#📚┃suno-school
1 messages · Page 99 of 1
wow do you know what I didn't know?
that if you take a song you did a while back and make it into stems now....the wobble and shimmer is worse than on the original track. why on earth would that even happen?
I feel that 100% but again if the work around is adding an extra letter, it's not too big an issue if it means that artist have a harder time suing Suno. It's why I'd rather wish for a better solution than no solution
"Throwing caution to the wind" for the sake of artistic integrity sounds badass, but when your decisions effect the livelihood of your entire staff, as well as the longterm feasibility of your product, it's less of an argument to say "so what, who cares"
Yeah, if only individuals could be held accountable and not the tool they used
Not an attempt to dismiss legitimate frustrations, but sometimes I forget how wildly mindblowing this technology is, and what degree of science fiction it felt like just a few years ago to imagine a robot making music for us actually was.
The fact that we're complaining about finetuning a product to give us exactly what we want on-platform is neat, when you step back and appreciate everything for what it is, warts and all.
There's an entire generation of new artists who are learning to hate the sound of rivet-cymbals just the same as I do 
I'm a song writer, well.. a lyricist.. I can't really sing though but my songs are pretty personal. I don't have the tools to bring them to life so having Suno has been amazing. I've even become a better song writer because of Suno- It's absolutely amazing
I'm also dyslexic.. so sorry for so many typos
It makes some great templates that I could give to actual artist and maybe even get real instruments and vocals in the future
Just occurred to me
Listening to one of my playlists and”when you walked away” had a sh/reload of shimmer the second time through, & MORE yet the third time.
But for now, I'm just so stoked to be able to hear my lyrics sound good in music format.. So I guess I should really be thanking those working on Suno for helping someone like me re-discover my passion. I'm fine with a few frustrations so long as Suno stays available and keeps growing
😄 the immediacy in being able to hear your words brought to life absolutely would have immersive effect on lyric writing 🙏 Suno definitely breathed life back into my musical passion as well, so I appreciate you sharing your story 🙂
Maybe I could share one of my most recent songs? They're a bit dark but I love sharing my work
Hey wonderful information! Thanks for expanding upon further steps for this, because yes... this WAV stem download method, isn't the be all end all. It can take some depth from your tracks, and it sometimes doesn't completely remove the issue. I think the base of the shimmer is gone, but what is left is still some form of low quality noise, or a phenomenon of a sound suddenly missing from the space. Just stopping at downloading the WAV has been more than enough for my satisfaction, but not for everyone.
I knew eventually people in the community would experiment with next steps. If you don't mind, I'd like to add your comment link to my posting here: https://discord.com/channels/1069381916492562582/1309224710692929548 
I should consider getting a mod to also rename the post to 'REDUCE SHIMMER', and amend the wording in the post.
I'm just keeping my spirits hopeful for the near future when they've figured out a fix. Which I am sure they will.
Content posted here should be AI related. Advertising a game promotion here is also against the #📜┃rules @unkempt ridge
Are you lost? I think that was for #🗣️┃chill-chat
I put it there too.
Yeah Woodhouse was meant to reply to you in that channel
Why are all the covers of the songs I generated suno, and the sound quality is also very bad. I am using v4
go for it, anything I can do to help
there are currently issues with v4 that the team is working on. Best to just wait it out for now and work offline on things you've done in the past to make them sound better.
At least, that's what I am doing for the forseable future.
I created a piece of instrumental music, but the song is too different before and after. I want a piano piece, but the style is too different. I hope you can solve it.🥹
Is allowing a Persona to Cover a song something that's going to be a feature or planned? It would be really nice if I like how the song was sung prior but want a different Persona
im still having fun in 3.5. and then i roll the dice on a remaster at the end. sometimes it hits, if it doesnt, i just stay satisfied with the 3.5
Hello !
Is it normal that since V4, when I let suno write the lyrics for me (by giving him ideas), it only generates lyrics for me in English? (even if for example the basic text that I gave him is in French)?
Did you try using French as a tag?
yep
That is awesome. Is this limit removed for Personas?
I noticed Suno did that to one of my songs once when I wanted a French-Waltz
Example : "Une chanson sur les éléphants en français" did =>
Ya all in my cast iron kitchen
Rollin' that dough
Deep-dishin'
Said you like to start from scratch
Hope you're calorie efficient
You start cuttin' onions
It's burnin your eyes
We taste-test the pasta
Cuts the taste with wine
[Chorus]
I start stirrin' the sauce
You start stirrin' the pot
And you're crackin' the jokes while we boil
We start makin' a mess
Baby
Lemon zest
Should probably replace you with foil
[Post Chorus]
That's why I like you to come over
When it's two in the mornin'
...``
As far as I'm aware it's for both cover/persona (they seem to use the same underlying functionality)
Can't I keep using Suno on my own discord server? Tried it again after a few months and the Suno bot isn't reacting
The bot has been shut down a while back -- it's just on https://suno.com or the iOS app https://suno.com/ios now
Suno is to music as Figma is to UX prototyping.
I have a small but annoying problem. I made a song with V3.5, consisting of 2 parts. I then generated the whole song from this and wanted to remaster it with V4. The problem now is that I get the error “Prompt flagged for moderation” when remastering. I know that the error is due to explicit lyrics. But why was I able to do the whole song with V3.5 without any problems and when remastering I get the error???
Maybe remastering has a censorship component to it now.
Hello! Whenever I want to recreate a section with new lyrics when it recreates the section it is the same lyrics. I have made sure to change the lyrics in the song details as well and it still doesnt work. Please help! Would love to use this amazing feature. Thanks
are we noticing it does that for when we use "get whole song" as well. Cause im now experiencing it there as well
It is a thing. They implemented new safe guards and even if the song passed them before does not mean it will again. They won't delete your old song or anything but you won't be able to make it like that any more. A remaster actually runs it through the AI so it is subject to the new safe guards.
Not sure what you mean. Get Whole Song shouldn't be limiting song length.
Anybody know any tricks that will let you change a song's prompt? Everything goes back to the base one after compiling the song whole, and I want to adjust the prompt pre-remaster if possible.
Covers sadly, muck with the pacing and such too much for my needs in this instace.
Cover is the only way I can tell you that can change the style prompt. Which you addressed. I can tell you that cover doesn't always mess with it. Try putting the original lyrics with all the same instructions back in the lyric section but change the style. I put things like remaster, or some other thing that really doesn't do much. Sometimes I just put my name.
I'll give it a shot. I think I might have a workaround.
You can remaster a remaster, but it won't change the style. If you find a work around let me know
yeah for what ever reason mine is being limited to 3:14
Covers, or using "get whole song" on an extension?
My thought is splitting it in two, since it's just under 4 minutes, re-uploading those parts, extending them just a titch to add the prompt, and then get whole song.
"get whole song" on an extension?
Ah. OK very specific situation...lol
Has the making of Persona been off for uploads for a long time?
I did use a persona at the first level which limited to 3:21
So if you take a 4m song, add a 2m extension onto it, and "get whole song", you're saying the resulting clip comes out at 3m14s instead of 6m?
correct
Yeah, I'm remastering a tune that I'm using in a music video that's 70% edited already, and I dont' want to have to muck with the beat since I've been animating to it.
Okay -- gonna test this now.
That is crazy. I haven't been using persona, but that is insane. I don't even know what bug would cause that.
Gotcha
Steps i took: Persona Song - it got limited to 3:21, created an extention - it was 2 mins, hit get whole song, created a "full song" which was limited to 3:14
yeah im a little sad cause the song came out so clean
I've been nursing the song for awhile, the original version was awsome in structure but was like chiptune underwater, and i've been remastering it manually, uploading it, covering, etc for ages.
worse case ill just grab both parts and combine them in my editor, but was hoping to not do that much hands on if i didnt have to
Gotta coddle the baby. I feel that.
as an additional note this is using the V4 as well
The persona limiting thing is already strange, but I'm sure if I could see the background code, I'd say ooooo...that's why, but the extension limiting after theharbor checks it, if it is constant then it needs to be reported.
doesn't seem to be an issue for default generations from v4 -- gonna test it with a persona
same with tracks extending from persona
yeah it didnt happen to me earlier in the day but now it is
well thats disapointing on my end then
im trying again now and seeing if it create correctly
nope still happening to me
If it's the same issue later tonight, it might be worth reaching out to support@suno.com so they can look to see if there's anything wonky with those tracks, on the chance they're able to resolve something.
prrof im not crazy
I didn't think you were crazy or fibbing 🙏 
ill try again later tonight and see if its just a hiccup on my end and if its still borked ill reach out to support
I've seen enough weirdness to know I need to trust AND verify simultaneously
Nah, crazy stuff literally rotates on suno.
as someone who used to be a QA specilasit it always sucks when you find a bug that peeps can recreate
is the noise in the background the shimmer afect
Wow ok that doesn't seem good... What time did you extend from part 1? From the end?
If it sounds like a wobble flanger cymbal then yes
3:10
Sucks more when you find a legit bug that nobody else can replicate, because they'll just say "doesn't happen on my system" and close the ticket 😛 (just relating to testing in general, not explicitly calling Suno out)
Intermittent is evil
the absolute worst lol
The worst troubleshooting ever. Like when dad gets in the truck in younger years and it stops making the noise.
Thanks for helping me look into it a bit back to the board to edit
is v4 able to split stems with vocals any better?
Are these parts definitely the ones that were used in the full song and if so, does it still do it if you use Get Whole Song on part 2 again? I generated a test song and an extension and got 3:58 + 1:58 leading to a 5:57 long full song.
and extending the first part at 3:00, getting 2:00 extension, full song for me is 5:00
I already did the same tests, posted above his.
Not sure if it's giving limited results from all songs or just a quirk from that particular generation though
Would be good to have links to the parts to inspect in more detail
How do I make faster flows? I put a no-vocal 21 second sample, and it spends too long on the hooks that it can't rap a usually doable 2 verses
yeah its still happening for this song, but it didnt happen for a different song i made
Is this what your looking for?
part 1 - https://suno.com/song/a4bb028a-ca29-4f79-9aaa-251933c3d649
part 2 - https://suno.com/song/6e7b48a5-f844-4c6d-8a47-52144b89617e
Yes - thanks! Can you send the "whole song" link as well please?
I just did a stem seperation and the 6.1 is only 7 seconds in the stem 😮
lol i think it broke for that song
the stem shows 2 mins but when hitting play it only goes for 7 seconds then stops
which would explain why when making the "whole song" its only 3:18 and not the full 2 mins
Is there a way to put two voices in the chorus, but like, one singing and the other saying oh..oh..oh.. or even singing another verse over that first voice?
for sanity purposes, don t know why the bug happened but for what ever reason the 6.1 showed 2 mins, played for 2 mins, but when stemed out it would only play for 7 seconds, which would explain the "whole song" being incorrect as well
Looking at the page metadata everything seems to be as you have described (extend time, correct song pieces) so I am really not sure what happened there. Something must be broken with that second part. I have some other concerns about how this is rendered so I'll forward these onto the devs to look at
awesome, thanks for looking deeper and forwarding, i also flagged it in the bug report from the tool.
positive note, i just did a new extend and got a better version and that worked perfectly
Oh good - maybe it was a sign! 😄
If you end up putting any of these in the trash, just leave them in there so the devs can at least see them
Hi, how can i use only the v3.5? couse when i try to extend a song, the page turn it to the v4, but i dont want it. HELP!
ill leave the ones i sent in this channel in the libray and let them drift into the oblvion that is my library
Yes with a lot of trying and using lyrics like:
lalalalalal (oh oh oh oh) lalalala
they might overlap with enough tries and prays. You might need to put duet, layered duet or something similar in the style and lyrics as meta tags.
My best example so far:
https://suno.com/song/d589386e-efd7-4ac8-871b-e609748fb19c
If you're extending from an audio upload, I think you may still need to crop the song first as a workaround -- even if you crop the entire song, essentially creating a duplicate.
wait, the devs can go and rummage through my trash can? Are you sure?
Why do you think you're encouraged to put lids on your bins when you set them outside?
Devs are essentially raccoons with access to coffee.
Does anyone have a suggestion what is the work around to make it generate the song as I prompt in "major key" but it keeps on generating in minor and a very melancholic vibe? thanks
no way, I am not ok with this
Have you tried using minor key, melancholic, sad etc etc in the negative prompting section?
no, it didn't occur to me, but I will try, thanks
I love negative prompting, because I feel like you can overload it without as much fear of it overpowering the other style prompts -- where if you put major key, major key, major, happy, cheerful directly as a style prompt, the other genre prompts might not be as well respected
Since it's negatively impacting the weights, it's like finetuning by subtraction instead of adding more of what you want when it may be carrying an uncertain/unstable prompt load to begin with
seriously, how do I keep raccoons from accessing the coffee? would deleting songs right after being generated and then download them into my DAW ASAP? so when they come for coffee, they see nothing... then they go to your trash can 👀
If you delete from the trash they should be gone forever.
But now even I'm curious about what secrets lurk within there...
I used today at least 600 on songs and 90% are unusable doing to the shimmering flickering background noise.. l hope they will fix that soon
Quick clarification, I don't think the devs actively go exploring & checking out people's songs -- I just thought it was funny to reverse-anthropomorphize machine learning engineers as trash pandas.
re: song privacy, afaik "trash" acts like a separate storage, but once it's deleted from trash it's gone -- but being in trash keeps it active on the servers in case you want to bring up back to your library later. I think furrybeans was just meaning that if people wanted to investigate something later, it helps to note delete-delete it
Had a fairly quick response on this: Basically something went wrong when generating part 2 (kind of obvious at this point since you tested with the stem split etc.) - if you extend part 2 it will also get cut short.
But I noticed that in the full song, part 2 sounds different in a few ways. I think the audio clip you hear for part 2 is actually not the same as is being used internally when you split/extend/get whole song etc.
https://suno.com/song/2aaf4285-60cb-44dc-a3fa-bef12ec0b20a What the hell happened here? lol I kinda love it, like a nightcore render, makes me think would be cool to have a bpm slider.
apocalyptic cyberpunk, acid beatbox, angelic gravity vibrant, overwhelming acapella harmony song. Listen and make your own with Suno.
Can someone explain why these lyrics trigger the censor?
[Verse]
Mama packed up all our things,
Moved us out to where birds sing.
Concrete jungle left behind,
Now it's farm fields in my mind.
[Verse 2]
Traded sneakers for some boots,
All the locals gave me hoots.
Cows and chickens all around,
Dirt roads where no horns sound.
[Chorus]
City boy gone country, what a sight,
Trying to saddle up a horse, hold on tight.
Fishing at the creek, ain't no streetlights,
Learning how to live those starry nights.
[Verse 3]
Tried to drive a tractor, Lordy dink,
Got it stuck beside the old hay rink.
Neighbors laughed and lent a hand,
Guess I’m part of this here land.
[Verse 4]
Swapped my skateboard for a quad,
Out here it’s just me and God.
No more traffic, just the breeze,
Nature’s way of putting me at ease.
[Chorus]
City boy gone country, what a sight,
Trying to saddle up a horse, hold on tight.
Fishing at the creek, ain't no streetlights,
Learning how to live those starry nights.
Never mind. Apparently it was the word "dink." Haha. I didn't realize that had any connotation.
What explicitly does the error message say?
(Oh, just saw you figured it out)
It said it couldn't create the content because of something to do with inappropriate material or something. Lol
Suno must be a jaded parent if “Double Income, No Kids” triggers it 😉
It popped up too fast.
Haha
Too many... "nefarious" creators out there I guess.
I'm late to the party on this one but I have seen this word cause problems with Suno before. Unsure if this is exactly the reason but one definition is: "Used as a disparaging term for a North Vietnamese soldier or guerrilla in the Vietnam War."
Seems it may also be used as a racial slur, linked to this.
Can’t the shimmering issue in V4 be resolved by excluding certain instruments or musical elements causing it through Exclude Styles?
Ah, interesting. It's funny though that Suno wrote the lyrics then gasped when I tried to have it make the song. It's like creative differences between writers and artists in AI style.
It's a bit weird when you consider it's the same service, but yeah it's essentially several different AI doing different things and they all have different opinions 😄
Yeah. That's for sure.
Yeah 😅 Just adding a bit of levity -- fun fact relating to how the neighbours "The Dinklebergs" were named in the show Fairly Odd Parents
It's less of an intentional choice made by the AI, and more of an unfortunate technical result. I have faith that progress is being made 🙂
https://suno.com/song/ab5f794c-96e6-45a1-af70-82d3ebfadedc - at 00:30 the chorus starts when it gets to the third line is the large overlap. It is very difficult to get multiple words to overlap, the best way that I have seen was by doing way too much to actually control it. I have been trying/wanting Broadway Style dual lyrics running at the same time. V4 is getting closer, but I think that this will be a limitation for a bit due to it not being anything mainstream. I think we will get Background vocals first.
I haven't used "medley " as a prompt yet. Maybe that would help.
Hey guys
Could anyone help me identify the style of this music intro please? So I could write it inside the Prompt box
?
[doubled] works pretty good
Interesting!
I usually add it in the style as well, "doubled hook", "doubled vocals" etc.. obviously helps if style you going for has it more often.
I've tried pretty hard to keep a whistle going during a verse/chorus, but not much luck unless its more like a flute
Never tried that. Very interesting
"There's a problem with the account's cover songs - it regenerates a new version every time I try to cover a song, and the images have watermarks"
Hello, I have a question, how are some users managing to almost perfectly replicate the rhythm and melody of famous songs like "Bohemian Rhapsody" for example, when Suno is unable to recognize existing artists and songs? How do I do this?
You can add musical and vocal descriptions over every line, so pretty much just a lot of work, and probably 100's of generations etc.. with replaces, extends until you get what you want
Really is that how it is? Isn't it just a case of downloading a song and creating a cover of it? Wow
If you upload audio with vocals in it, you can share by link but can't make them public, so easy way to tell if someone likely made it from scratch
But how does this work? Do I put the description in the lyrics or in the Style so that the song comes out the same as the original version? Like: "Create a melody in such a chord, such a tone, etc."?
Extension tokens? Like a single token used for the model being able to extend its context window by sliding its window effectively. Just came to me, everyone quick... You better @ me on twitter when paper comes out... hymba, on extension and pause tokens trained on combined merge of LONG summarization and repo level refactoring or I mean totally whatever the equivalent for music is lol
I mean you have unlimited ability to be creative I don't really see the point in copying what already exists. But just like making an original song, style and negative prompt are the overall driving sound of the song, and the lyric area is where you can fine tune some details.
But you can get very detailed such as: "[Instrumental: background whistle in d minor starts with long, sustained notes then slowly ascends from D to E to F and builds a melodic interval flowing continuously, simple intervals, vibrato, reverb]"
Has anybody had success in transitioning a song from one genre to another. Say pop to rap or rock to country?
Keep an instrument from the country as a subtle part of the background, then bring it forward when you want it to switch, and use a [breakdown: how to change here]
pls help-me
If it's doing that for all the providers, probably an adblock or something
i don't use adblock.
Was just a guess, vpn?
I don't use a VPN either.
I would just try different auth provider, or browser then, or try phone vs computer etc.
Hey, everyone!
I am new to the generator and it does not seem to follow the instructions I gave it. It reacts to lyrics section very well, but fails following the 200-character prompt. I clealy know what I want out of the machine, I have voice samples, reference tracks etc. But after burning 500 credits for nothing -- I decided to ask for your help here 🙂 Prompt example: ""Thick 127BPM A Major TECHNO. Deep male vocals with clear and proud pronouncing (NOT SINGING) of lyrics. GERMAN SOUND. Sound of Burning Man main stage. Refer to uploaded track before composing."" Can I have a link to Suno prompt engineering guide, please?
I tried asking that question in general-suno chat, but it seems like the messages get lost in flood chaos there.
Hi Denis, think of it more like you are influencing the AI then telling it what to do. So the order of the style, and words matter. Filler words like "of" or "and" are usually pointless
You also have a negative prompt section, so use both styles to achieve what you want. Use technical terms, so "Not Singing" = "Spoken Word Delivery"
Ensure you have a space after each comma, drive the overall style with a style prompt, remove things you don't want with the negative or by modifying the style, and add complexity and texture in the lyric section with tags [Breakdown: pitch down, 120bpm, vocal drawl, vibrato, sustain] etc..
Thank you very much for such a helpful and fast response 🙏 Will adjust my prompts now and try again!
Also be specific yet concise, so "Thick" what? .. piano? 🙂
thick rolling sub-bass, pocket delivery, reese bass, 125-140bpm, panning, haaz effect
Then you can control the bpm per section
[Intro: 125bpm, slow build, etc..]
[Bridge: 140bpm, fingerstyle guitar, etc..]
For keys, [thick rolling sub-bass, key A] or [thick rolling sub-bass, key Am] seems to work for me to save characters.
what's up with "replace section" not even blending into the existing music?
watch the video on lyrical alignment, but if you have already done a bunch of edits sometimes just no luck.,
Does it just not work well for instrumental tracks?
ohh I would assume for instrumental tracks it would work perfectly, I usually have trouble with more fast paced vocal tracks
no I think I pick the perfect replacement, then I apply it, and there's an abrupt skip after teh replaced part
Maybe try with 3.5 or 4 (aka different model). I haven't done instrumental replacement so just guessing.
@fathom siren Highly appreciate your depth of response. Thank you very much indeed!
No problem, couple other tips, you can do tags like this in the lyric section, so you can get really specific but in general I try to make them not sound sentence like.
[Instrumental: background whistle in d minor starts with long, sustained notes then slowly ascends from D to E to F and builds a melodic interval flowing continuously, simple intervals, vibrato, reverb]
Also you can do
Gold leaf mirrors *distorted* distort the scene..
or
[distorted vibrato on "distort"]
Gold leaf mirrors distort the scene..
[slap bass, light guitar. stand up bass]
In this tag, "stand up bass" will probably get interpreted as lyrics because of the period, and be sung.
I tried the work around for Shimmer and it did eliminate the shimmer but it does not sound the same when you combine the stems
Losing audio quality even with WAV files
Any other ideas until they figure out how to get rid of it?
I notice not everyone is dealing with it but in my Prog Metal generations it's so prevalent in every song I make
I haven't seen that [breakdown] before do I need to put everything that you typed in the bracket?
no that was just example, [breakdown: mute guitars, fluid, rhythm builds, slap bass]
etc.. just how you want the song to change at that point.
I need this pretty much, if it's possible
[Pop instrumental]
[Dynamic pop female singer]
Lyrics blah blah blah
[Door slams]
[Hard rap instrumental]
[Deep male rapper]
Lyrics blah blah blah
I would probably do [pop] [interlude/instrumental][pause/silence/break] [door slam] [pause/silence/break] [breakdown: baritone vocals, beat-driven rap, pocket delivery], then re-enforce the vocals on the next rap verse as well.
Do you think I will need to burn through quite a bit of gens to get what I'm aiming for?
Just depends if I have something specific in mind I'm trying to get, It can take a LOT of gens, If I'm just kinda flowing looking for something close, probably not that many.
I feel like we should get some sort of credits back for the shimmer effect issue. Is there any chance of something along those lines? Granted its like $10, but more just principle, since most of what is generated at the moment is hard to listen to.
That would be awesome, especially since I need it to all happen within the 1st like 30-40 secs of the song
Why not just generate on 3.5?
Still there too
Since i got monetized few days ago, looks like youtube has higher criteria for monetized users/content creators.. I made compilation of 25 blues songs created with suno v4 and edited custom video and effects.. Youtube recognized even four songs as others people music or covers and offer me limited monetization (i must share ravenue) I have to removed these songs in order to regain control and keep ownership to these songs.. Can developers refine suno v4 to be more creative in order to avoid patterns which youtube could recognize as covers or others people music?
Wouldn't it be better to talk to youtube about fixing their algorithm.
its impossible.. they never admit mistakes on their side... Chat support and mail support are useless because first and second level support staff are useless idi0ts who can't really help nor have power to change/fix anything. They are only very kind but for any serious problem they have to contact someone in their hierarchy who is more skilled and have more power... in most cases its dead end for serious problems. I already tried
Well from a technical perspective it's almost impossible as how do you move the AI towards that goal, without having access to the algorithm that is flagging the content.
anyway, suno could be really make their engine to be more creative.. trained data should be used to create brand new content, not copycat and plagiarized patterns which youtube and similar services could detect as covers or songs of other people.. i don't have similar problems with 3.5 model but in time when i was using suno 3.5 I didn't have monetization active
You can just write the content right?
i like to make instrumentals and then mixed these 3-4 minutes suno artworks into one big 60-90 minutes mix.. i have some normal songs (3-4 minutes) with lyrics as YouTube videos but i always prefer instrumentala
https://suno.com/song/f8447a53-5451-4ce5-a93d-2ec1e9e93e08 took me a couple generations, but this shows start contrasts.
1:14 for transition
Has Personas creation been disabled for uploads since yesterday or longer?
I believe that personas are disabled if a vocal is detected in an audio upload
Best way to force end a song?
Yeah, I thought that too but when I uploaded an instrumental track I got the same msg.
If you are planning to remaster: extend to 3.0 and [End] it
For 4.0, it is more luck. See
#📚┃suno-school message
#📚┃suno-school message
As some possibilities
(Unless you are okay with a fast/abrupt end, in which case you can use the Crop tool)
Guess they disabled it altogether to help avoid copyright issues violations from Audio Uploads
It worked now with an instrumental that was clearly without any vocals. Just a piano instrumental. The thing with V4 though is that it creates distortion on really quiet piano music. Hope that something they will improve.
Hi there, It's pretty weird, Suno seems to not follow the structure of the song, I dont understand why. The verse 1 doubled without any reasons, according to my knowledge
https://suno.com/song/ae29d6b2-5a84-421a-96ea-4040d4ba2c12
French Eighties New Wave, Slow, Emotional song. Listen and make your own with Suno.
Still, you can be nice—nobody likes being called an idiot. They’re just doing their job, and I’m sure you are too. I’m frustrated as well about not being able to use the service correctly, but being angry won’t help the situation.
Is there a way of telling suno to change just a part of the song. verses are on point but I'm not satisfied by the chorus. I see there's a "replace function" but i do not wantt to change the lyrics, just the way it is sang
Hi! Yesterday I payed for the suno yearly plan on the phone app via google pay. Now, I login to my account via webbrowser from pc and see that no subscriptions are active. How can that be?
yesterday we have had a similar issue, all our credit were gone, it cam back on its own, everything's good now
groovy blues metal, brutal glitchcore, dark acid rock, experimental noise, chopped and screwed song. Listen and make your own with Suno.
can you repost that here instead https://discord.com/channels/1069381916492562582/1083129341870878780 ?
what Should i add for Putting in a snyth solo or any other instrumental part of a song
I think you have fallen for a scam app. Suno's Android app is not yet released, currently on pre-registration (https://suno.com/Android).
AFAIK, there is no current method to pay with Google Pay atm?
What to do:
- If you logged onto a third party app, immediately change your password as it may be compromised.
- Enable 2FA if the login authority has such a feature available.
- Cancel/dispute your transaction.
If you made a song through the third party app in your main account, contact SUNO Support ASAP to let them know what happened so there is no flags raised on your account (support@suno.com)
i am always nice with them i just show some frustrations here with my comment... they are nice but in most case useless to solve real issues caused by their so rigid or bad AI which they use to scan ane evaluate uploaded content
it's annoying, but it happens
as long as you don't change the lyrics, the replace section will keep them the same
but sing them slightly differently
Very, very slightly then. It's barely noticeable
#1256301881966198805 message Terminator Extensions make it relatively straightforward 🤘
clip 1 - 0:00 > 0:27
clip 2 - 0:27 > 0:28
clip 3 - 0:028 - end
is there a problem with replace section? It doesn't seem to generate, it's loading indefinitely
You just keep trying :p
also, anybody else noticing that the "exclude styles" thing does the literal opposite of what it's meant to?
to me the style prompt is more often than not, not respected in v4
Yeah suno does not seem to give a sh
If I get one more generative image with balloons imma ....probably use it
This last days it as more than 80% of the generation 😦 It's consume lots of credits
How suno playlists work? What happens if I have "open" playlist with "hidden" songs in it? Will they still be available for people through it or not?
You can't "hide" songs -- private just means they don't appear on your suno profile & can't be found using the on-site search.
The only way to "hide" a song is to not set it public & never share access to the url
yea, that's not what I'm asking tho...
What heppens if someone opens open playlist with hidden songs - will songs be visible or not?
I gave a checkmark-react to say "yes, songs added to a public playlist will be visible" -- I was giving follow up information to say "you can't hide songs in the way you're thinking
Something tells me your song structure is very short
If there aren't enough lyrics, it'll just make some
or copy existing lyrics onto other parts of the song
so basically like youtube, good
Setting a song as "private" doesn't "hide" it in any way, beyond making it not show up on your profile or the search feature.
It WILL still show up in playlists.
So download the wav and the stems and throw it in the tash?
I guess
or is trash also discoverable
I'm not sure that relates to what the user was asking about...
As in, yoink the actual song, and then delete it from the site
or was the question not about how to make something truly private
I might be confused, but I have no idea why you're sharing this with me.
I was helping another user, answering questions about the "private / public" thing...
it never was about making something private , I just wanted to know how playlist visibility works 😄
ahh
makes sense now
I've had some success with this
distorted vocals, distortion, industri, reverb, harmony, layered vocals, stacked vocals, multiple vocal tracks, chorus effect, chorus reverb, chorus harmony, backing vocals, backing harmony, auto-tune
in exclusion
I wanted some songs to be visible, but outside of main page, so that's perfect, just have seperate playlist with more stuff, main page just for goods
to try and avoid the extreme amount of reverb and backing vocals it loves to generate
Exacty; it's a great way to organize content for others -- I only have a handful of public songs, but there's a playlist with ~60 experiment tracks that would otherwise clog up the main page
@charred merlin do you have any other tips to further reduce the persistent layer effects?
I'm going crazy
overdubbed?
it had a little effect at times when I used it as positive, idk
Sure, but bringing out those effects is much easier than suppressing them
I'm talking about the exclusion section
we need prompt weights, like in image gen, (genre:1.2), (style:1.2), less important prompts...
I have these in inclusion acoustic vocals, crisp vocals, unprocessed vocals, clean vocals
I'm still waking up, so I'm only operating at half-speed atm, but do you have a clip that demonstrates what you mean by persistent layer effects?
I don't want to pigeonhole people into feeling like they need to use the same terminology, but do you mean the effect that others are calling "shimmering"?
I usually just go for "raw", but it often leads to rock...
which also help, but because they directly conflict with what it tries to do, they are sometimes just rerouted to other styles. "Acoustic vocals" sometimes just summons an acoustic guitar for example
The helicopter blade noise?
I don't actually have that problem most of the time
very rarelly get distortion issues either, probably due to the exclusion tag
My issue is that the reverb and backing vocals persistently show up (especially during the chorus) despite my best efforts to keep the vocals dry and clear
And when the reverb shows up, it almost always comes with some autotune artifacts
So I referred to em as "layer effects", though I realise in hind sight that reverb isn't applied as a separate layer, but backing vocals still are so I'll take it :3
okay wtf, I'm trying to cover an instrumental music I made, adding lyrics so it turns into a song, and it outputs an instrumental ? ô_Ô
Did you flip the switch?
(don't worry it's not copyrighted material, I'm trying to create the OST of a fictional game)
oh I'm gonna try something: put lyrics on the original track details
I'm not sure that I'd have any "aces up my sleeve" in terms of better prompts to use, but I notice that each of the prompts you're mentioning also includes "vocals" -- acoustic vocals, crisp vocals, unprocessed vocals, clean vocals -- perhaps comprising 50% of the provided prompts is impacting the vocals in ways that would seem detrimental?
Id remove all references to vocals imo
Any tip om how to make Suno not to start in a middle of a new track? Can’t understand why this is happening when using specific genres. Have tried to use the intro tag but that doesn’t help. It particularly when I try to create some chill tech music.. Never happened with other styles.
From my experiences, I had this issue more with v3 than I did with v3.5, and I'm having the issues way less with v4.
Can you share an example of it starting mid-song so we might be able to see if something in the style prompts might be affecting that?
I just using this style prompt: ambient tech house
ok it works but it's hit and miss, it will sometimes generate a full instrumental
I really like that one, I wish it would have the chorus follow the melody of the intro : https://suno.com/song/a6eea455-d862-4ace-b5d4-6a1a9d91c496
new jack swing, male vocals, upbeat, melodic, happy, jungle, tropical, rock, fast-paced, memorable, dubstep song. Listen and make your own with Suno.
Hello I just flagged my own song by mistake how do i undo it XD
can [Chorus with intro melody] work ?
This is where extensions shine. It may take several gens with tweaks to the tags each time, but I've always been able to recall the melody from previous sections... with enough credits.
This is from the second gen; the melody is there, tho maybe not as promenent as you'd like and the v4 shimmer certainly doesn't help. The next gen may nail it... or flop.
https://suno.com/song/22f1e430-439c-4643-86a7-ab3b4e6c16f5
new jack swing, male vocals, upbeat, melodic, happy, jungle, tropical, rock, fast-paced, memorable, dubstep song. Listen and make your own with Suno.
welle, that seems like a no lol
I reffered a friend and never got the credits
If it's something from the ... > Report > ... menu, nothing to worry about
You'll need to email billing@suno.com -- Discord doesn't act as an official support channel, especially for account & billing related matters 🙏
alright thank you
I don't see how billing will solve my issue :))
I came here to ask you guys if I am making anything wrong.
Your issue is related to not receiving the credit balance on your account, right?
You invited a friend using your affiliate link, they created a new account & you both made 10 songs, now you're waiting on the credits?
this is nice, but what I want is the singing to have the melody (the one we hear right after the chorus), I specifically wrote them so they follow the syllabic structure of that melody lol
I'll try to remove thos dashes, maybe it will be better
extensions... 
...and credits.
I misremembered both parties needing to create 10 songs -- looks like just your friend needs to make the 10 songs.
That's all that's mentioned on the invite pop-up. As well as you can only gain a max of 2500 throguh this method.
Well, that's all the user-side technical issues sorted.
We could speculate further by asking whether your friend successfully used the link you sent them, or just signed up on their own after being told about the site, but at the end of the day, the only people who can grant the credits or investigate the account-related matters is the Suno Support team.
support@suno.com and billing@suno.com go to the same people, but understandably the support queue is probably a little bit longer since they field general inquiries.
I was thinking that since it's billing-adjacent, you might get quicker results with billing@suno.com
Hi guys, is there any chance to fix a song that I really like, but came out cut in the finale?
Get whole song button still not working for weeks
extend
yes, use the "extend feature and paste the lyrics it's missing
trust me, it's working
That's absolutely not a system-wide issue -- it might be related to that particular song?
I'd recommend contacting support@suno.com if using "get whole song" doesn't add the stitched "Full Song" to your library, especially if the issue has persisted for multiple weeks.
The problem is that is cut in the instrumental part of the ending... when I tried to extend, Suno gave me a totally different song O_o
I did, not even get a reply
You can use ... > Create > Get Whole Song to stitch that extension onto the original.
i'm a user but you can DM me and i'll help you out
got some experience with stuff like this
Oh... I didn't see this feature. I try. Thank you, I'll tell you if it works
Thank you, too
Same
if you are haveing style issues, DM me, I may be able to help
Thank you so much ❤️
We can also help right here in the channel. 🙂
I have a discord for creators, we are a small community where we share helpful stuff to make suno songs into art pieces
like tips and tricks for prompts, style inclusions, personas, and more
Cool!
I'm in, if you give me, then, in pvt 😄
For help, I'll be here, anyway... I think I'll need
Many times xD
Am I going crazy or isn't Suno always $10/month? I'm not seeing any discount here
I'd actually get a year of it for 40% off
Annually -- click "yearly billing"
that doesn't look like 40%
I agree, but that's not really a technical-discussion, but I agree that's misleading.
I don't doubt that you're experiencing issues with it, but I attached a video to show that nothing is bugged with the feature in general.
If it's been over 5 business days since you emailed them last, I'd send a follow-up in the same email chain -- otherwise it might look like a separate request. Hopefully the support team can look into the matter sooner than later
then where would I discuss it
There's a couple discussion channels on this server -- #🗣️┃chill-chat, #💬┃general-chat, #1308499132356362383 is practically the wild west.
I mean, discuss it here as well if you'd like -- just didn't think it was a technical matter that users could help each other with
Meh, since we're already on the topic, I can at least share that if you have a monthly premier plan, that's $360/yr billed monthly, swapping to a yearly plan would be a 40% discount bringing it down to $216.
Still, still "save 40%" and "save up to 40%" are VERY different sentences, especially when it doesn't say "read here for more details"
Hey guys how's the shimmer effect? Has it been fixed yet or mentioned?
for me there's no sale, it's 20% only
#📢┃announcements message -- any further updates are likely to be mentioned there or in #🚀┃updates
Not sure if it's a regional thing, but you might just need to refresh the page?
Getting this too
same, refresh not helping
Suno - messed the bed. Same here.
Yup, fixed now.
😦
question, my renewal date is 15th, how do I get Black Friday discount
Heya any song upload enjoyers here? Clip has to be 2 mins but I have several parts that I want AI to pick up on. Is there a way to get my whole song in or do I just pick the best parts?
So, here's an interesting question for you all... will Suno ever decide on a non-integer BPM?
I've calculated this one I'm working on a bit more down to either 87.5, or 175 BPM, and for removing a bit of silence in an 'all-rest'/'false ending', I want to keep it on-beat
🤔 I guess there are some technical issues?
I think it would be wonderful to see in a future version the option to merge any of your songs together.
Has anyone else gad s problem with v4 cutting songs off at 3.20. Last 10 songs have done that.
If you're using cover or persona, this can produce shorter songs.
There really needs to be a better way to organise the songs. Folders or something. When adding songs to playlists. it doesn't move the song, and you can't even tell which songs have been added to playlists from the main song list.
Good news. My account should be fixed. Hopefully by Monday I should be able to regain access to all the songs made long ago at my old account. Hopefully it'll be fixed by Cyber Monday so I can gain access to My old songs and I'll be able to subscribe to the Black Friday. Deal that's going on currently.
Would a cheering group sfx be considered voice on an upload?
Probably. Try it.
Wouldn't be surprised -- even digeridoos can be flagged as false positives because it can sound like human vocalizations.
Then again, there's false negatives that make it through as well
electric guitars, too
I've been begging for nested songlists like Udio does...but it seems like it just falls on deaf ears. In the world of dev work, this is what we call low hanging fruit and could probably be done in a 3 week sprint.
v4 is not really usuable in my humble opinion. too many issues. it cuts off all the songs, never really ends a song. (its rare for me at least) and the audio quality with the issues (shimmer, wobble, distortion) is too much.
I don't think that it falls on deaf ears, I just think they don't give it the priority that we wish they would.
UI should always be a priority on a new application, hell even on an old one. we still update our UI based on User Exp on our game and its 10yrs old
Maybe this fits here, but I don't know. One of my songs disappeared. It's not in Trash and I didn't save it to my laptop. Is there any way to recover it?
that happened to me and I have no idea how to recover it.
Sorry to read that, for both of us.
its a bummer man. it just vanished. who knows how many other songs did as well I've got like 4k made so far. most are trash though.
This one was one of my faves.
but I mean. that's expected. i've got over 10k midjourney pieces but more than half are throw away, its how we learn (me and the AI)
I want to take advantage of the Black Friday special, but it appears Suno doesn't accept Paypal. Is there a way to use Paypal that I missed?
Does the black friday special appear to you? For me it is still 10$ and not 6$
Well, I managed to pull it up and play it using the link I copied. I can get an add to playlist option, but it doesn't seem to work.
damn I should have thought of that before I cleared my browser cache
Yes, I see it. Maybe a location/time zone difference? Maybe you're viewing the subscribe page from your browser cache?
I wonder if covers will be extended to 4 minutes? It's nasty when I have a 3:50 song but either cuts off or is faster when covered.
What are the copyright rules for songs shared publicly on the homepage?
Would anyone know if apple devices will be included in the Black Friday sale, or not? Any and all help is appreciated! 🙂
Immediately after sending an email about this to support@suno.com, I checked Trash again. The song was suddenly there at the top of the list and I was able to restore it to my library and save it to a playlist.
I'm trying to download .wav file but getting download failure message. MP3 seems to be fine.Anyone else having this issue?
Hi Rose, i am getting failures for download wave files as well.
i dont get an error, at least on MS Edge, it just does not download at all.
me as well.
I just tried MP3 and it seems to download fine. But not .wav file.
I'm using Chrome
I can't find one that used to be in a Playlist. I have the original and a few extends to make another, if wanted. But the final song-- nowhere to be found. So odd. It was just after a downtime moment, if coincidence. I hadn't downloaded it yet. Oh well.
sad thing is, the team is off until Monday. soooooo. it's up to us to help us.
wave download is broken in all browsers i have tested.
Hmmm that is sad
It was limited to 3:14 until yesterday. Now it can be longer but it apparently depends on the original song being covered or used for persona. Given that gens are limited to 4 minutes, uploads get extended to 4 minutes minus the upload length etc. I expect there is a hard max limit of 4 minutes for everything at the moment.
Yes I learned that in a hard way yesterday that the songs are limited to 4 min, as my lyrics were not even completed when the song just cut off
Wav dl worked for me. Just to give you a comp.
you can do an Extend @bitter minnow to get the song to complete.
Even though they gave 3000 characters allowance, but mine was only about 1800
Thank you!
Oh it does? What browser are you using?
Chrome browser , but on android phone. I should have appended that info.
some songs simply will not download with wave still
Thanks. I'm using Chrome but on PC. I just tried Firefox and got my .wav file. Chrome worked for mp3.
Thanks to all those who responded. What a great community here! Not sure where you are, but it's Thanksgiving in the US. Wishing you all happy days!
With wave still... then make wave move. rim shot A Dad joke. Carry on.
It seems that there are many of us with the same problem of not being able to download wav files, does anyone know if there is a support email to expose the problem to the SUNO technicians?
Funny "carry on" is in my song that I just created and am trying to download
Chrome didn't work for .wav download but worked for .mp3. But I just tried Firefox and it worked with .wav download. So you may want to try that too. I use PC
Yes, I use Firefox and Chrome on PC too, but nothing works for me
It may be a skeleton crew available with it being a Holiday today, and so email may likely take a couple of days, but then it could improve by then. But support can be emailed support@suno.com
This channel is just to share knowledge among members. If a bug or issue, you can report to the bug channel
https://discord.com/channels/1069381916492562582/1082775596767117312
Now you have me singing "carry on" by Kansas.
did you find your song? It took me a while to figure out that all the (accidently) "deleted" songs are in trash bin (look for the icon on the top right). There all the songs you ever created are there.
My song is 4 min but I just got my download
That's been a common glitch since I joined 9 months ago. I get it often when clicking a song in my list. I have to click the previous song first, let it start playing, then click the adjacent song that i really wanted to hear. Just refresh and try again. I've gotten used to it.
I see, thank you
haha... Where are you in KS? I was in Whichita for 1.5 years back in the 90s
No. KANSAS is a prog rock band (from , hmmmm.... 1977?)
carry on....wayward son
But my brother and sister went to Freemont, Kansas when they were 15 on a church retreat. So we can keep going. Lol.
Love you all everyone! Very thankful to all your kind help and advice! Continue to be creative and shine!
Carry on my wayward son. 🎶 🤘
Anyone else getting this weird wavey effect when you remaster songs?
Everyone is 👀
#📢┃announcements message they're working on it, though
Honestly, I did a few yesterday and today and got a wavy effect, I would relate to a demagnetized cassette tape sound, always at the end of the songs. My songs are a bit period dated sounds, the wavy bit doesn't hurt them, but yes, noticeable.
It's mostly distorted guitars that are getting it on my remasters. 😭
If it's at the end, you can crop the song. Sometimes it adds weird sound mostly at the end, but at times in the middle.
Blashpemy!! ::cough:: ahem... Sorry. But we must preserve the guitars!! 
It succeeded with Two sets of cheering and a set of screaming
I use Suno for music, but I'm in real need of a tool to read my scripts naturally
I'm trying to use Suno
::C bl- O -as- U -ph- G -em- H -y !!
I'm starting to get some ressults
Dialogue, storytelling, reading, script. song. Listen and make your own with Suno.
but in no time it gets to singing mode.
But I can see this being an epic voice actor
the acting is epic in some lines: https://suno.com/song/00e22092-e174-41d6-abcf-0fe009476ee8
Dialogue, storytelling, reading, script. song. Listen and make your own with Suno.
very nice! We can do some cool things with spoken word, tho Suno often needs a reminder not to break out into song at the wrong time. Here's one I did with a couple of "characters" from the Discord. https://suno.com/song/75f07f11-cb9b-4b1a-8e8f-1d10a944ac25
Dark Horror Poetry, Spoken Word, Glitch Audiobook, Creepy Male Vocals, Eerie song. Listen and make your own with Suno.

lol
I check what commands you used
Theh funny part is that it feels like some sort of musical, it actually sounds quite decent, gonna check yours
It becomes a musical. 😄
Yea... it would be cool to have suno spin off a script only tool. But then the background music is what makes it too.
they're removed from that one... but not dissimilar to the tags in this track: https://suno.com/song/135504dd-0dc6-4ec0-9d75-dbbdf0118622
Dark Horror Poetry, Spoken Word, Glitch Audiobook, Creepy Male Vocals, Eerie male vocals song. Listen and make your own with Suno.
that would be awesome
or just more tools for incorporating spoken word and sfx
dude, nice ressults!
I see your's also gets into song mode XD
heh. That was intentional, at least.
I guess the trick is going to be to trick suno into thinking somg comes later
the intro of the second is epic good
this could be an insanely good narrator for video game and such
I kept this one with spoken word and sfx https://suno.com/song/173ac910-bc00-4c67-8834-ca70eac7de79
that's actually part 1 to other one I posted that switches to song
dude, with all the SFX built in
I got the wind effect with [*wind*], which (at the time, at least) only worked in like 1 out of every 10 generations
I'll be experimenting with theese , Spoken Word, Glitch Audiobook, params
hi.. when will suno ai update their black friday sale on ipad.. i want to pay with "apple pay" plz help how can i take this offer????????????
hey, i have little problem with the voice, i want female voice, and i still get the male voice insted. Even when i write only one "female singer" or even when i write this "female of mysticism, trance, hands up, female Vulnerability in speech, female ballad, female Minimalist expression, female, female Emotional depth, heartbreaking, extra slow, hands up female, female, " do you guys have some ideas how to repair it?
Any swedes (or perhaps 🇳🇴 / 🇩🇰) on here have tips for pronouncing the letter "y" in proper Scandinavian? It tends to sound more like our letter "u"... :/
Just did this "live" intro.... I've going to try more. They're fun. I did about 30 iterations, this isn't the best -- the best actually has some audience rabble... but it the cleanest song of all. 40 second sample attached. https://suno.com/song/6723b56c-1504-4d4d-9b23-b73faee2d08b
This is my new duo (they're actually partners as well) The one on left is Janis, and Im working a solo album for her, but they both speak in this clip.
Everytime I listen to this song, Suno radio plays a Death Metal after. Haha.
btw V4 and V3.5
Have you tried Elevenlabs? I find their text to speech models to be excellent
sound nowhere as natural as in Suno, IMO
they sound good tho. But a bit soul less
in Suno it feels humanlike IMO
I've a couple of saved audios files from elevenlabs, I share
@woeful grove I sent you a DM with the intro for part 3 of that horror trilogy with all the tags still in there.
tx, I check
Your best bet is to report it, as many have, to the language issue channel. I say many have, I mean, everyone is detailing what the issues are in their languages. So it's a helpful channel to keep updates. https://discord.com/channels/1069381916492562582/1308617723328204880
I think the prompts without "female" are good prompts to emit a female voice, so those are good. From my small experience using Suno, id say it may not be needed to use female on every prompt in styles, assuming the more the better. I kind of think the opposite. Each prompts ,being weighted, means "female" may be deducting than adding to the prompt. So, a [female voice] makes sense, and a "happy piano" for instance will get a bouncy playful piano..... but a ballad is a neutral prompt, as is expression and depth, so "female" might not be recognized and take away from prioritizing those prompts to be as influential to get a female voice.
I would have female vocals in the style among these other prompts, but take out the rest of "females" and focus on pre-emptive metatags for each line and stanzas you want female in the lyrics. At least to compare results.
@naive quail. I have a theory. I've read another bring this up as well. It's more karma based, but I feel if I create a gen, and it's wrong, such I get female instead of male vocal.... Suno will respond better -- more attentive/corrective -- in the next gen , if I let it time out, emit a time stamp on the present gens . So always about 1 minute's wait. Then click create again. It's like, it needs to dump its "karma cache" in the recent job just done else it will tend to repeat the error on the next gens if done too quickly.
The second I do with covers and I see others try as well: if covers keep repeating I throw in an "opera" cover to shock suno and then either do a cover of the opera one with the actual prompts I wanted, or just do a 1st round cover again, having stuck in an "opera" one. Then as Opera doesn't work, I go with "death metal". Either of those two, seem to make a noticeable change.
What do you think about the first point. Things just seem better on the next gen if I just fully wait (dont play, dont click create, etc) for the present gen to "cure", as it were.
i using it in style of music, when i use in "lyric" [female] or [female voice] or [female sing] it's same. when i use it in "lyric" in backgroud is (female "ohhh") but when it ends, male vocals arrive
Does the community have a term for the strange high pitched shaky engine like artifact that appears in most tracks at some point in the background?
Likely referring to "shimmering" -- as noted in #📢┃announcements message, the team is aware of it & working on a resolution as we chat 🙂
Cheers! Just wondered what keyword I should be looking out for in announcements.
I could imagine a world where they decide to leave it in as an audible watermark and charge for versions without 😛
I'm glad you gave it a try. It's not exact, and even as a premium member I have times I run several before it works. It's a balance with the prompts, the order and placement of your prompts, the prosody or theme and ideas presented in the lyrics, the metatags, and other fCtors, all the right balance to get a good results.
So, I can use up credits to test things such as: just keep your lyrics and tags the same, but only use "female vocals, kpop" in the style, to show that's its not the lyrics that's are pulling your song to go male. But if you can not test various ways, then ask around more to see what you can do.
There is always compromising what you think you want to finding something nice that works, that you can be happy with as well. So go in in a different direction on the styles maybe.
Also use, "female-vocal-prominent" , " female-vocal-priority" .or "female-dominant" in style prompts, with the dash marks. Although that last one might gat an aggressive voice due to the varied meaning of dominant.
anyone else having touble with downloading WAV files?
Does anyone happen to know if apple users will also be receiving the Black Friday sale along with the web users? Any and all information is greatly appreciated! 🙂
Interesting. I'm going to play around with this over the weekend.
This doesn't make sense as an intended behavior, but weird bugs are possible.
Oh this makes perfect sense though: "if covers keep repeating I throw in an "opera" cover to shock suno and then either do a cover of the opera one with the actual prompts..." that for sure will be very different and "if your cover isn't changing enough, try Cover of a Cover" is good general advice
I'm hoping now that they've introduced black friday sale, the fixed V4 will roll out.
Wouldn't imagine the two are inherently connected
On the 22nd, they mentioned "the model you have in two weeks will be better than the one you have today" (#📢┃announcements message) -- so that's when I've been eyeballing 🤞
Another mentioned that songs withing the same create page can have a reflective quality. Ghosts in previous gens sneak in to future gens which should not be. I'm not sure about , same page, hypothesis, but I've said before I could be working on a Christmas song, but change tracks and just do something else, and , as a made up example, sleigh bells will appear 10 unrelated gens in where they haven't in months before. Or the word "christmas" show up in a hallucination.
So , I'm trying to do more refreshes, close and reopens, and other neurotic moves to both give me more assurance and also persuade people to walk on the other side of the street.
Maybe make sure Chrome isn't autofilling Exclude Styles or something
Thank you for sharing! Didn't realize Suno can do the dialog part. Sounds really good with a mixture of dialog and music!
I was liking mine back when 4.0 started.... but the recent works are less likeable. A few (3?) have recently gotten a cassette tape warble affect not heard before at the ends. Similar to listening to an old cassette left out in the heat or demagnetized. So I'm feeling it's more going in waves than a cummulative improvement. Dam a river, you can flood a village. Fix one thing, it might affect something which worked fine. It all a process.
Yes I reported this earlier when I couldn't download WAV files from Chrome on my PC. Later when I switched to Firefox on my PC, it worked
It always gives me a chance to prep/write more songs to guitar and keyboards to have them at ready.
Why does Suno completely change my instrumental upload when I lay some vocals on it? It changed part of the beat and took out 90% of the sfx
good day, may I have a question for the developers, I have an important question that should only be answered by SUNO developers or someone who manages the website and registration and uploading music / voice
@untold briar is there any dev from suno or if they could answer my question here?
Firefox mobile gives this error
same here
same!
Hi all. Do you know if there is ANY chance to get folders in library anytime soon? To me this is the key game changing feature and I am losing hope already if that ever will come....
Very occasionally devs pop into chat but unlikely to answer specific questions. Best chance is go through email support@suno.com
it is within the framework of Europe and GDPR
does SUNO meet the conditions?
Does it comply with the laws and protect the user according to EU law, is everything subject to directives and EU law or only US law?
Foxfire PC browser issue
I don't want to start an now create email just because of a question
same here
i have had 3 days straight "a server issue occured", happens when i try to create something.. i have cleared browser cache etc so what to do?
Discord mods and guides are not Suno employees, but email will go to Suno employees, so it is best way to get official answer.
i tried even with konqueror. Same error
sent email to Suno but is here any employees/support/admin?
We can take guesses why something happens.
For instance, suno tends to "clean up" what it thinks are noise. Any sound effect, to me, would not relate to a comparably trained instrument sound and may get "cleaned". Any gritty, vocal fry or unique voice seems to get "washed" more.
As for uploaded themselves, I can't say Suno duplicates uploads note for note musically anyway. Yes, it follows melody, but not so much the instruments that played it. So much may be lost in the process of adding additional data such as lyrics on top of that.
Again. All guesses until you get an actual answer.
Just tried on Chrome on PC (as well as foxfire) and got the same error:
and i started to have this just after i paid and subscribed so really pissed about this out come and user exprerince
I've never had and issues with SUNO, im sure they will fix it asap once they know there is an issue
i had this as well, went off after cache was cleared
i think it happened cause i force try all browsers and different log in options
Suno worked like a charm until i paid them 😄
Can you at least answer my question?
it is within the framework of Europe and GDPR
does SUNO meet the conditions?
Does it comply with the laws and protect the user according to EU law, is everything subject to directives and EU law or only US law?
I think support should know this information
I've gotten or get this error often a week for months. But not to much consideration. I refresh , close/open or wait. Things seem to normalize quickly.
Firefox on PC works for me though
Thank you, I closed and reopened a few times and now it's working. I think they were updating the site because now there is a "What's New" and +1 for notificaiton. Thank you all for the recommendations!
I do not care about this to be honest
I bet they had a huge influx of new people with the black friday 40% off sale
u in EU ?
therefore, the laws are different from the USA and the EU
what may be legal in the US may be illegal in the EU
on the contrary
right, support is through Suno email.
If no member here has that knowledge, as this room was set up for chatting the technique side of "technical" than the actual technical (or legal or policy) side, then it might take the other channels or email anyway.
or I may consider contacting the European Commission regarding a possible violation of the GDPR and other European laws a they may wonder if it is true or not
it is technically because within the framework of the GDPR, users must not have access to the US and EU within the same server
these guidelines are from 2018
yes. still don't care 🙂
can you share what is your issue? this is a music app. why this topic here?
so you know that every EU citizen has the right to file an objection to any company that somehow operates in the EU or is allowed for European users to connect to the given service, so they must meet the conditions of
Hillbilly Death-Metal Big Band song. Listen and make your own with Suno.
Yes. I follow. Excellent question. As a legal guy I'm always curious about things like that.
In mine opinion, I did not agree with the last amendment regarding the GDPR inside suno
are you a lawyer
so if they were watching them
so we'll let them check again or we'll link to the GDPR page opposite here
because I can't find the GDPR and other amendments regarding the EU at SUNO
so I take it that they are not and were not granted
therefore I do not exist, therefore I assume that it is violated
the second thing is that the database users within the USA and the EU must not be connected
Unfortunately, this is also covered by the GDPR directive
you mustn't
You must not be a US server when you give the SUNO product and AI models to Europe and sell them
so they would have to have servers in the EU, but one admin wrote that SUNO is purely a US company
I won't argue, I'll be surprised at the servers I reach, maybe I'll find something new or interesting
LEGAL !
Yes it sounds very good. The composition itself is very nice. But I was trying to get dialogue only. As of now is very tricky. In Japanese is even worse as will start singung right after a few lines no matter the commands or style.
That can be a question or merely informing them of something they might not know. Either way, I would think it could be quite appreciated to know.
I can send an objection to the EU regarding the investigation into GDPR compliance
European universities find out that it is not, they will receive a similar amount as Facebook received
Is there proper procedure for filing these objections?
on the website of the European Council, you can click on it, it's hidden somewhere, there's an email where this is sent
it is done by taking screenshots of both the website and the service as a link, and so on, they will pass it on if the client's claim is met with an objection, and then they will ask SUNO for a statement, and if they find out that the servers are only in the USA and that I am selling the service within the EU, then it is 100% of that it will be like facebook
I expect that SUNO will be disconnected from the EU until it reserves its servers and rules under the EU and GDPR
🙏 I hope you hear back soon, however the matter plays out.
Is this something for which you're requesting assistance from other users?
I'm talking about the user database
their files and so on
Europe is such a thing that the files of the US and EU residents are not allowed to connect within the framework of the GDPR and so on, they must be transparent and I know that I had to change my database because of the GDPR
This has nothing to do with technical support lmao. I am also in EU, write in to their support stop babbling on in here.
so we support could throw a link to some GDPR
I meant passing on this information to suno would be informative to them and likely appreciative. Their legal may have this information, but personally, I never mind someone telling me something I may already know.
so you probably know that the GDPR should be visible and it should be possible to read again at any time what your rights are and so on
yes, everyone of us who lives in eu know about gdpr. 90% of US tech companies know about gdpr and data residence. you're not having an epiphany here. if you are concerned write in to support. stop spamming here
it's like someone telling you to do what you want but then judging you for breathing someone else's breath
can you see why I don't have them on the website?
Have Suno fixed so the black friday deal works? Or is it still the 10$ issue?
probably a 404 error, call the moderator, maybe the HTML table will list it
Not at all. I'm quite ignorant to the details you've proposed and find it interesting to read to the point of a concern.
After which, there's nothing more to be gained in a casual chat room about techniques for creating songs.
I hope I hear more about privacy laws and the process. When I have time I'll do more reading. That's as far as I'm willing to discuss here though. Good luck.
are you having a stroke? are you an ai bot?
While we can appreciate your concerns, this really isn't the place for them. We're just Suno users (like you) discussing the technical details of using the service. You really need to contact support@suno.com to express and hopefully resolve these concerns.
ok, but I plan to ask about the enamel as well, but if I find out that I don't have anything, I will file an objection with the EU
i really ? me english by good ?
I don't know why I'm arguings with other me
I will write my framework and of democracy anyway
I'm just afraid that SUNA will be disconnected from the EU, nothing more
hmm
@worthy sleet What has been your style of music to write in Suno? Being in Europe , others tastes in music is far from mine as a USA guy, of course, so it's fascinating to hear songs from around the world.
well, my friends found something called the golden treasure of system errors
Dang. Sounds wicked. What's that?
but it's a classic that they tried to recruit for a couple of months until they succeeded
Ok. I'm a bit confused. Is that a style of music. A new technique. Sampling?
@polar spade You're a saint. 
I will create test music in the suno file
and audicraft
I doing it like a mini battle
When you get one done... post it up in here... that's definitely a "technique" that we can learn from. Cool.
did you know that if someone gives you permission to use music and then someone claims it, they do not have permission to continue using it, for example a service (SUNO)
Oh.. well, at least post the suno version, even if it's a rough draft.. I like to hear just what we create from the suno side before one takes off site and uses a DAW or processor (Audiocraft)
Yea. Strange huh. Lots of YouTube vids about that. It's a wild world out there.
if it is taken, I have access to the dataset which is 2011-2024
so I have the possibility of training on the contents of the data
the only thing holding me back is money
I'll tell you a secret, do you know that Google lost the court house in the USA? 😄
Well the best plans are those kept hidden. I'd say don't give away too much information here.
BINGO 😄
Have I understood that correctly? So to follow the chosen persona as closely as possible, you need to set the styles right?
was it on me?
No. @azure crane is starting a new question.
im ask DC why write you ? 😄
Yeah, I just wonder how personas really working. Have upload an instrumental that I’m using for a persona. Now I try to figure out how to use it correctly.
Basically, yes... if you load a persona... the persona's styles usually load in prompts as well. But if you then manually change the styles in prompts , you'll get mostly a genre change. The persona's should stay the same or similar.
Having said that, I have noticed by changing prompts let's say, from folk to disco, then can be some variations in the persona's voice which don't match up all the time. That's just anecdotal and personal to me, so most times you should still get the similar persona style sound in their voice regardless of what you put in styles prompt.
If I want my folk- persona to sing my disco song in her folks style, yes, then I will make sure my prompts are back to the original folk-persona prompts that are in the persona's data.
Ok, but the upload doesn’t have any style so to have the persona working you must add a style that the persona sounds like. Seems that Suno doesn’t find my style setting similar to the personas feel and sound. I’m only working with instrumental music by the way.
Oh. Ok. It's unknown who one responds to unless they also show the tag of the comment. I would suggest to right-click and choose reply, so that one knows who you are responding to. Otherwise every comment is open for a response by anyone.
Personas seem to keep a small chunk of the original song as context to use as reference -- It's sort of like extending from a clip, where you're allowed to put whatever you want in the style prompt section, but usually keeping it the same will give the best chance of replicating that same style.
In your instance, being that the original clip didn't have style prompts to begin with, I'd be curious if leaving the style prompt section blank might keep it on track?
Ahhh... one issue though. Suno doesn't have a way to add a persona to uploads or covers, so you won't find that choice to add a persona on an upload.
Let's say you uploaded an instrumental and want to add lyrics for suno to sing. In this situation, you can NOT also add in a persona style. It won't let you.
Yeah, but you can’t create without a style prompt so that is impossible.
there is a reaction if you are react to see her there again
You can -- just add anything to the lyrics section -- a period, a backslash, a space
Just noticed that right now 👍.
I mean, Suno low-key insults you for doing it, but it's a viable method 😉
sad
@charred merlin doesn't know the word "impossible".
The result was totally different from the Persona 😅. Hmm, have to nail the persona style then so Suno creates similar music as the persona.
Sometimes Suno creates exactly the same music as the persona.
Ok. That's my time at the Suno comedy hour. I'm here the rest of the week. Try the veal. Don't forget to tip your servers. Happy Thanksgiving y'all! Drive home safely.
Am I the only one witnessing a specific rattle in the background of every generation of a song?
If no, why is that? Its quite disturbing imo
#📢┃announcements message The team is aware & is fixing it
Thank you for linking the announcements to that!
Not a worry -- any news will likely be shared in #📢┃announcements or #🚀┃updates -- you can turn on notifications for those channels so you're the first in your friend group to know 💪
Are the 10k credits you get here permanent ?
Like, can I buy one month now discounted and then use them at a later month when the v4 fixes roll out, or will the credits be removed once the month ends?
Monthly subscription credits
They aren't extra top-up credits that roll over -- the premier plan receives 10,000 credits every month.
They don't stack, and reset to that amount at the start of the next billing cycle
So I have to use 10k in one month for it to be worht it?
Yes
tbh -- this is the first discount I've ever seen from Suno. The top-up credits are the most expensive way to use Suno -- it's a convenience to have everything on the same account, but it's remarkably cheaper to pay for a second pro/premier subscription for a month (you still own all of the songs, and it's only against ToS to have multiple free accounts -- just make sure you don't forget about it and get charged for months you don't need)
Let's say I buy the cheaper Pro Plan to get "Optional credit top ups" will there be any worthwhile top ups I can buy to make up for the 6$ I will waste on the subscription?
Or for that same $30, you could get 10,000 credits 👀
I guess the answer here is no 😂
Thank you for the help!
You're most welcome 🙂
"Caveat Emptor" -- I have a reasonable idea of how many credits I'll spend per month, and use Suno pretty regularly, so any extra costs seem frivolous to me, but for someone who uses Suno in bursts & sprees, it might be worth the extra cost to buy extra credits to use months later -- subjective preferences
I can't handle time constraints so this subscription model is very much out of my ballpark
I completely understand that -- it throws a lot of people for a loop, but usually it's after their credits have reset & they're wondering what the heck just happened -- no small amount of Kudos to you for informing yourself on the setup before diving in 🍻
I like suno but I don't like how they will just take my credits away if I don't use them in-time, at all
How to fix it? Can anyone help me?
I can understand how it would help a still-young company better predict & balance their server load, but there's absolutely the complexities of human-life that don't always blend with business efficiency.
I'm hoping that one day they'll announce "credits will now stack from month to month" in a headline that will blow users away -- but until then it might be advisable to start off with a Pro plan & see if you're more commonly running out of credits or left with some at the end of the month.
Most times you just need to wait a couple seconds and try again 🙂
If it's persisting after a while, it'd be helpful to know if it's impacting every song generation, persisting after a browser restart, occuring on other devices as well, etc
I have been trying for 10 minutes but it is not happening
It's impacting every song
I also restart the browser
Please try for me
It doesn't appear to be a system-wide issue
If you're using a Chromium-based browser, you could try swapping to Firefox (or vice versa) to try another "housekeeping troubleshooting step", but if it's persisting with every generation, it might be worth contacting support@suno.com so they can look into the matter, as it might be account-related.
Is there a way to stop suno to repeat lyrics (starting at a random spot in the text) eg. verse 1, chorus, verse 2; in the middle of verse 2 it starts the lyrics in verse 1 again repeating and then continuing....?
Anyone able to get the expected 40% off on suno yearly subscription? I'm just getting about 25 ie. 72 euros from 96 original price
Try to add a [break] in correct places, mark your tags for verses and chorus correctly.
I mark them like [Verse 1] [Chorus] if thats whats ment... At which point I should put the break?
Appears it is -40% with respect to the monthly pricing. Yearly does not include the additional -20% discount, like one of those "does not apply to other discounts" things you see on coupons 🤷 (see #💬┃general-chat message for the math)
Hi, I'm new to this forum. Do you know how to remaster a V3.5 track to V4 without changing the harmony or melody? Every time I remaster, there are changes I don't want... I would like to keep the same melody and harmony.
Is it just me or is it slow to generate?
the producer tag is "Dear diary, it's me again" pretty sure
You could just have an unlucky track that doesn't remaster well, but if you really want it to work, sometimes if you reformat your lyrics in different ways, it can effect that and could help. Specifically try removing punctuation and capital letters, and overall making sure the lyrics closely match the real words that are sung in the song.
@charred merlin sir, I have an audio ready but I want an instrumental intro, so can I upload the audio and generate match instrumental audio for the same composition?
I am starting to think this “shimmer” is an audio glaze to prevent other AI audio models from scraping Suno to train on. All my previous gen songs that were clean now have the shimmer. Some have it so much in the mix that it is a distraction from the vocals.
Wow. I've never had it do that. When I click on create remaster v4 for a v3.5 track, it always just gives me the same song back with no changes, other than cleaning up the vocals. Same length, same melody, same lyrics, same everything. I don't understand why it gives me two remaster versions however, since there is hardly any perceptible difference in them.
It's not intentional. Suno are working on removing the shimmer.
Does anyone else noticed that the quality of the V4 on the day it was launched the tracks were incredibly perfect but right now it just seems to be a slightly improved V3.5?
Wow had a weird one, after a [break] with lyrics in the song, the whole thing cut off and start right from the beginning, it's a shame because I was satisfied with how the song was turning out....
Hi, i've a technical question, i hope to receive and answer. Is it possible as FREE to decide what kind of voice will sing a Verse? So for ex i want one Verse with a female and one Verse with male etc, there is a sort of prompt i neet to use?
You just need to distort the lyrics enough where the censorship isn’t triggered but the AI can still pronounce the phrase. Try Hiii… mee, izzz… mee ’gainnn…
[male vocals], [female vocals] tags in the lyrics should do the trick
Yeah, $360 if you pay for the monthly premier plan -- 40% off is $216, but "up to 40% off" isn't displayed on all texts, so it can be confusing.
But why would it be showing up on the 3.5 songs from weeks or months ago?
Does it have any use also use [Verse 1] [Chorus][Outro] etc?
It always has been there. Welcome the frequency illusion
I can hear it in my 3.0 gens, including a piece made ~2 weeks since its initial release.
The frequency illusion (also known as the Baader–Meinhof phenomenon) is a cognitive bias in which a person notices a specific concept, word, or product more frequently after recently becoming aware of it.
The name "Baader–Meinhof phenomenon" was coined in 1994 by Terry Mullen in a letter to the St. Paul Pioneer Press. The letter describes how, a...
[Verse 1, male vocals] should work, I think.
Maybe you can guide the AI further by putting "Duet" in the style prompt, but it's trial and error at this point.
Interesting.
If you upload the audio, it will be added to your library, and you will be able to "Extend" from it.
It is easier to generate new content after your upload -- it requires some extra work, potentially using simple editing tools, to use Suno to create an earlier section of the song.
It's strange, i use some thing like [Verse 1 – Female Voice] and it doesn't work, maybe have i to use the comma instead of -?
Tbh Suno lacks a slider to have more control over whether it's creative or stick closer to the prompts.
I've always used commas in tags
Kerrogann nailed it. Here is a playlist (https://suno.com/playlist/11ff3827-bdcf-4ef5-bd04-da8647c6cdb6) of the different methods (#💬┃general-chat message) I try to prompt the genders for duets
But it doesn't work all the time. In my experience, when it does work (~60% rate), half the time the AI swaps the gender roles
So is it better put the voice genre in the Lyric instead of the Style of Music?
This is how i write in Style: "A gothic metal sound blending distorted riffs, haunting melodies, intense percussion, and alternating ethereal female and deep male vocals, creating a dark, dramatic tension."
Duets are the holy grail in AI generation. Direction works and sometime it doesn’t.
I get good results when using style tags such as vocal interplay, dynamic vocals, duet harmonies.
[Layered Duet] sometimes works in V4 to get them to sing together.. at least from my limited V4 testing (~3k credits worth but not all for duet stuff)
Will have to remember to try that tag.
sometimes it only adds harmonies though and sometimes it does nothing at all... but so far it was never sung so can't hurt to slap it on top to maybe increase chances
Yes I had the same issues, and then realized the songs were cut off at 4 min. I liked the melodies too but had to shorten the lyrics and do over
I wrote the quartet parts out in the style and I got the result I wanted after only a couple generations. https://suno.com/song/957a0422-7c4d-4fa1-8c10-577a762132df
mixed barbershop quartet: female tenor male lead male bass and female baritone for rich balanced harmonies song. Listen and make your own with Suno.
But it took me many different genres to narrow into a barbershop quartet.
On V4 I have gotten layered spoken whispered harmonies of male and female. That was on a roll while using my “female breathy 01” persona.
Yea V4 can properly overlay... but still difficult to prompt reliably..
this is my coolest one so far but it did it not in the way that was prompted.. still cool though
https://suno.com/song/d589386e-efd7-4ac8-871b-e609748fb19c
melancholic garage, post-punk doom scape,doomed dark wave,apocalyptic cello fusion,british female singer-songwriter, mixed gender rubato duet song. Listen and make your own with Suno.
Am i writing right the Style?
When extending a song from 00:10 does the new extension start at the beginning of the 10th second? Meaning only up through the full 9th second is kept from the previous part?
Yes, unless you Get Whole Song on part 2 first and Extend that instead. In which case it uses up to the last 2:00.
so i've been feeding old demos into suno and seeing what they can come up with and i'm trying to figure out how to keep my voice, at least the tempo/melody, etc. I actually have a ton of questions, but lets start with that
i know its possible because ive done it
but it doesnt happen every time
for example
fed it this
got this
this is me playing around with 20 year old demos that I don't have the raw files for anymore etc
how long does it take to answer one email?
1 h / 24 h
1 day / 7 day
31 day
365 day ?
The two main things to think about tweaking are RNG and Prompts:
- **RNG ** just keep changing whatever you can change because even small changes can make a big difference. Different clips, different duration or crops of the same clip, different Extend From times), vary whatever Suno lets you vary.
- Prompts is well, prompts. But for example you might try Extending an upload with blank style, lyrics, or both. Or you might try Extending an Upload and having Suno first recreate a verse or verses that were already sung in the Upload, itself, and then sing new verses with your voice. And if that doesn't work for the new verses, Get Whole Song and Extend from after a recreated verse (a la RNG) and try again.
As far as I know, the team that responds to emails work from <t:1732888800:t> to <t:1732917600:t>, Monday - Friday, and presumably respond in order that the emails were received.
The turnaround time will vary, but usually seems to be within a few business days.
Thanks for the info, seems good to know. 👍
I created a persona of a song and i created like 30 song with this persona but with every song i got hallucination with the lyrics idk why, i lost almost 300 creds
The cello was nice on that one.
oh yeah this was an example of it working. still rough but it actually follows me. what im trying to avoid is what it keeps doing which is this:
where I don't hate the music, but the vocals are clearly not my melody/tone
Maybe later than usual since it is the US Holiday Thanksgiving right now.
Persona’s do not prevent the AI from hallucinating. It just might do it in the style of the Persona instead of a random voice.
Lol funny, Crush 40 and Johnny Gioeli are not blocked from the style prompt (although if the style is similar because I added "rock" to the prompt, the voice is not even close to Gioeli's lol)
Oh and if you don't know who they are, listen to Open Your Heart by Crush 40 or Live and Learn, the singer is Gioeli loo
Except here, the voice is pretty close to the real one, I just regret the chorus having a forgettable melody : https://suno.com/song/6335e02e-aacd-4a63-8726-0dd527713bd2
Hi, haven't use Suno to create songs anymore since V4. Where the heck is the create text button gone from the text prompt window ?
You need to click Write with Suno first before entering your prompt, then Write Lyrics when done. For a step-by-step guide, see #📚┃suno-school message
Thanks, but is does not work that way or in another way! There is no button ! And I cannot copy text there if I click first the "Write with Suno" button.
Recently I've been having some problems with the ending of songs, the song will end and out of nowhere the AI starts singing another chorus, does anyone know what I should do?
Type in the mini field box, select the model, then Write lyrics
my GOOOD ! What a stupid idea to hide there a text field !!!!! Who had this stupid idea ?????
I would have not found it in 1000 years without help! That is what I call now an example of the total opposite of an "intuitive" user interface.
🙂
Yeah, would have been a great idea to make the input field more distinctive, like a pink background or border
to place a mini textfield on the top of ANOTHER textfield that you know already and to place 2 buttons on it and an X needs some prize 🙂
and it is also super tiny !!!!
and the window has a super small font now !
time USA or EU ? 😄
it is divided when it is - 7/8 o'clock or + 7/8 o'clock
The new lyrics generation also caps the input to 200 characters, which means you can't use the lyrics model to do some things that used to work when the lyrics box itself was the input. For example extend an existing set of lyrics or to rewrite lyrics in a new style.
The timestamp shows in your local time but it is approx. 0900–1700 EST (UTC-5)
https://suno.com/song/da55b4b6-5f12-4824-a4f3-b72a72b38c53
Finally got a satisfying result after 1300 credits burned for this song! It's not Gioeli's voice at all, but the style fits perfectly for what I was trying to achieve, I love it!
Crush40, Jun Senoue, anthemic rock, dubstep, upbeat, fast-paced, happy song. Listen and make your own with Suno.
I'm I in the right place to ask technical questions about suno?
Because Suno's training should wipe out references to artists, singers, etc. to avoid breaking the law (mainly things like impersonation which is why the ELVIS ACT has been enacted in Tennessee).
Try describing the voice instead like [Male: gritty, deep], etc. (but it is hard for the AI to get the tone you want...)
Great! I've made this song in Version 3:
https://suno.com/song/ec6600c2-f987-4e22-8644-d82e67223b2f
I had to modify lyrics so it would sing the words "I.T Guy", but I'm still having a problem with the word glare. I changed it to stare, but ever time I remaster that song using V4, it never can pronounce glare or stare. Here's a V4 remaster:
https://suno.com/song/1ed19e04-534d-4f07-86e4-422f0f68059e
Any suggestions?
rock electric song. Listen and make your own with Suno.
rock electric song. Listen and make your own with Suno.
No need now, I'm happy with what I have. It's not impersonating Gioeli, but it's great nonetheless!
The problem lyric is either:
"her glare could melt the concrete"
or
"her stare could melt the concrete"
Suno can't seem to pronounce either.
I was hoping someone her could advise. I'm not familar with discord, so I hope I'm not out of line asking this here.
I'm very old compared you guys most likely, so please be kind and helpful 😉
In the future, it would be helpful to provide a time stamp to the issue (01:07 in this case).
Sounds like the AI is skipping the word due to flow/speed? Not so much as mispronouncing it (as can be heard at https://suno.com/song/18fa7805-6888-4a84-b954-aa34748e3f80 where I was testing the pronunciation, chorus at 00:45)
Hmm, funny, I swapped stare for stair to see if the flow would skip it but they worked fine:
The song lyics somehow converted english into panish lyrics even though I had an english prompt written down
Hm, using stare normally worked fine as well
- 01:18 https://suno.com/song/bc14a7d8-8093-4233-90be-3be5edb88894
- 00:56 https://suno.com/song/0732f0d0-7c43-4de9-a845-91257847fd8e
Guess the issue was the flow in your faster paced song? 🤔
if you guys can help that would be greatly appreciated
Are you using ReMi? It can do that, it's unhinged
uhh how do I uncheck it?
Trying adding "in English" at the end of the prompt
I cant seem to find it
Change from the ReMi model back to Classic by clicking the model name
Well, if you like Simple Mode, you can state something like:
In English, sing about horses and hay
For more fine control, enable Custom Mode
To generate lyrics in Custom Mode, see #📚┃suno-school message
You're welcome, hope that helps 🙂
does it use creddits for generating lyrics before making the song?
It should not. There may be a visual glitch showing deductions (I think there is some spaghetti coding) but refreshing the page will show your correct credits (well, after your generations are done)
Hello where can i get help, i pushed the "make persona" and all my old music disappeared 😨
Sounds like you may be in the persona list. Swipe the options to the right and you should see the option again (Or go back to https://suno.com/me)
Did you create a song with those crazy lyrics?
Aah i found it, there are multiple rows in library, but something happend i cant scroll all, but ok happy they are still there 🥳
you might try glair or stair. I have found in some cases when Suno AI has trouble with pronunciation that changing the spelling of the word to be phonetic or a homonym of what I am looking for helps.
I forgot to also mention about the pagination but I assume that is not the issue now :x
Question / re: Replace Section. I did a replace of a line. It worked out fine.
Now I'm trying to replace in the same song, using the Final song of previous replace, a line near the beginning. But when I highlight the area (drag over the graph) , no lyrics show in the replace lyrics box and no lyrics highlighted to be replaced.
We can replace lyrics more than once on a new full song from the last replace?
If so, I'll give it time and come back to it another time.
I had the same problem a few days ago, as well as a few others. This seems to be a bug. The less-than-ideal workaround that I have found is to extend the song (I just add onto the end of the song and put [End] in the lyrics and it usually adds little to nothing, and you can crop it back out later if you want), and afterward the lyrics will show up again the same way it did the first time. This has worked for me on a couple of songs, although no one else has confirmed it works for them - if you try it and it does would be great if you could confirm here for future folks with the same problem!
Oh yea? Thanks. Put a step between a full song from a replace and a new song from an extend. Cool. I'll try it tomorrow then.
I am trying to make symphonic metal songs, but I find that many generations have a particular noise that is challenging to describe that sounds... bad. Does anyone have any idea what instrument might be used to describe this sound in the negative prompt so that I can try to get rid of it?
One clear example I have is the background reverb-like sound between 8 and 15 seconds in this song. Like I said - hard to describe but the percussion (?) sound that plays with the fastest tempo during 8 - 15 seconds. https://suno.com/song/d54f4aeb-db51-4e3b-9239-89ab9a02fc69
symphonic metal, sinister, carnival, dark, high-pitched goblin-like voice with raspy undertones song. Listen and make your own with Suno.
I'm not sure if you're already doing this, but I end all my songs with these in the lyrics section: [Outro] [End]. I don't seem to have this problem more than maybe 1 in 100 generations. If I did have that problem but liked the song, I'd use the extend function beginning just after the last part that you like and in the extend lyric prompt just put [End]. Usually this will cause the extension to be <10 seconds and smoothly transition to silence. If you like the end already and just want to get rid of the extra chorus, you can also use the crop feature and just select the song from 0 seconds up until the end (before the extra chorus). Hope this helps!
So ive been working on a song for 2 days now and i cannot get it to say the words from the promt ive tried using reuse prompt and a few other way but it seems stuck on older iterations verses new versions anyone know what i can do to fix this I've made like 300 versions all are different musically but the words stay the same even though they are change and getting the same results
anyone else notice a new upload window now? that lets us trim the clip?
The upload audio top left ?
Wood house...not in the mood
Killz, may I see it?
Woodhouse splintered me for typing your name all caps
Change the words...in every available window, Say...reuse prompt or edit it...and under the lyrics section where you can insert an image, change it there as well as in the creation interface,
Then if you do use the change section or whatever it's called get your timestamp count and extend it from a defined point of your song ..like after a crescendo or post Chorus... Interlude or break ..it doesn't have to be labeled as such just match a calm spot 20 seconds give or take surrounding the words your altering...I did a Funk And Groove song ...and Suno got messed up somehow and my next project was littered with F * * * *bomb while the filter was at its most strict setting during the start of the lawsuit ordeal...idk has nothing to do with anything but funny..I saved it
Meaning your right it can get hung up or confused almost like it habitually will flow and say the wrong words even after revamped
anyone have a good guide for trying to mitigate the shimmer with other programs?
see what Goelem ?
so V4 has this shimmer effect on the drums snare how do we turn it off?
Thank you for the reply. You've certainly proved that it can be done, but the problem is that I really like this song the way it is except for the words I can't repair. Is there any way to keep all other aspects of the song while repairing the glare/stare/stair issue. I don't understand the UI completely. The closes I've come to keeping the same song and changing lyrics, is to edit the lyrics of the version 3 edition, and then remastering to version 4. Seem like their would be a way to alter the lyrics slightly, and generate the song just like it was but with only a word different. Its not clear how to do that in the UI. If made the UI, I'd just make a button under the lyrics editor, so that after you're done you can regenerate without creating an entirely different sounding song.
Taiwanese mandarin, Chinese traditional, do, fast, drawl male singer, High pitched singing, Shamisen, Shaoxing Opera song. Listen and make your own with Suno.
之前V3.5是沒有的,爾且響度也不夠響,氣勢沒表現出來。這是V3.5的原曲https://suno.com/song/acaac58b-e944-4b54-a551-097409a56dac
Taiwanese mandarin, Chinese traditional, do, fast, drawl male singer, High pitched singing, Shamisen, Shaoxing Opera song. Listen and make your own with Suno.
Does someone know, why suno generating a text something like "this is a fake app...". Using Chrome on android, and Firefox on PC gives same result .-.
P.s. Extend and generate from scratch works fine, but Cover and Persona generates spam lyrics about that i am using (somehow) a fake app and need to proceed "suno.com/android... etc"
Thats disgusting
If anyone else haven't figured this out yet, I think the monthly Black Friday deal only works if you haven't had an subscription before..
They really should add that to the announcement.
Due to people abusing Suno's internal API for their own gain, Suno have been detecting for any accounts that may be related to these third-party services (e.g. non-affiliated parties, non-official apps, IP address, etc.)
Looks like your account got flagged. Nothing to be alarmed about as there have been several false flags (mainly due to VPN).
Please fill out the following form and the Suno team will investigate your account to remove that false positive: https://sunomusic.typeform.com/to/zGsJSzsf
Be aware it may take some time to resolve as the investigation team is small and only work during Eastern USA business hours
Turn data collection into an experience with Typeform. Create beautiful online forms, surveys, quizzes, and so much more. Try it for FREE.
This is so strange, never used any king of API
But thanks, should try to fill the form
I hope it helps
Like I said, could be having an IP address that an abusing service is using, esp. if you use VPN, a common issue for these false flags
Yep, sometime forgot to turn off vpn in the past using v3.5, maybe thats an issue
is the v3 down or not working for anyone else. i just did the same prompts in 3.5 and 4.0 but it wont generate in 3.0 for me? i still think i get the best results and sounds from version 3.0, i wish it would just go longer than 2 minutes or 1 minute for the extensions. Also it almost seems rare to get a 3 or 4 minute track these days on any version, has anyone else noticed this or have a solution to get longer tracks?
after typing that 3 of the 4 tracks I made were over 3 mins lol
In what context are you trying to use 3.0? (E.g. cover, extend)
I can see it in my model drop down list in Create
i was trying to make new tracks from scratch.Keeps saying system error occurred at the bottom. Made the songs in the other versions but wont work in 3.0.
Hmm make sure that the lyrics are < 1250 characters for v2 and v3.0 iirc, that may be the issue (esp. if they did not fix the visual bug for the max lyric char for it)
Please let me know if that was the cause of the issue
ok i think it was the character max, its 2500 characters of instructions in brackets [ ]. i forgot to shorten it to 1250 and didnt see the reminder. Thanks.
Does Suno learn / improve from thumbs up / thumbs down?
i have a question
Not directly.
For beta features and beta models, the developers use the feedback from those to make adjustments.
Outside of that they have no direct effect.
i think it lets me use the v4 model kind of when i write it in the style of music, it provides me better sounding music structures and overall better sounding music than v3.5
is it normal? or Is it intentional?
you can test it by :
experimental-{your choice of music style}-v4
in music styles 🙂
and tell me is it just me or is this a thing.
You can't make one model use another. They are separate implementations.
The prompt will likely be doing something but not what you expect here.
I usually apply a compressor and EQ with audacity, do you guy have a good AI mastering tool to recommend ?
Hey all, so I was wondering about something.
I made something and I used the "Barbershop Quartet" style. Now I want to create a song, but I want it to be an all female choir. I was wondering if anyone had any suggestions for styles, other than female choir, since I made numerous generations and that didn't seem to give me what I really wanted.
Try something like
a cappella, Female, Female vocals
A lot of people recommend Bandlab.com for AI mastering
Got it, I'll try that!
(maybe offtop) TBH I don't understand v4 remaster
- It doesn't follow lyrics. It cuts lyrics not even in half, it's 20% of original lyrics
- It produces damaged song structure because of 1.
- Sometimes it's 2 second song (yes 0:02 !!!), sometimes 1:14 out of 2:04 original
Yeah, music and vocal quality is better, but in general currently for me 4 tracks generated all 4 are not acceptable
-
Does the lyrics on the song reflect the original lyrics being sung? If not, please make sure to do that as it follows the displayed lyrics closely when remastering (cannot recall if you need to fetch rid of the tags and account for hallucinations)
-
???
-
Hmm I will assume it is because of point number 1
My remasters had lots of noise...
Likely due to shimmering which devs are still trying to get rid of. We should get a major update about that in a week or two.
For now, you can see #1309224710692929548 message
Yes, song matches lyrics
Ah, strange then, not sure how else I cam help you, sorry
You might as well just Reuse Prompt lyrics in this case (when shimmering issue is resolved) as Remastering does cost credits iirc
Yes it is that shimmery/wavy reverb you hear
Hm...I see
You can hear it in the original too due to the frequency illusion but, whatever process v4 uses to clarify the vocals, made the issue very apparent.
In 3.5 that's more like a microwave sounds, I got used to it 😄
oh yea in the beginning it's very apparent too, wasn't paying attention
Now you are cursed with that knowledge. Welcome the frequency illusion
(Hmm, sounds like an eldritch curse or something lol. Who said knowledge doesn't hurt?)
The frequency illusion (also known as the Baader–Meinhof phenomenon) is a cognitive bias in which a person notices a specific concept, word, or product more frequently after recently becoming aware of it.
The name "Baader–Meinhof phenomenon" was coined in 1994 by Terry Mullen in a letter to the St. Paul Pioneer Press. The letter describes how, a...
nah, I was aware of microwave noise from other tracks, just wasn't paying much attention on this one, idk
Also v4 inherited bugs introduced in some iteration of 3.5 (I've posted complains here)
Generation from scratch.
There is lyrics left, it's 2:09, it's not silence. Why end of stream?
Paired generation even worse 1:36 (forgot about 3rd verse, stopped just after chorus)
3rd generation 1:44 (forgot about last chorus, hard drop on no-silence)
4th generation 1:41 (ok lyrics kinda fit, but where is my outro?)
It's hard stopped right at first kinda silence, so I need generate outro with extend
It ignored explicit outro
[Outro]
[Fade out]
[Silence]
[End]
Yeah, v4 is like v3.5 where it does not want to end.
Some solutions that worked a few times for me but I do not know how reliable these are: #📚┃suno-school message
(If you are planning to remaster, might as well use v3.0 to [End] it)
If it doesn't want to end - that's not big problem. Extend + [End], right
But in my case it ends way before lyrics prompt ended. It just trim at some position as udio do
Ah, sorry, it is rather late and my brain is not working properly.
Sometimes the AI ends when it thinks it reached a good end point (probably because something in the lyrics?)
You should be able to make it longer by having things like an [interlude], [___ instrumental], etc. where it cuts off
If you are extending, make sure that the extend does not happen during the silence portion and a bit away from the last instrumental. Sometimes you may need to [break] and bring an interlude/short instrumental to continue on.
Guys, learn already how to use some basic DAW instruments or even some simpliest online services
Cmon now, you cant do the simpliest fade out by yourself or just cut the song where it needs to?
Never had this type of problem
WHAT THE HECK !!!!! EXTEND DOES NOT WORK WELL ANYMORE IN V4 ! It glues the parts at the wrong position. 2-3 seconds. At least for me now.
This never happened before. Is this another known bug from V4 ?
I find my extends to transition flawless. Could be the point at which you chose. I try to get close the end of the current measure. @soft gyro
Can’t see any option to extend in the iOS app. Trying to extend a V4 track.
I like to use [Fade Out].
Is there any way to sort library by both text string AND liked simultaneously?
To hide unliked matches for the text search?
Has the "stuttering" sound artifact been fixed for V4 yet? Some of my remasters have this choppy "helicopter sound" on them that I'd rather avoid in my final V4 remasters. Here's an example song, pay attention to the intro in particular:
https://suno.com/song/e96ff84a-4d36-45cd-b700-e34dd140b213
greek folk metal, technical death metal, melodic harp, high-energy crossover thrash, satanic hardcore punk, progressive song. Listen and make your own with Suno.
Whom to text here for account related issues ?
I've been getting some odd background squeaks and stuff on a few occasions lately.
Yes, we refer that sound as shimmer (because of the wavy reverb). Please see #1309224710692929548 message
Re: using replace for prompting.
Good morning. I haven't tried yet. So asking if anyone has used Replace for prompting? Will it work that way, or just recognize words. Can I put brackets prompts in the replace section and suno interpret it as a prompt?
I think, if I recall, only words appear in the lyrics review. So likely it only reads any text as singable lyrics. 🫤 I'll find out soon.
Discord isn’t an official support channel, it’s just users helping each other out.
billing@suno.com for account-related matters . 🙏
Billing.suno is not working. Support one reply not provided yet. Thank you for the information
When can we change this channel name to "technique" discussion? 😅
Techniqual Disqussion
or tracking how active DASVON is?
Pretty sure this is where we talk about the techniques we use when we play Discus
from my point of view, all technical questions that should be on SUNO or something should go here
from stupid questions to more elaborate ones questions
Does anyone notice that after you created the song, suno put an extra space after the "," in the style prompt?
So when you resuse the prompt, you have to remove the extra space if you want to fix the prompt. I hope they fix it 🙂
Is there any way to v4 master a v2 track?
Yeah, you should be able to remaster them the same as any other clip 💪 Just tested
Hm oh. I asked because the option doesn't show up on my track where it usually is... Weird.
Hrm... there might be some other reason preventing the song from being remastered, but nothing is jumping to mind atm
Is this a song from your own library?
Yes it is. And it seems to be visible on any of the parts (for example 2nd or 3rd section of the song) but on the "full song" it's gone
If you'd indulge me in a curiosity -- could you open the Full Song in crop, then "crop out" the entirety of the song, essentially making a duplicate
I'm curious if the same restrictions apply on that resulting track
I actually tried - cropping won't work on it. Might be bugged in the system then. Oh well, I'll just "remix" to the same genre, seems to actually work well 🙂
seems to keep every quality of the original but upgrading the quality like I was expecting from v4 mastering, so I guess that's a great workaround
Does "get whole song" cost any credits?
Or image generation?
Just got a message from a user that v4 isn't available in Israel. Is this true?
I need help. My song generate in V4 is cut off at the eng even though it doesnt exceed 4 minute duration
this happened quite frequently
Send in the whole song settings
here?
probably better to open my link https://suno.com/song/6969b8bc-aa84-4426-b397-05a6628b136f
dance-pop,k-pop, female voice song. Listen and make your own with Suno.
Get whole song does not cost credits. Just the Extension. 🙂
Tried to extend or get whole song?
my temporary solution is extend the song
i did. But i still dont get it why i have to extend everytime i generate a song. its just waste credit
Write me a DM
maybe people know better in here...is black friday 40% off for new customers only (as in new, not non-subscribed), is it for US only, yearly only or everyone? Cause I'm not a new customer, not yearly, not subbed and not US but I dunno if till be a surprise charge or not
This is a bug dating earlier than May and they know and still have not done anything in 9 months
Song time is bugged with covers and personas right now. Fresh generations will go to 4 minutes but covers and personas end early
Usually at 3:14 in my experience
I need some help with v4. Things have changed in the app and I can’t figure it out.
I just created an awesome song and I don’t know how to change the lyrics slightly here and there with keeping everything the same (and maybe add another verse) now there’s just Remaster with no way to edit the same lyrics?
I am pro plan now ..so i can post music in itune now with or without email to suno?
I don't know whether they are working right now at the V4 model, but the audio quality right now is so bad as never before !
The audio deteriorated with the time, even without extensions now... such a bad sound! Sometime even starts bad!
Comparing to the Udio sound this is like...nothing.
This is just a credits waister. The will need a remaster of the remaster of the remaster of the remaster.
Besides that, I tried to extend a song, I made 10 gens and it produces the same melody everytime, which was not the case before !
It is the first time I use V4 to generate songs, but V4 is a joke !!! They should return all the credits spent for 4.0 gens.
Everything is buggy & miserable. Not to talk about that stupid text prompt.
Do you have the same problems or it is only for me ?
it is impossible to finish the song...
v4 is in beta so fixing issues with it is likely their main focus at the moment.
I know for sure they are trying to fix the "shimmer" noise that affects many V4 songs. I don't know when the fix will be ready though. It ruins many of my songs.
so they do changes while the system is running ? dont they take a separate server for this ?
I had earlier with 3.5 some "bad days" where only I had problems... or some.. but not all.. so maybe it could be like this again ?
Credits still don't carry over right?
I don't know their infrastructure details but I expect they have a development environment and there'll be updates to v4 during beta up until it's released. Similarly, cover and persona are in beta so can change.
The main, non-beta models (used for normal generations) don't change.
the credits expired without being able tto use them
Correct, they do not
@worn flicker Not that either of you would necessarily care all that much, but just figured I'd give a quick update to happily report that I've now managed to finish "take 2" of that song and get it a lot closer to how I originally wanted. I still had to throw audio into Audacity and make a few additional edits, but all the advice from you two really helped with this. Thank you so much again!
I think my original desire still stands: I hope Suno can one day introduce some tags which help you more directly influence how each section in a song can sound, but at least for this particular song, the job is done! 👍
Do you anyone know, once shimmering issue get fixed, will that fix all playback of existing songs or it will apply only to new one? I mean, does it make sense to spend credits now if the result will be most probably broken?
It's still being worked on so I don't think anyone knows for sure at this stage. Maybe remastering after it's fixed will help since that regenerates the song.
there no way to pay with paypal? for some reason my card fails whenever i try to pay in us $ (am UK btw)
I’m curious if there’s a list of meta-tags that can help guide the creation of a song. If so, I’d love to learn more about them.
There are many lists. Some work some don't. There are a couple wiki pages as well. They are just a little dated. The good thing is they will give you direction:
https://sunoaiwiki.com/resources/2024-05-13-list-of-metatags/
Oldest
https://sunometatagcreator.com/
Thank you!
That will probably have to go to support@suno.com
hi
recently i make two remasters , they eat my credits
but it dosent show up on create or library tab
https://help.suno.com/en/articles/2421185
PayPal isn't supported atm 🙏
When you create an account, you automatically start with a Basic (free) plan. Our Pro and Premier plans will give you benefits such as more credits, c...
Strange, they should if it was successful. Please make sure:
- You are on Page 1 of your pagination of the library
- If you still do not see it, refresh the page
- If you searched for something recently, the search result may be saved (though should have been erased when refreshed). To do this manually, type anything in the search box then delete it.
Note if you are trying to remaster a v2 or lower version, it may not work/be supported
Every song I make with V4 now has the strange "Shimmer" "Bells" effect. Has the issue gotten worse or is there something I can do to lessen it? It sounds like a glass is shaking on a table in the background. I would have assumed it was my headphones but everything else sounds fine.
The team is working on it as the message not long ago mentioned. The team should give us an update of the progress in a week.
For now, please refer to #1309224710692929548 message
I am in there @mental lava lets chat. Its an artifact
I got a question about suno yearly plan. If I have a lot of credit left for the monthly plan and I upgrade to yearly plan. Do you think those credit carry over or they just reset them?
I see a promo and I am tempting to go for the yearly plan 🙂
Hey guys, question about plans. Got old 1000 extra credits as “buffer” :):):) and Pro. Always using exactly 2,5k monthly. If I take Premier (on multiple accounts, to have 100k or more credits), do I need to use all credits always until end of month cycle? For example:
Start of month: given 10k credits, got 10k
Used during month: 7k credits,
End of month: Got 3k credits left
Start of next month: Do I have 13k credits, or 10k (and 3k is wasted)?
/chirp
Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com
Credits are just a freebie for paid members. So each month they renew without roll over.
You CAN purchase credits which are yours to use, so they stay ongoing until used up. No time restraint.
Thanks. Sorry for my english it’s ununderstable :):) so 13k credits or 10k? :):):):)
10k... each month at renewal. The remaining 3,ooo are forfeited.
Thanks. Yeah, but 4k credits for 30 USD (as I remember) is much worser than buy Premier for 30 USD and get 10k and cancel before month ends, and after you use it, do same.
💡 Right! I use about 9000 a month. I average 1-2 hours a day testing and creating outputs.
I used to work on Suno 10 hours daily for first 4 months on v2 🙂 yeah, much used 🙂
Where is that "Lifetime" subscription for $1500? Hahaha.
Nah. Checked Black Friday. That’s just monthly subscription in price of yearly divided by 12. No bonus for yearly subscription, am I right?
No additional discounts, correct
Kinda like the coupon saying it does not compound on other sales/discount type of thing
Sale: $216/12 = $18monthly ( per a year of Premium). (1 year limit)
Presently premium is $30/month or $288yr/($24mnth)
Hmm. You’re right.
iZotope
That's like you get to eat 1 free dinner a month! Mmmm
I read somewhere that Personas don't actually work with V4 but can't find it again. anyone know if that's true?
If you use one on 4.0 and it has a similar voicing, that seems the test...I thought I've done a few, but honestly I'd have to check to see if it went through 3.5 or 4. By the end I've remaster many through 4.0 and still most carry some characteristics throughout.
I've forgotten, are persona's not available to free users?
Great question and this is where I'm at. Holding off with credits until I know shimmer is fixed. Pointless generating until then for me.
Hello, After generating numerous tracks with version V4, I’d like to share my feedback. The sound quality is undeniably superior, more stable, and for less complex tracks, it often approaches WAV-level fidelity. However, when it comes to pitch accuracy and emotional expression, I’m not convinced there’s been any improvement—in fact, it feels like a step backward.
Version 3.5 seems significantly better in terms of creativity and conveying emotion in vocals. With V4, the emotional depth appears to be lost, and I’ve even encountered off-key notes or atonal sections. Furthermore, when converting a V3.5 track to V4 via the remaster function, in 80% of cases, the emotional nuance disappears, and sometimes false notes are introduced.
Am I the only one who has noticed this?
There is a matter of a voice morphing through a remaster, but that's also happened in 3.5 before 4 was released.
It's been mentioned as least by me. It's disheartening. Just s monotone style melody where sometimes the chorus is the same as the verses with little range. No emotions, you're right.
And correct. Both melody AND instrumentation are off pitch, off key, or sadly stay the same pitch or chord when clearly it should be a different arrangement to what melody is being sung.
Right as 4.0 was released, just got about 4 songs that would make me cry, excellent arrangements, nice vocals..... then it seemed to fade away.
How can we explain this regression? How did Suno not notice it? Without meaning to offend, it seems like we are moving towards producing soulless audio. What made Suno truly exceptional was its remarkable ability to mimic emotional expression. After all, isn't that what music is all about—emotion?
anybody know how can i extend a instrumental song for real?
0:02 or 0:03 seconds is not a good extend from the original 3:38 song 😦
**bug reported
try removing some seconds
start from 3:31 by exemple
try more
still few seconds
This happens to me quite often. I think the model struggles with songs longer than 4 minutes—it seems to run out of ideas! ^^
not sure, over my mind, if i pay "more credits" , the time "will be extended"
V4 has become a bit lazy!
how can i push this topic to the team? o.O ?
I have a Pro account with 10K generations per month, and I’m experiencing the same issues. Not sure if adding credits would improve the length of the generations.
i have a pro account 2, but i never extended songs before
V4 struggles to extend songs beyond 3 or 4 minutes. Personally, I create in version 3.5 and then remaster in V4 to retain the groove and emotion. I’ve stopped generating vocals because they no longer meet my expectations for emotion, placement, and harmony. However, the instrumentals are still decent—about one in ten turns out really well.
Good evening and happy generating to everyone!
there is a problem if a try to "remaster" upload audio?
this is what it fails
they eat the credits and i dont see de remaster
thank you for your help, happy holydays 😄
Well, it's disheartening to me since I feel emotions, but I assumed there are parameter or gates that are set to try to limit hallucinations, and if suno can be emotional, it's in the hallucinations. This might keep it from going too far out of range in singing as well.
The odd arrangements, mixed in with some excellent pieces, and odd notes just seems a tweak thing as well. But yea, ive got a playlist just for "repairs and edits" until most revisions seem optimal. So between Suno taking vycadin and other challenges, I figured it's a known challenge, but other matters are prioritized.
I'll post a piece of my monotone song if I can remember which, just wrote this week. Hey, and I'll say, to Suno credit, the song could be monotone I think moreso because it's trying to emulate the indie, 90s, Indigo Girls sound which is the character of the persona used. So that may be all it is for this song.
I should test lyrics with a couple more persona's and just random styles. But other songs just seem flatlined from 3.5, so I know what you mean.
I want to get that deal but I still have 7k credit left. I think it's gonna reset once I upgrade to 1 year plan.
Is there a way to get SUNO to reduce the signal during rises & Falls if they are going to clip?
(I am teaching myself to upload my work from a DAW so as I am not a pro the Mixes vary)
remasters sound clipped
I am reducing the signal on all my tracks prior to upload. I was just wondering if SUNO was capable of that?
Is it just me or is in V4 for Rock and Metal Style a metallic strange sound in the background. Switched to v3.5 and there I dont have that Problem.