatags aren't able to 'command' the song in the same way that we program a computer, but they're able to 'nudge' the direction the song is taking, determined by what's used in the Style Prompts section.
The site I linked might be slightly outdated, but has tips on best practices to make sure metatags are used effectively, and to try to reduce instances where they're sang as lyrics.
https://discord.com/channels/1069381916492562582/1308617723328204880 There's a megathread opened for language issues concerning v4 🙏
has anyone created a singing avatar to sing their song? If so, what software do you use?
Suno down ?
its working for me
hello, I upgraded my plan, now I want to get Stems from my music. When I click on Get Stems, nothing happened. Just loading icon... whats the problem?
The stems get added as the most recent tracks in your library
I see its in the creating, but they will not finish.
loading icoin is still rotating, and never finish.
I take it you've refreshed the page / closed & reopened the browser; how long has it been stuck like this?
when I refresh page or login again. These Stems will dissapear
im waiting for hours, and still nothing
..few days ago. it was working
how thanks 🙂 I'll try to rework it then 🙂
I am creating songs in Brazilian Portuguese, so it may be a little more difficult than English
Does this happen for every song, or just for specific songs?
and could not find where to search for personas
only for specific song
So what is the high pitch sound when the vocals start any one eles getting it to in v4 and it's not in v3.5 https://suno.com/song/b50cbf72-1194-49a7-a6bf-8a919e5e7258
somber melancholic rock song. Listen and make your own with Suno.
Aka shimmering effect if Im wording that right
Please don't post the same thing in multiple threads -- you'll see many posts about this kind of noise in the #1308499132356362383 channel. The team is aware of it, and is working on a fix, as noted in #📢┃announcements
It might just be an issue with that particular song -- a common workaround for a number of issues is to use the Crop feature & select the entirety of the song -- this will add a duplicate to your playlist.
If you still have the same issue with that duplicated track, you'll need to email support@suno.com for assistance 🙏
OK I know where the problem is. Why I can't create stems from music created in v4 ? I can get stems only from music created in v3.5
Doesn't appear to be a system-wide issue, as I'm able to get the stems for v4 tracks.
Its not working for me. where the problem is?
on my side, it's an error on imported sounds: impossible to create an extension or a cover
Unfortunately, discord isn't an official support channel -- we can try to give advice, but the only way to ask Suno directly for answers is to email their support team -- support@suno.com
When I get stems from v3.5 track , I listen music in vocals. WTF
Being honest didn't see that
Appreciate your support !
how to restored trashed track ?
Go to Library, click trashbin symbol on top right. Choose the song you want restored. Right click on three dots and click "Restore to Library". 🙂
Little button in the library -- if you're on desktop, you can use typical desktop selection methods to select multiple tracks at once
Jinx! 😂
Thank you its work !! and i trying to use V4 but my netive laguage only work with 3.5 , vocal perffect work with only 3.5 and v4 totelly messy pronouce . do you have any suggection to fix this issue ?
https://discord.com/channels/1069381916492562582/1308617723328204880 You can report language issues with v4 in this megathread 🙏
Until any changes are made to the model or the language recognition, spelling words phonetically can help
I don't know if anyone can help with this, but when you crop an portion of an extended portion of a song the option to add that cropped part back into the full song dissapears, which is sad because with outros sometimes Suno will create two outros - and you can only use the first half of said outro in your song. Anyone else experiencing this?
If I understand, you'd like the ability to extend from a song, crop that extension, then use get whole song to stitch the cropped extension back onto the full song?
Exactly. So I have a song that needs a new outro.
I generate the new outro using "extend" and it makes a track that essentially has two outros stitched together (outro 1 plays, stops, outro 2 plays, then stops)
But I realized that the Outro 2 is a LOT better than outro 1 lol. Is there a way to just get that part back into the song?
Thank i just did and how i fix its by myself ? is there any way to do that ?
Ooof -- if both segments are in the same extension clip, it's monumentally easier to clip off the tail end than it is to "disappear" a section from the beginning.
Not impossible, but just a lot less straightforward, and either requires some 'luck of the dice', or to edit parts of the song externally.
Until the team reviews the feedback and hopefully optimizes some of the language-related issues, spelling words phonetically is the best way we have to workaround these issues. For instance, if it was pronouncing "matthew" as "matt hugh", you could spell it as "math-you", which would be pronounced the same way.
I was afraid of that being the case :c. Oh noooo. Oh well I made a post in the feedback channel so maybe they'll consider adding something like that in in the future.
thanks anyway!
At that point the crop is no longer an extension of the original. You have a new separate piece of music which now would need to be "spliced" onto the original and blended in as if it was an extension. So if thats the case, we could splice two songs together, or splice an extension from another song onto this song we're working on.
There is no splice tool yet set up to do that kind of work. Even if you cropped off the end or right side of the extension, Suno only sees that new output as a new "song".
Thank you ,, i did but still wired voice ... i think i will wait .. 3,5 work fine for now
Gotcha, it would be super cool if crop had an option that "inverted your selection" at least so you can tell it what to omit, and it then stiches your song back together without the part you selected. Oh well lol maybe in a future update we'll have something like that.
me and @light bloom have a fix for the shimmer check it out in v4 feedback
100% all genres
Yes. Really cool. I'm still working on their making a splice too to cut out middle parts -- which would then double as a crop tool for the end parts. And we wouldn't need the crop tool. Grrr... 😆 But AI is a different bear to deal with and maybe reasons why it can't be done.
And that link would be?
Is this staying inside Suno app and parameters -- for those of us who don't use DAW?
we dont use a daw
Cool. And that link would be?
Thanks brah
np
Cool idea. I'll pay attention as I'm working songs to see which version I'm in.
Just keep the same style prompts when covering, eh?
yuh what ever you wrote in 4 write in 3.5
extend will work as well
Mean concept and simple. You all dat and more. Thanks. I'll try some out.
Sunos content policy can be quite strict lol. It forbids me to create a song from a two hundred years old poem/ballad
Should muffled screams and distorted laughter be in brackets or parenthesis?
no, nothing
but I guess the content of the poem could be interpreted in a way that's controversial today. It's just funny as it's a well known poem
that's weird, never encountered any forbidden words
also it's just the "cover" feature that goes wrong
The original song was generated, I wanted to get variations of it
cover is great for instrumentals, not so great for lyrics in my experience
exact same prompt, but I get the content violation thing back
I didn't change anything. I just used cover to get tiny variations of the same song.
ouch, I really don't like hearing that. Never gotten content violation but I sure don't care for it
it's not a big issue. I understand that it's difficult to balance that
it just weird. Think about entering an Edgar Alan Poe poem as Songtext and it tells you the text is not appropriate lol
I write a bit of Viking poetry and I get content warnings in chatgpt when I do research.... so yeah, it's bananas. Mythology from 2000 years ago...
still, Suno never say anything
Hey guys, does Suno ever respond to support emails?
Yes
I just uploaded a small snippet I made and tried to extend it, but the result is WAY different than the uploaded clip. Is there any way to make extensions stick closer to the source material?
I find the getting stem function always leave very obvious voice residue in the instrumental track. I guess this has to do with how AI music generating process works, it probably don't generate singer voice and instruments seperately. But is this problem addressed by Suno team and will be fixed in the future?
Question about when I master a song. Both times I've tried it there's a weird chirruping noise almost like an old school Nokia ringing tone. It happens liberally throughout the song like the mastered version is picking up something and trying to enhance it but it really doesn't sound good
By the way, I found a trick to work around it, which is using Remaster on the instrumental track and getting a similar one without voice residue. But not ideal enough, it will create other extra strange fuzzy sounds
Having spent the weekend deep diving into V4 my initial gut reaction is fidelity on the vocal arrangement is a huge leap forward but the song composition and the meta instruction limitations have been dumbed, down somewhat. The edge of 3.5 has been lost and a somewhat blanding of melody construction (even after detailed meta promoting in song) anyone else found this????
Anyone els who can't get down to version 3.5 ?? every time i click it nothing happens?
Hey ded... haven't seen you in awhile.
I bank on the description of the action to be in brackets and the quality to be in parentheses, though parenthetical words often get sung or spoken as well. So lately I try to put 1 to 2 spaces between them. Anymore and it , the ( quality) gets sung.
Example would be:
[Spoken] (whisper)
I think we're close.
[Laugh] (distorted) "Bwah-ha-ha"
[Crowd: Female] [scream] (angry)"We want freedom, we want freedom"
But I have seen others using patterns such as:
[Muffled female scream]
I've not focused as much now as most times, I click several creates and it eventually gives me what I want from those styles I've given. So to say these work always on the first couple of tries is not true.
Is it stated somewhere that, whenever you cancel a subscription you will loose all credits?
tldr; had 1500 credits (from plan, not bought), cancelled plan, cycle refresh, 50 (free tier ones) left
Canceled or active accounts, credit are renewed each month. No credits rollover or are retained other than extra credits purchased or gifted.
Under subscriptions (at the bottom) in your account profile menu
Ah cool thanks Dasvon
I'm trying to create an interlude for something of mine and needs all kinds of sfx which is seeming pretty hard to get all of them to generate.
"You dam a river, you flood a valley"
Yea, it's hard to get every metatag to work throughout the song. Some work on one while others on another. I tend to concede to whatever combo makes the song work as a whole.
Yeah, that expire if you don't use model really took hold with all this AI crap.
@hybrid arch Bring back the old generation window. New one cannot continue the lyrics. It generate new lyrics from scratch.
Yeah this is my setup and don't think there's a way to make it all work.
Trying to get all this to work
[distant shot echos]
[trailing cop siren]
[Muffled screams]
[Haunting wind blowing]
[Voicover echos]
"Late one night..a shot rings out. A mortitiacan in a closed funeral home. FOUND dead, welcome to the murder mystery."
[typewritter keys clacking]
[White noise]
[heartbeat fades in and out]
[Somber piano fades in]
[distant, distorted laughter]
[Voiceover intent]
"Help me...find the killer"
[Piano becomes distorted like underwater]
Sorry about the wall text
You mean "generating lyrics" not "extending songs"?
Yes.
When I enter the lyrics part, like one verse and chorus, and want it to generate the next part.
It worked before. And now with all the model selector, it stopped. It just generates new lyrics from scratch.
I mean on site.
Ahh, I see what you mean. Even in the old model, you can no longer use the lyrics text area as input.
I understand. The "Write With Suno" popup text box is limited to 200 characters. Previously, you were able to have Suno look at all the text that was already in the lyrics, and if you already had lyrics, it could add more on to these. You can post this as a suggestion in suggestions channel, that it would be useful is Suno lyrics generation could have access to the existing lyrics, and not just the 200 character instructions, which is the way it used to work. You could even write "make these lyrics funnier" for example.
Sent. Copied your wording, if you don't mind?
@vivid wasp I don't think Suno recognized cohesive words like "in" and "and" and "the" and so on. I haven't seen differently. --- except for verbal phrases such as "fade in" or " fade out".
My thought is it treats each word as a style, of sorts, and any it can't recognize tends to neutralize the rest.
So try [typewriter] [clacking]
or. [Clacking typewriter] or ( ) can emit sound...
so. [Typwriter clacking] (tapping) or
[ clacking typewriter] (tap tap tap tap)
Or typewriter(tk)(tk)(tk)
Hi, I have a question about the V4. Will it be free? like after 6 months or it will be behind the paywall forever? and how will it be then? there will be updates only and only for V4? or it will expand to V5
So like [shot] [distant echo]
[Screams] [muffled]
Hi, I have a technical question. I want to create a persona from my uploaded audio , but I can't. It keeps saying "Personas from audio uploads are disabled at this time". Is there any way to turn it on or do something about it? I haven't been using the program for very long, so I don't know how it works yet.
Oh and has any one figured the right style prompt to exclude instrumental bg yet? It would help in creating interludes
Audio uploads containing vocals are restricted -- you're unable to make personas based on them.
How's the shimmer going? xD
I understand, thanks for clarifying.
Anybody master their Suno tracks after you've downloaded them? I'm curious what your tips are if you do! I generate solely instrumental music and notice the tracks sound more clear when I cut in the middle and boost the low and highs.
Has anyone else noticed a weird fluttering warble in all of their rock tracks with v4?
It seems to happen whether I'm generating a new song or trying to remaster an old one
This just changed my life.
Yep. It's referred to as "shimmer" and the devs are working on addressing it.
Ah, yeah that makes sense lol. I've wasted so many credits trying to do remasters and it seems to ruin almost every one of them.
I think I may stop using Suno for a while until they address these issues with v4. 3.5 is also having some oddities and the sound quality for some reason has gotten a lot worse.
i hear a crazy wobble in everything I make now. it's in every song. I only download wav's and its just in everything, no matter what version I am using.
im using EVENT Studio Precision monitors and despite their age, they are showing me everything.
https://suno.com/song/814e63e7-9416-43b8-ac54-d6788b6dc3d2 you just gotta make the song so destitute that you can't even hear the shimmer 🤣 🤣 🤣
its the "shimmer" has been discussed many times and noted by their devs, they are working on it.
as I was just mentioning...its in ALL tracks. DnB, Rock, Metal, doesn't matter
this is more than a shimmer man. this is like the audio is underwater or being played thru a fan
this isn't a glittery sound. its like its underwater
it's so weird, every time i play a song on Suno, my phone starts ringing, and when I pause the song my phone stops ringing, how does my phone know when Suno is playing
that's a feature not a bug.
just you wait, in 2 months, people will be ASKING FOR SHIMMER
you know the crazy bit is. I'm listening to songs I did a month ago on 3.5 and it was there. Using ableton now instead of Acid Pro has put all the noise and crap from these files up front. Ozone can't even get rid of them
!!! exactly! it was just more subtle, didn't sound like a rotory phone was superglued to your head tho
yeah its bad now. so I'm just going to try to master all the stuff I worked on prior to v4 in Ableton. I have a bunch of Waves and Izotope plug ins I figure that with Audacity I can clean this mess up a bit and salvage what I have.
Question for you peeps:
I got the perfect vocals but im not happy with the music, has anyone found a way to keep the vocals but swap out the music?
i did a remaster and it workd
the vox and music will stay in time, you then get the stems and piece it together in a DAW
i did it for 4 songs and it was perfect.
awesome i wasnt sure that would change the music altogether, ill keep at it then, thanks
it "shouldn't" but its AI so...it can. make sure all the prompts are the same and it should work fine.
if you do a cover its gonna change the music and maybe add or take things out.
yo, why are all my personas just repeating the exact same lyrics they're based off? Kinda defeats the purpose
yeah im already seeing some differeneces (not the best, but different 😛 )
I do the stem/join trick which does remove a lot of the shimmer. Not completely of course. Can't wait for this to get better 🙂
What's with the pronunciations now? Vocals can't pronounce "Belt of Orion" instead of "Uh-right-on" it says "Oh-ree-on" 😅 and there are other words pretty wrong.
If you generate the stems for the songs, the "shimmer" is reduced significantly. I make mostly instrumental music, and when I did this on tracks that the shimmer was really bad on, it became hardly noticeable. I think what's happening is Suno is trying to generate some type of ambient sound, it's just not fading correctly.
So I somehow, while extending a song for the second part got a great result, but it didn't come out as an extension T_T
Is there a way to piece it together in Suno or am I gonna have to download both parts to piece it together myself?
Not at the moment, but it's a great idea for a feature, can post in suggestions. Glue/Fuse/Merge/Join/Blend/Anti-Crop/Piece-It-Together
That makes me sad lol I have this aria, and it came out amazing, but now Im worried it won't sound right with me having to piece it together.
I suppose I could piece it together and remaster it with the new feature after I reupload it.
There are problems with replace and extend too
you will hear a huge volume and tempo difference, it would be nice to notice users
dont know what happened but suno creating very short songs today - even using same setting as earlier i get 1/2 - 1/4 parts of songs, like 1 min, 2 min long, cutted half lyrics. very, very annoying, and just using credits without getting what i got a week ago if i compare. sb have idea what to do with it?
3,5 and 4 doing the same short, cutted things now
OK, I'm going to ask one more time (my past 2 attempts went unanswered). I cannot get Suno to follow my prompts in a duet between male and female voices. It either ignores them or swaps them. Does anyone have any suggestions on how to get it to actually listen? I'm sick of wasting credits trying to get it to come out.
I've got this:
[male vocals]
I’ve walked this path with shadows near,
Through every trial, you’ve been my guide.
But now, the silence grows too loud,
I can’t keep this truth inside.
[female vocals]
Your words are flames that light the dark,
A warmth I’ve never dared to seek.
Yet with each spark, my doubts take root,
Can I give what you truly need?
[Chorus; male and female vocals]
Oh, the echoes of the heart, they bind, they break,
A fragile thread we’re too afraid to take.
Between the light and shadow’s embrace,
We stand together, yet worlds away.
And it either makes the song all female vocals, or swaps them. I haven't gotten a single attempt that's correct.
I've had some success using [Harmony] but it's not given...
Like for the chorus part?
i had the same problem... i use this :
[Male singer] : Tu es mienne…
[Female singer] : Je me perds…
[Duet] : Dans les ombres, nos cœurs s'éclairent.```
and the result was
[Female singer] : Tu es mienne…
[male singer] : Je me perds…
[female singer] : Dans les ombres, nos cœurs s'éclairent.```
or after some try :
[Female singer] : Tu es mienne…
[male singer] : Je me perds…
[Duet] : Dans les ombres, nos cœurs s'éclairent.```
It just gives me stuff like this. I really don't understand why it can't follow the only direction I'm giving it. The song would be just what I want if it would just swap the vocals...
https://suno.com/song/69444c09-75e3-42c9-835a-26c64b5037fe
Operatic metal duet; male and female vocals; tender and heartfelt melody song. Listen and make your own with Suno.
Yeah, I could try telling it the opposite and seeing what happens. ^_^
i tried, it didn't swap XD
LOL great
Extending remastered songs are just pure bugged, right? Every single extension doesn't sound similar to the remastered track at all, like the extension itself having a different "seed" or something
I might have finally gotten something to work, sort of... by giving it "male vocals" in the music style. At least it starts with the male verse...
LOL and it only worked once. Sigh. I give up.
Yeah, it's really weird and unpredictable.
Are you looking for something like this @solemn lintel :
ethereal, love ballad, duet, cosmic, heartfelt, male vocals, song. Listen and make your own with Suno.
Give it a listen and see, I'd be happy to tell/show you how I got it.
The displayed lyrics won't show the how of it, I tend to edit those so the words flow easier for the reader lol
I mean, I'm getting exactly what I want, it's just giving me a female voice when I ask for male, and a male voice when I ask for female. I threw a bunch of covers at it (maybe 40??) and it finally gave me one with the proper genders, but the version itself sounds kind of bad, so I might try messing with extending it... Your song sounds really nice, by the way! So I guess I'd still love any pointers you've got!
Stolas and Blitzo! 
I just got really in detail each time I wanted the vocals to be the same, for the gender thing. As for consistency, when I needed to extend the song, I continued from part way through a verse rather than a whole new one and it only took a couple attempts for me.
Let me show you the original wording, one sec
Yas! 😄
[Verse 1 - Man like Bryce Pinkham as Stolas, theatrical, British accent, high pitch, regal, playful]
In the twilight’s gentle embrace,
I find a world where we belong.
Stars align, in your eyes I trace,
The echoes of an ancient song.
They question our love, they can't comprehend, in ignorance they roam,
But in each other's arms, hearts to defend, we find our home.
[Man like Brandon Rogers as Blitzo from Helluva Boss. Cocky, sarcastic, mid-range pitch, American accent]
Through the shadows, I sought the light,
Drawn to you, with love so pure, my guiding star.
In your presence, the world feels right,
In your wings, I would wander far.
A realm that doesn't understand, tries to tear us apart, to divide,
But in each other's embrace, we'll always abide.
[Pre-Chorus - Man like Bryce Pinkham as Stolas, theatrical, British accent, high pitch, regal, playful]
Together we rise, above earthly bounds,
In the symphony of the night.
With every heartbeat, love resounds,
In a world of doubt, we'll be the light.
[Both]
In this realm of stars, our love will shine,
A bond unbroken, forever divine.
With you, my love, eternity’s nigh,
Together, forever, we will fly.
[Chorus - Both]
We are eternal, beyond time and space,
In this dance of souls, we find our grace.
With every whisper, with every sigh,
Together, we ascend the sky.
Love's flame a symphony, melody untamed,
Together, unyielding, unashamed.
Sorry for the text wall lol
use tripple ` tags
Oh, I see. OK. More detail then. Hmm. I just kept hitting the nail with the same hammer.
@naive quail If you think THAT sounds like Stolas, check this out, one of him singing to a living follower, to hint at the real world mythology he's based on. The voice is almost exact, had to save it as a Persona lol:
[very emotional, resonant rich baritone male vocal with a charming accent], cosmic, ethereal, [love ballad], [metal] song. Listen and make your own with Suno.
@solemn lintel Duets are one of the more challenging things to get with Suno. The "easiest" :cough: way is to do it in sections, extending from just before the parts where Suno uses the wrong voice. In some cases, you may need to do an extension for each section... other times you may get a run of 3 or 4 sections in the correct voice, after which you start the extensions.
Suno seems to prefer starting with female voices, tho that's obviously not a hard rule. I usually only include "Female" in the style and leave the male tags to the lyrics. Remixing your track, I only tweaked the tags in the lyrics up to the second chorus, but Suno did a decent job getting that far: https://suno.com/song/9b99e603-05db-419d-8ef6-07603d53cdec
Operetta, Metal, Female Vocals, Duet, Harmonies, Harmonized, Tender , Heartfelt song. Listen and make your own with Suno.
Yeah, I'm having a hard time now getting the chorus to actually be both voices.
Getting consistent harmonies is mostly just dumb luck, unfortunately.
LOL yeah...
Welp, I'm almost out of my credits for the month (which refreshed on the 20th, sadly), so I might have to stop poking at this for a while. ^_^
But this definitely helps, thank you, all.
I've been messing with the extending option.
This sounds good, by the way
I wouldn't spend too much trying to get the harmonies. The moon and stars have to be in perfect alignment: https://suno.com/song/cd8dc22c-b192-4734-bf20-88ec0c939ab8
bright voice, political, hillarious, comedian, quirky, American, cynical, comedy, artistic, satirical, funny, polka, Jewish, whimsical, funny, energetic, ska, high energy, fast paced, hilarity song. Listen and make your own with Suno.
That's quite the goofy voices! How'd you get those?
I used my Plucky Diva persona 😄
I've bad mixed results with [full band]
Love it.
Playful Kazoo, Humor, Comedy, Cartoon Voices, Corny, Kazoo, Silly song. Listen and make your own with Suno.
That's the track I used to make the Plucky Diva, but that one itself was made with my Bacon Time persona, which... has language. https://suno.com/song/fb1ee1d9-e553-46d2-944f-0312cfa4b14b
LOL Bacon Time
I love the song. It even has Donald Duck in it.
Here's what I was finally be able to put together. Not perfect, but... I can work with it.
https://suno.com/song/13424276-6cb5-451f-a99a-ad79ba7bf143
male vocals; tender and heartfelt melody song. Listen and make your own with Suno.
It took generating a song, doing a cover, then extending 2 times. And using... like 600 credits? I lost count.
any news about the fix for v4?
It'll likely be in #📢┃announcements or #🚀┃updates whenever news comes out -- make sure turn on notifications for those channels so you're the first of your friends to know 💪
sounds about right for a duet 😄
I've had a few lucky one-shot duets (like the Astronaut one above), but all of the others needed extensions at some point to get the right voices.
Only creates songs with 3:14. And when we create other parts and ask to create just one, the sound is cut off, right at the cut where one part ends and another begins. Do you have any solution?
The 3m14s limit is a known bug at least regarding covers in v4 -- we're all waiting for a number of optimizations, as per the last #📢┃announcements post
Has anyone else noticed that the songs created with V4 are much quieter compared to previous versions?
I made a song and I like the style but the lyrics are ai generated. I want to copy the style to my own song
How do you do so
You could try getting stems, then doing a cover using your own lyrics. You might even be able to just do a cover, skipping the stem step.
Any ideas on how to get replace section right? I try to be as exact as possible but it seems that after a second or two the song gets weird for a second
Hard to explain
Any tips on how to specify the initial song length (without extending) ie adding something into in the lyric / style boxes - getting a lot of songs generated in the 2min range
Replace is always weird. Like everything else, it's hit or miss. I actually had luck with it the other day, but it took a dozen tries before it wasn't weird.
Yeah… it looks like it will be good and when it’s finished it’s not
I’ll try covering the song to see if I get it better
Have you tried turning the device off and on again?
if you got a $1 every time you've answered this same question, you probably would've earned enough to book a trip for two to the Caribbeans 😂 don't you wish you could pin your answers somewhere in here so you don't have to answer it over and over? Or wait, maybe you get paid $1 every time you answer, the world may never know 👀
I freaking love the crop feature. Sure, I could do that in my own editor... And I still am going to, because there's still no way to fade out the audio within suno. But this is a great first step to not having needless padding at the end of a song.
doesnt work lol
Yeah, the only problem with editing outside is that then you can't keep anything in a Suno playlist.
Does Sunk know what stutter house is?
Edit: suno keeping the original for the funny emotes
today im getting even 1 min long ones, just wasting credit... it doesnt care how long the lyrics...
this is what makes me disappointed most often, it just not care how long the lyrics, not even tries about 4 min, 1-2 min and done, who cares its just a piece, and when im extending it, ofc it always changng sth what should keep
but at all i dont understand, i didnt used it for a dew days, and now when going back to it, all generations are between 1 and 2 minutes, v4 and v3.5 too, even with longer lyrics
I have like a disturbance i almost every song in v4 like a buzzing sound on the background.... like it feels like every good song is a waist of my credits .....
That is the shimmering effect mentioned in their last announcement: #📢┃announcements message
Do you own the rights to this song?
then use the same prompt you used to generate it?
That probably means yes?
Ive tried
But other music with the prompt sound very different
I used "energetic synth" I think
make a persona with that song, use the persona when creating new songs.
How does that work
then songs you use with that, will try to keep the same vibe and vocal style
especially if you use the same prompting.
This sound inspired me, thanks boy
what style of music u Use?
Was having fun with v4, then it just stopped working for me; are there known limits?
I've got original lyrics, etc. Trying to figure out what rule or limit I breached.
Now custom song will not generate in v3.5 either
you can create a persona of it, that will do what it can to maintain the styling, as long as you use different lyrics (and not just REVISED, brand new)
seriously, what the hell with it today!? again just getting 1:06, 1:11, 1:47, 1:36 long song-pieces instead of a song
have no idea why the hell it doing this
makes me crazy why it doing this
does anyone know if you're supposed to use up all your credits? all my credits just disappeared from last months purchase
Yes credits from subscription reset at the end of the month, sorry
that's an odd system, free users get 50 free credits a day (if they use it up) but paid subscriptions get their credits taken from them and no free credits added to their subscription daily
so we're being punished for supporting
that's interesting
i personally don't like being forced to do things, and i feel it might be a bit manipulative due to the fact that users generally don't calculate their usage
i personally wont be purchasing further until this is sorted
** paid users get their entire monthly allowance given to them at once, plus a sizeable chunk on top (even more for premiere users)
you get 50 per day when your credits are empty
i havent been checking in for a few days, hows our shimmer going
both of these assume a user will use all their credits
it's not upon the user to calculate credit usage
as glittery as always
lol damn
This is true -- it states on https://suno.com/account that the credits don't roll over between months.
it's upon suno to also remind users of credit usage
but that's not the argument
I'm not arguing, just stating that the way it works is clearly laid out for those who want to read.
the argument being that suno need to improve their system as users who pay for the service aren't getting what they pay for unless they take a certain amount of 'uncalculable' time out of their day - @charred merlin i apologise, i'm not stating that anyone is arguing - i am stating that this is my argument
if anyone wants to defend suno, i respect that but i'm going to also build upon said argument
otherwise nothing will change
Not even really defending it -- I'd prefer it they stacked between months. But they're up-front with it.
the point i'm trying to make is, since it is the case that credits do not transfer over - it should be more than smallprint, there should be pop-ups, warning messages, and a lot more to reassure paying users that they do not need to 'work' to enjoy a product
There's not a lot of fine print on that page, tbf
i mean, it's about losing credits you pay for really
The only time I didn't ue mine u was when I got pushed int more days than I wanted, and otheriwse I've been into free credits many times
i like the system but not losing my credits, and i don't really want to have to calcualte how much i have to use per day and then actually go out my way to put in those hours
it seems like i'm paying and then working for them
I understand making the assumption because of how other services work, but it's not as if they make it difficult to find this information.
After you use all of the monthly credits, you go back to 50/day. So spend away
again it's working
you just clarified my point that users would need to go out their way to find it
that's not how a paid service should work with losing something you are paying for
i use a.i to automate most of my creative works, it would be counter-productive for me to have to go out my way to do excess work in calculating credit usage (i had to buy a lot of credits because i was over-spending small credits and they give you deals on monthly usage)
yeah it's literally at the bottom of all the UI
that's the only message users will ever see
if they see it
It's on the same page that you sign up on -- it's a minimal amount of reading, seeing as how the top part just says "contact us for help".
It's subjective, I suppose, so if that's "out of your way", I suppose you're allowed to think that.
again you're assuming users should read everything
Yeah, man.
that's not how consummerism works
I understand people don't read the entire ToS, but this is 2 sentences. I wanted to have some faith in mankind.
do we use a.i to waste time on pages?
again you're assuming people even scroll down
after selecting what they want
you're making a lot of assumptions about how consummers should act
we're just people wanting a service
we don't want all this nonsense
Okay -- have a good night, dude 🤘
no matter how you explain it
you keep defending things that are going to change exactly the way i explained it anyway
"Epic symphonic metal; operatic vocals; rhythmic, intense build; heavy instrumentation"
i'm just letting them know as someone who pays for their service, i wont anymore - and now it's a public complaint, that is all - you don't get paid for this
Though I think I modified it a little for that specific song? I took a long time and was able to create a persona that sounds like Nightwish, and then modified the style from there.
I added "tender and heartfelt melody"
If it helps, you can give direct feedback from this link on the Suno page, since it's just other users that read this technical-discussion forum.
I'm very happy it inspired you!
thanks, i also like sharing with the community so they can get a glimpse into something they might have missed themselves
a lot of people with add / ptsd / autism
i like to support those people
a little reminder goes a long way
when ReMi spits out a bunch of hacked together existing popular songs lyrics XD
People with ADD would have no issues blowing through all the credits per month. don't go scapegoating them
what you break? lol
@crude jetty
these were 2 snippets LMAO all from the same song
[Verse]
Fly high like paper get high like planes
[Chorus]
I like the way you work
Kid no diggity no doubt
lol
remi trying to get me busted
i have ADD, we're not herd of cattle
just funny when people try and defend its not trained of real stuff, like i love suno but be real
allow me to slow-clap for you, I DO TOO
so why are you taking away from people with ADD that struggle differently than you?
why do you need to be that special
because you're using that as a scapegoat for your inabilities, that's not a symptom, that's just you trying to lump everyone together to make yourself feel better
that's not what i'm doing but you've shown clearly what you're doing
you're clearly struggling to comprehend the message
and taking it personally
idk what yall are talking about but its annoying
you know nothing about me, don't assume anything
Speaking of ADD. A commedian named Mark Lowry says sometimes in his performances that he used to have ADHD when he was younger. But the hyperactive part was tiresome, so now he only has ADD. 🙂
i apologize that we share the similar mental ailments but you're distorting my message - i am stating that there are people with different mental disabilities that cause difficulties, i don't know why you would assume the difficulties would be shared but it is what it is - the point i am making is that mental health comes in different shapes and forms and it's not compressed to any individual, weather you want to use me to set an example for whatever you're trying to say - it doesn't change the fact that there will always be users with different health issues and i happened to use what i understood as an example to comparison (nothing more, nothing less) you decided to take that personal, my intent wasn't to attack people but to highlight difference
sorrty, I get very passionate when people try to blame their own shortcomings on something other than themselves
nobody was doing that, you're gaslighting me
here, i can even dm you the message i sent to suno to help clarify if it would stop you trying to gaslight me
otherwise i suggest you keep quiet
Now you're trying to tell me my opinion is wrong, that's typical
you're turning this into an argument
well yes, if i disagree with you that's correct
i think your opinion is wrong just as you think mine is wrong, nothing wrong with us not agreeing with each others opinions - that's reality
i think this is getting out of hand as you have successfully distorted my message
shrugs I have songs to make, and you are nto worth my time if you're unwilling to do anyhting other than accuse me of things you don't understand
i've already sent feedback and i'm relaying to the community on a public form
i don't know what you're expecting of me
you literally said people with ADD couldn't read #📚┃suno-school message
you're further gaslighting me
i never even said these things
you know you don't need to dig this hole, right?
oh your own words are now gaslighting?
you're free to quote me and even take this to dms if you want
this is an ai music forum
You would think that they would have left it with the private message they sent, rather than drag it into public. Didn't get the response they wanted to and went on a warpath. I have nothing more to say to them
still gaslighting
why is it difficult to have a discussion with someone without asssuming their intent @crude jetty
you've actually caught my attention now because your name is literally TruthDemon, you should be someone that seeks truth... and instead you have engaged with both assumption and emotion towards a message you misunderstood while trying to maintain assassinating my character, i really hope you can stall this behaviour and maybe keep it to dms (if you want to talk about me, i'll open them for you) but you're literally turning this into drama and i have to now respond to the nature of your being
manipulation isn't a form of finding truth
for things like this, this is why I use an external way to generate lyrics, and then I edit them myself. it does a better job than anything you can use onsite
🗿 5 songs a day until pop-up arrives
exactly
this is way for now - 150 songs in a month
there should (most likely will be) pop ups, email reminders, texts, data of usuage per day/week, the sooner the better as it's a rising platform
that's all that's missing from the foundation
i paid for 3 top ups and then decided to upgrade but then the month i upgraded, i didnt even get time to use suno as much and i felt pressured, and on top of that i started to forget nearer the end and then literally half the credits were gone - i knew what i signed up for at the start but by the end of the month i wouldn't have minded some reminders and i might have been able to spend a bit more or even determine my monthly usage
it's a work in progress
50 credits a day for free seems generous compared to other a.i applications
it's a ghost
this is technical discussion, i think it's okay?
🗿 ghost
are you meaning to put this in suggestions?
that's fair
that's not why I questioned it, it's ok 🙂
I promise those are not the lyrics I put into the song
I don't even know what it's suupposed to be saying..
I mean.. it's SO far gone, I don't have a clue how it got them. while the 2nd branch was fine
not how I wanted it, but the lyrics are fine
Hi take how many day to get commercial use license for 8 songs?
Hello, if you made songs during a subscription, you have all the rights granted to you (as mentioned in the TOS).
If you want Top Music Attorney's take on the Suno TOS, see #💬┃general-chat message
If you made it during the Free plan, sadly you cannot as Suno owns it. However, it is not unusual for Suno to grant you a retroactive license if you reach out to them via support@suno.com when you subscribe
These links from #1294423389515284552 may be helpful for you:
https://help.suno.com/en/articles/2425729
https://help.suno.com/en/articles/2425345
Unfortunately, no! Subscribing to a Pro or Premier plan does not give you retroactive commercial use licensing for the songs made with a free plan. ...
A Basic plan allows you to create songs for non-commercial use, while songs created with Pro and Premier plans include commercial use. Here's an...
anywho one else getting problems with 3.5 not using their prompts correctly?
people stay try diffent one to me, which makes no bloody sense!!! As that woudnt be want I wanted duh!
So, this is not my best song, but I did manage to get an overlapped vocal that sustained for a decent amount of time while another lyric sang. Not sure if it was luck or not https://suno.com/song/ab5f794c-96e6-45a1-af70-82d3ebfadedc - 00:30 sec where the chorus starts and third large line in that belts high.
I have been sticking to v4, but both have been more stubborn lately. Especially with tags.
Has any one been noticing old songs, made on 3.5V that you have created months ago (and ones your very please with, that met your standers) are now coming through distorted and at a much lower volume, and turning it up doesn’t help at all.
That without a shadow of a doubt, that you would bet your very life on, the audio has changed it self, from how you remembered it?
I dont own 4
am free I always been happy 3.5
Do you have an old, downloaded MP3 of a song still in your library? If so you can redownload the same song and compare the two files.
Try downloading the MP3 anyway in case it's something happening in the browser that is causing this
no.
also I send the wrong message
I ment to say that 3.5 is broken and no matter how reword my certain prompt I dont get what I want
suno killed 3.5
anyone else getting songs capped at 3:14?
Any tips for alternating vocals. Trying to get it where it switches back and forth between normal sad male vocal and angry 'demonic' vocal. If thats even a thing that suno can do. It did it once accidentally and I want to try and recreate that.
takes some tries. You can try and prompt the vocal style before the verse e.g. [Verse 1 - Male vocals singing]
[Verse 2 - Demonic screaming vocals]
If you're using cover or persona, those tend to be limited to 3:14. Unsure why. If it's intentional, I have requested for this to be documented as it is not obvious.
ah yeah it's persona, never noticed it before. Weird bug I think
hey question, if i create a track with good 20 sec of vocals but bad instrumentals can i grab the vocal stems from the song and upload it via the "upload audio feature"? to try to get better instrumentals with the same vocals.
or does it violate the big scary pop up text
def isn't a terms of use violation. Worth a try I think. You could try using the persona feature if you have a paid sub
Hey so i been using v4 for a bit and in one song i noticed this weird "rattling" or sounds like a soft sounds of sticks in a cup in the background its faint but when i noticed it i cant not un hear it. And its in every song i made. Changed lyrics still there, changed the the style of music still there, Even made this track (cropped it) with no style and random lyrics to see if it was still there. Am i going insane or do you also hear it ? Is this a v4 issue ?
If you own the rights to the content being uploaded, that's fine.
it's not intentional
Allright thanks, i do have the premier sub so i'll experiment around.
V4 has problems right now, read the announcements
ahh i guess its the shimmering effect
yeah thats the curse of the shimmer. If you get the stems, then put them back into one song in audacity, it seems to fix the shimmer most of the time
Interesting
I will have to save up my points till its fixt then. Glad its a known issue and been worked on 
you can use V3.5 still
what's the v4 way to end a song these days?
Cropping XD
end a song without an abrupt end*
try [music stop with style]
Alright, I solved my problem by generating at v3 which provided a full minute like everything Suno is doing nowadays with every conceivable exit prompt 😛 Fortunately, of that 1 minute, the first 2 seconds fade out long enough to trim appropriately. v4 for whatever reason is afraid of silence more often than not
[Fade Out 0.5] [Song End] usually works pretty good for me
Hm, I actually did [Fade Out] and [Song End] on separate lines. What's the 0.5?
Hi, Can I change female voice to male voice ?
Can't tell you exactly, some arbitrary timing that seems to work 🙂
Me: Soon:
[Intro 26.1]
...
[Verse 1 8,901.00001]
...
[Abritrary all the things ...
]
Thanks 🙂 I'll try that next time. Fortunately, Out of 9 generated minutes, I got the 2 seconds I needed 😉 https://suno.com/song/712019d8-0972-47fc-8947-b4c1b6b7cc69
anyone else messing with this Remi lyric bot they put in beta? its secretly dropping some nonsensible heat
You guys seriously need to fix your AI problem of trying to END a song - V4 has the same darn problem as 3.5 where it won't freaking end
are you tagging with [Outro] for the final verse?
I am so tired of having 80% of my stuff end up with what I can only describe as cotton candy synths.
Like bro I want some deep/low tones with bass, instead of being seemingly limited to these high pitched popping synths.
If you're a sub you can try replacing the desired portion of the end and see if it'll generate a better ending
I agree, since I started putting (directions) into my lyrics including ... Chorus, Verse, Outro.. and the AI has had no issues ending the song correctly
Just make sure keep the same length setting is off
That works but sometimes it still embellishes more than you might like. I had it do a fade out end when a good riff and drum hits would have sounded better
But regenerating the end got me a better one
Yeah it has its spurts at the end but you just trim it when that happens
It still has an ending though
Has there been an update on the "shimmer"?
whats the "shimmer"?
The issue with generation, where in the audio you can hear almost like poor reverb. They've dubbed it "shimmer"
It's hard to describe because the artifact doesn't really behave like other audio artifacts
It kinda comes up in similar situations as pre-echo in low bitrate audio encoding
Yeah I imagine it will be a tough one to sort out
But it doesn't really SOUND like pre-echo
Is there anything I can do to get my songs to not be so heavy with the treb? Like there is no bass unless I turn my audio up all the way, and even then its stupid quiet.
You would use a different program to essentially equalize how you like
Everything is so high pitched it makes my ears want to like run away
At least what I've had to do
That only works if your not making an Extended song - I'm making an Extended song
Imagine being able to request certain guitar types and drum setups
I'm sure eventually the prompts can get deeper
Math-Rock, Metalcore, Heavy Drums, Experimental, Progressive, Angular, 7 String Guitar, Sultry male singer song. Listen and make your own with Suno.
'12 string e flat', 'bridge humbucker pickup'
Best I can do is crop it
If you essentially code your lyrics the AI can take pretty good direction
Found the work around for the shimmer
There's a post about it in the V4-Feedback
When I was trying to force end a v4 song, what I did was:
- Made an extension start a few seconds before the point where I want to end
- Set style tags to: End, Short
- Lyrics only contained:
[Instant switch to End]
[Fade Out, End]
{END}```
Pretty sure having both the Break and Instant End helped end that particular piece.
- Break because it was in a middle of a long instrumental section
- Without the Instant Switch, it wouldn't end for that particular song
Ooo Keep that in mind for the next song
Mind you though, I only did that particular command twice and, of the 4 tracks, one failed to end right away
My credits and subscription page aren't loading
Same
The song has been created, and it sounds very good; however, the vocals are not. Is there a way to use the same, complete song, but change the vocals?
😭
#💬┃general-chat message
https://suno.com/song/8d021965-316b-4e8a-a73f-c4b7a23fb0de
After the calm intermezzo in this song I try to do a final part and outro (extending from 02:32) in the same style as the beginning, but it results in muddiness and different styles. Any ideas on how to force suno to complete it in the same style?
Already tried to keep the style box clean when extending, but that only gives worse results
Yes, I could, of course.
But wouldn’t it be simpler to have a section with all the saved prompts, organized in a list and accessible with a single click? All graphic AI apps have this basic feature, which greatly simplifies the workflow.
A "save prompt" button
A "load prompt" button that opens the list of saved ones.
Click and done.
What you’re suggesting is cumbersome and inefficient. I have hundreds of prompts, which means I’d need a playlist with hundreds of songs. And then what? I’d still need to open it and resume. But most importantly… try finding a prompt among hundreds in this way.
There are users with prompt libraries, and copy-pasting from an external source is neither satisfactory nor productive.
✌️
Today, it seems that Suno decide to duplicate certains parts of my song like the Verse 1. It's weird...
Can't download WAV....
Hi there I changed some lyrics but when I regenerate the lyrics don't change. Anyone know what the problem can be?
is there a list of words you cant use on Suno?
Try downloading in Firefox, I hear it works fine with that (error likely has to do with the filesize from the blob being made)
Ask ChatGPT, it is the censor (or you can ask it what is wrong with your lyrics)
Most should be common sense like no excessive swearing, no inciting violence, no death threats, etc.
I'm not using crazy words :0) Just changed a few lyrics but Suno keeps using the old lyrics. Even if I regenerate.
Producer tags and artist names also can trip the filter (like "You're dead" which is apparently is someone's production tag? Your dead will work fine, maybe even youre?)
I assume you mean Regenerate video? If you are trying to view the clip on Discord, it caches your old one. Add a / at the end of the URL will trick it into thinking it is a new link and cache your new vid
things like im a loser or stupid, will that be ok?
Should be fine. If it stops you, it will say something like it is being reported to a moderator (which is itself - in which case you can GPT itself what is wrong with that lyric)
ive had the problem if i make the same song like 10 times in a row it will make up new words sometimes
if using a persona minnor twaks will not work
When you use song structure tags without numbers appended, e.g. [Verse] instead of [Verse 2], it has a chance of repeating if you Reuse that numberless tag (very helpful for [chorus] really to repeat without needing all the words again)
The other reason could be poor song structure. See #1261146873331318835 message
Honestly, I don't see where the difficulty lies.
Maybe I'm not explaining myself well, and I apologize for that.
-
I write the prompt.
-
I press the 'SAVE' button, and it gets stored.
-
I want to retrieve that specific prompt and, for instance, modify it.
-
I press the 'Saved' button, and a list of all the prompts I’ve saved over time appears.
-
I search for the prompt I’m interested in, click on it, and it’s already pasted into the "Style" section, ready for use.
All graphic apps have something similar, which is simply reasonable.
And let me assure you, I have hundreds of prompts organized in third-party software, which I need to access, copy, paste, and then reverse the process if I want to save a specific prompt.
I don’t see where the challenge is. Really.
Let’s discuss it.
The producer tag "You're dead" belongs to a music producer known as Yung Gud. He is a Swedish record producer, songwriter and member of the production collective SPC (Sound Petroleum Chemicals). that's dumb
Yep, very
It has been a week
No reply since Saturday
Subscribed a pro plan, but still normal
While money gone
🗿
alright... I'm getting pissed off again... I have a 3rd Pre-chorus section that refuses to conform to the same cadence and melody as the previous two, while all 3 have the same syllable counts and patterns... this is on v3.5
The difference between this one and the others? It starts at about 2:10
You may want to remove that screenshot as it shows your email address.
Unfortunately nothing we can do, this is a community-ran server 🫂
Perhaps send another email to support (or billing) and ask for an update is all I can personally suggest
😭thank u, suno really disappointed me
So how do I make suno actually perform the section properly? This is a Cover of the original song, and a previous Cover of the same song got it right extremely quickly
Hmm, honestly I have no idea personally other than extending from an earlier part and then drafting the new part using a DAW (and instrumental as transitions to make the cuts less jarring).
there is no earlier part
Even more strange, ugh
This is an Original, and a single extension
Hm I wonder if assigning it a singer label like [Jane] might help but, honestly, I would not be surprised if it ends up having the same issue
That is the nature of AI. Sometimes it does its own thing.
I'm almost convinced this is 100% due to the fact that this section comes after 2:00.0 in the song
the first 2:00 is always nice and coherent, but once you get past that, it's like it dumps its entire memory and restarts
the previous Pre-Chorus secton comes at a little after 1:00 in, so even if it's 'looking back' a minute, it should be finding it to use
I mean, should I just be making a Persona of the original song and having it completely redo it with all the same lyrics?
With some work, I managed to get exact songs from exact artist generated by Suno
uh... the Persona is keeping the old lyrics
...right up to that all-important 2:00 mark
so WTF do I do? Seriously...
and then there's this...
Replace Section also horrendously sucks beyond... you guessed it, 2 minutes
Does anyone know how to remove the shimmer effect in post that Suno mentioned in the recent announcement?
There's a thread about that topic in the v4-feedback.
https://discord.com/channels/1069381916492562582/1309224710692929548
I've just checked with the devs about this as it comes up quite often. The limit for cover/persona is intentional. I'm continuing to push for this to be documented.
Is it possible to Force the same voice when editing lyrics? Every edit seems to be a new voice
Weird, I’m trying to create a Cover of a song made with V4 model, yet it has switched to 3.5 and won’t allow me to select anything else but 3.5 🤷🏼♂️ Anyone else having this issue?
I have a question, I have made an EXENDET, how can I add it to the song?
On the new part, go to the "..." menu and choose Create > Get Whole Song
I’m working on a deadline so it would be awesome if there is a possible fix for it asap
I clicked on it but nothing happened.
It should generate a new clip in your library (you might need to refresh)
What does it mean "refurbish" credits? In more detail, could you say what you are trying to do?
ty , have it , ty for help
This could be a placebo thing, but it kind of feels like when using personas in V.3.5 now the quality has become worse than before V4 was released?
So as most know, remove section is broken-broken. But is it even helpful to Suno for me to keep using it and accruing all the down votes in this state, or is the broken aspect something they just gotta fix on their end rather than users reviewing it?
To be clear: do you mean crop or replace section? Using best logic, I'll presume you mean Replace as that would most likely warrant some thumbs downs.
Remove section would be a great feature.
Hey, hello ! I don't know if it's the good channel to ask that;
But is there a way to switch voice between a song ?
Like i want a female voice on my first verse, and the 2nd and 3th verse a male voice, is this possible ?
I have been getting "An error occurred while generating the song." over and over for days. Anyone else having issues or know how to resolve this? I've only had a total of 20mins of use in 5 days. Every time I try to use it I get the error.
The improvements are daily and ongoing, so if the thumbsdown continues the algorhythms will continue to correlate where issues remain, and they continue to tweak and improve. So , if you're attempting replace, I'd say, sure keep clicking thumbs down on problemed outputs. But also thumbs up on the good ones to have a better comparison. Otherwise, as I've done, I just test a replace periodically as i create a new song, just for updates. Anything else, I throw on a list and wait.
Is there a way to prevent "Remaster" to alter so much the original song?
That's a frequent questions. No problem. Here's is a great channel to ask that.
We have all tried ways. And it's not consistent to say what works.
Here are a couple of ideas that seem to nudge suno in the right direction:
In the lyrics, note your preferred vocal before each verse. Leave a good line space before these prompts, but I don't know if spacing before the prompt and verse, or having the verse butt directly under the prompt is preferred. I see both as it might not matter.
[Male vocal] or [vocals: male]
[Female vocal]
Depending on your genre, Suno tend to favor one gender over the other. So Rock or Country , you may get a male. Ballad or Kpop, maybe female. Thing of those style which may incorporate both and add them towards the end of your styles list.
Some may include, opera, ensemble, broadway, kpop, rock opera, and maybe CCM.
Example of a prog rock style prompt lost : "rock, glam, fusion, epic, dark, film noire, ballad, opera"
<< this included "opera" at the end because in the lyrics you are going to metatag for different voices or a duet.
Your lyrics , if written with "I"s and "our" and other lines that infer a conversation or a relationship, might get picked up, where Suno will automatically sing it as a duet, assuming it logically fits. So a song about love would be easier to tag prompt in the lyrics for a vers to be sung male or female, where a song about bombs and radiation, maybe not. It's called "prosody", which suno seems to recognize.
If you space after the gender sung verse, adding a line space (which is normal) indicates suno to sometimes reverse back to the previous gender. This is obvious and normal. Just understand, even if you don't say [Male vocal] before a verse, Suno may still switch if the previous was in female voice.
Yo guys need some help here. New in Suno I read the Sunowiki but can't really do what I want.
I'm trying to emulate style of a real song and i ask to some AI to write the description of that song for me but every time they get me back descrtion not usufull for Suno. How can i do?
Don't be surprise if the voice qualities change per stanzas/lines or a male voicing morphs into a female voice. It happens. Then we edit and try to correct it or just run a new gen or a cover or a remaster and hope it corrects itself
First of all, Who is "they"? Do you get an AI to write a prompt to sound like a real artist or band, but when that prompt is put into Suno, Suno says you can't mention a band or artist name?
Is that what you are saying?
Thanks for the explaination ! I did what you said, even add this prompt in the music style :
"different singer in each verse"
Didn't know if it help but it work pretty well !
Of you write a prompting sentence to just include the styles that emulate the artist, and not mention the artist, there should not be an issue. Just don't mention the artist such as, don't say, "in the style of the Beatles"
It did not hurt it... so that's good. I'm glad those things appeared to help. It will never be 100% for the moment. Sometimes it takes click a few generations to find one that works.
You might try "alternating vocals" Suno seems to know the word "alternating" better that "different".
mb english "they" is the AI. nono, i send you an example:
I ask to AI for the type of song and write description for suno and gpt send me this:
Prompt for Suno AI
Cinematic orchestral anthem, male vocals, 140 bpm, rising intensity, powerful drum crescendos in choruses, energetic strings and brass, uplifting rhythm, evokes triumph and teamwork spirit.
But when i use that on Suno didn't sound as i want
So maybe there are something online when i can put no a mp3 and get a good analytics about the song (style, type ecc...)
Oh. Thanks.
I thought it blocked you. But you just did not get the sound you wanted. I see.
So what sound or band or concept are you trying to emulate? Is there a specific artist? So anyone here might suggest something.
trying to emulate this song style 2WEI feat. Edda Hayes - Warriors
Ok . Let me see what I find. I'm going to use.... https://audioaidynamics.com/genre-finder
One more think i'm not sure about. Is this correct to put in suno Lyrics (how is this write)
[Intro growing drum BOOM, BOOM, BOOM! growing intensity, they join the rhythm]
[Verse 1]
.... Song Text...
[ECHOES RISE, DRUMS ARE LOUD, growing drums only for 2 sec]
[Chorus]
.....
Is this correct for say to suno:
during the intro i want growing drum with growing intensity
and before the Choruse i Want growing drums only for 2 sec?
Or i have to write * ECHOES RISE, ... * ?
Here are some suggested prompts:
dark ambient, experimental, neofolk, darkwave, neoclassical, electronic, rhythmic noise, ambient,industrial, black metal, stage & screen
Others members may suggest more... like.. cinematic or other
ty
What about this?
Do I write like this or maybe better puto something in caps lock?
This is an annoying limit.
Tagging on YouTube, what tag for the Offical Suno YT?
It will always be hard to mimic a song that is composed by a human with just AI. "Warriors" is a powerful cinematic anthem blending orchestral grandeur, modern electronic elements, and fierce, evocative vocals. Try the following tags.
Heroic energy atmospheric layering emotional resonance stirring harmonies driving rhythm cinematic brilliance
Anthemic power thunderous beats vivid emotion orchestral layers gripping pace electrifying tension
ty
Woodhouse is sure wielding the stick today.

I know we're being silly, but it's genuinely frustating to have my attempts at helping users be swatted down by an automoderation setting.
Ah...that. Overactive autoimmune response sometimes. Let me find someone who can see about rescuing those posts.
Hey how can I help? All I see in logs is a YouTube link? Was that what you were trying to post?
Suno's YouTube channel is very lacking.
Here's the link assuming this is what you were trying to post:
https://www.youtube.com/@Suno_ai_
I think what causes the autoimmune response is that every link try’s to add a large embed inline. Unless, they are cracking down on links now.
Got me a new laptop. Maybe I can do more with i7 and win 11. Yay!
With great power comes great responsibility.
Oh crap. Not that. I have to be an adult?
To be a villain just disregard the previous comment as rules do not apply.
What type of laptop did you get?
#1154780167047217233 message If may be in Chinese but a nice list to remind us to use/try Italian instead of "bpm" . Suno likes to feel international.
Oh. Errr... an HP from Walmart. Only 15 in. though... I always like 17, but can't be picky. I needed something.
Please tell me it was an amazing bargain and you have plans to min-max it out so that this makes sense.
Oops
is hypocrite, just bought a tablet from Walmart themselves.
Hey!
Who wrote that????
😉
Power/ water / mortgage/ Suno.... the computer was a tight squeeze as is. I didn't win Timbaland.😁
Did they announce the winners already and I missed it?
Thank you, looking forward to trying that.
I noticed that with all my remasters, the first half (intro, verse to half of the chorus) has really good quality but after the first half mentioned, it kind sounds just like the low quality older models. I never got luck on any different songs I remastered in my list. Mind you I have 2 accounts subbed to the monthly pro.
Anybody else been having issues with suno vocals lately? I swear I never used to get vocoder effects this prominent two days ago.
same. early week of v4 was a blast. Vocals were just how i wanted them. But lately it's been bad as well as half of the song quality
I wonder if there's been some regressions as a result of attempting to correct the audio artifacting
Perhaps I haven’t been wording things correctly over the past week or so. This **goth rock, neoclassical synthpop, darkwave, Cinematic, 80s Film, ethereal , baroque, symphonic, ** is the prompt I’ve been using for months, and it worked flawlessly until recently, with the release of 4V.
I find that quite suspicious. Regardless, it refuses to work now—no matter how I reword it, nothing. This prompt used to reliably produce the sound I wanted without any issues. So, I was wondering if anyone with a trained ear might suggest new prompts (I’m at a loss on what to write now) that could help me achieve this sound again. Thank you! ❤️
am looking for this kind of sound...
At least I got a couple bangers before then 
Any update on that wobble/shimmer fix?
I need help! Please!
sigh. I was hoping something was said.
I have a sorta workaround but it's not perfect. The shimmer comes mostly from the music not the voice, at least in my tests.
Use Audacity as @humble pawn suggested a while back - This doesn't actually fix it, it slightly reduces the volume
- You don't need to make it one file, you can take the stems and keep them as stems, the reduction is noticable on both
- In Audacity highlight the selection of the track with the most offensive case of Shimmer/ Wobble,
- go to EFFECT/NOISE REDUCTION and with the section of music highlighted GET NOISE PROFILE.
- This allows Audacity to analyze the noise
- Go back and highlight the whole song
- back to EFFECT/NOISE REDUCTION and to to step 2
- Settings that worked for me were - Noise Reduction DB 9, Sensitivity 15, Smoothing 3
Click OK
This will apply it to the entire track, this does reduce it a bit but there's more work to do...and if you want it even less crappy, read below but...that's gonna be using outside support in plug ins and if you don't have them, buy them now as they are all on sale big time
OTHER STEPS (and this isn't for everyone I know but it seems to do the best job)
in your DAW ( I used Acid pro and Ableton to test it)
Open the stems
- Use Waves X-Noise and set it to Pink noise "BOTH" on the music stem
- Then take a compressor ( I used Neutron 3 compressor and Waves V4 Comp for testing) apply this using the Drum Pump in Neutron or Master compression in Waves
- Apply Ozone 11 master
This will take away 90% of the noise. it is still there because its an artifact created by the audio render from the source, so NOTHING is going to remove it 100% at all unless the devs redo the audio engine or allow for individual stems
Waves, Izotope and Native Instruments are all having insane black friday deals right now, A LOT of these plug ins are going for under 30.00
It's being worked on. I expect nothing will be officially announced or changes implemented until they're confident in the solution (if they release a fix that causes other issues that are just as bad, that wouldn't be good.) I get the frustration though, the shimmer bothers me too.
Until then I've stopped making music with it.
So, we can assume that unless they officially announce a change, there are no changes, at all? No subtle weight tuning right?
Cause I don't want to go crazy thinking things are changing when they arent
Can't guarantee that tbh. I meant specifically in regard to this issue, they won't say things like "today we tried X Y and Z to fix the problem", if they say anything at all it'll be when it's fixed.
does get stems cost credits?

Every song ive done today has been 3 minutes 10 seconds and it cuts off the final bit of lyrics. I thought the limit for v4 was 4 minutes? Is it something I'm doing wrong?
they were all fine yesterday
What's with the captcha popups in the middle of our workflow recently? Is there a way to just do it once and be done with it?
Are you using Personas or Cover?
personas
same as I did yesterday and they were fine
same ones*
Personas are limited in length.
@unique stone mentioned this limit up above. #📚┃suno-school message
Maybe if we bug... err add feedback enough the devs will lift that unacceptable limit.
Interesting. I must not have done the persona on the longer ones
Ok, thanks. I'll send an email about it, help make them either document it or change it maybe lol
thanks for directing me to that
Sometimes Persona's will be shorter and a bit longer than 3:14 seconds, but most of the time they cut off at around that time limit.
Covers apparently have a similar limit.
good to know. Appreciate it!
I don't know if someone could please help me with a "Producer Tag" issue? Please ASAP.
I have a song with the lyrics "Hi 'me', it's 'me' again."
And it won't let me recreate it or work on it because it says "It's me again" Is a Producer Tag. If this isn't an error, I have some serious concerns with not being allowed to write "it's me again" in my lyrics
I don't know where to put this so i hope this chat is OK? If not, I'll ask elsewhere but I'd appreciate if someone could help or direct me to someone who could, please?
it's me again = 🚫
it's mee again = ✅
Following my request to have this limit documented, I'm told that this has now been changed and is no longer limited to 3:14 however it will depend on the length of the song being covered (shorter song = likely a longer cover, longer song = likely shorter cover.)
But why? I understand there can be work arounds but how the hell is "it's me again" allowed to be sanctioned off like that?
Same reason "i don't love you anymore" is blocked 🤷♂️
But what's the reason? I'm assuming it's just unknown??
Sorry, it's just a little frustrating
Straightforward answer is they're blocking some content known to be present in other songs -- as to how the terms are determined is entirely unknown.
Usually, if something is blocked like that, it's the name of a song or something.
Cause I write my own lyrics from top to bottom and they're usually vent songs
That's crazy, that could limit a lot lyrics
Korn has a song called "it's me again", blame Iowa
Annoyingly some music producers decided to use some regular phrases like that as their tags. Which means they use them in many songs, that Suno might have trained on, leading to a stronger bias toward them in the model simply due to the same audio being repeated several times causing the model to "learn" it.
So basically it's to prevent someone from getting Suno to replicate such tags (intentionally or not.)
Well, at least that makes sense.. sucks though
I think?
I hope there's a better way to do it, cause it's not like I'm using a tag, they're lyrics in my song
Well, thank you guys for the info and assistance. I hope a better solution is found for the future
Sounds like a simple misunderstanding that you can workaround with misspelled lyrics -- if these restrictions make it easier for Suno to exist, even as a stopgap method to show that they're trying to prevent bad actors from ruining it for everyone, it's an acceptable evil in my eyes.
Yea. Like I said, it makes sence. I just hope a better solution is found
But.. to be fair.. if it's as easy to bypass as just a misspelling.. it it actually working.
I don't fully understand how suno works
Hi, i dont know where to put this but i disliked a song by accident and i cant find it, where can i find it? or is it gone
Downvoting only removes from the create tab -- should still be recently new in the library tab, and the downvote will make it easier to find.
For me its not in there.
I think (nobody knows for sure apart from the devs!) that it's trained on sound clips alongside text representations of those clips. So, the literal text "it's me again" is probably somehow associated with the producer tag audio, whereas "it's mee again" probably is not. That's my guess.
Either that or it's a lot more simple and they just block the words in the hope that nobody tries to work around it with the intention of getting the producer tag to generate 😅
I assumed the former as well though I fear the latter
It's not a perfect implementation, and there absolutely could be improvements, but imagine the alternative, where they've put ZERO efforts into preventing copyright.
ChatGPT had individual authors suing it because of the lack of restrictions put on how people were able to use it in order to try to get it to regurgitate training data.
No no, I do understand and agree. I just didn't have the understanding before and now I just wish there was a better way. Once again, thank you for you assistance and all others who have helped. I'll find ways bring my lyrics to life.
Hope you all have a wonderful day and take care
