#🪲┃bugs-web
1 messages · Page 8 of 1
when u have a solid style combination you can expect as initial result usually 2minutes
49 seconds would ring my "personal alarm" ^^
This sounds more like a #📚┃suno-school chat guys.. not bugs 🪲
i see it the same now 3daizy
isnt it a bug since i cant get a result and constant coin refund on continuation
wasnt clear until he told me about the 49seconds initial result
after 40+ tries
no thats probable in AI, especially when u narrow down the possibilities of it so dramatically
a bug would be if you dont get the coins back 😉
yes but this is the bugs channel.. if it is not a bug but technical we have a different channel #📚┃suno-school
We don't want to overcrowd this for the Devs
hey anyone else having issues with v3 when it comes to getting the full song after u extend it ?
hello, why i can't download the song, when i download it it just the cdn, not the audio
i just subscribe today but i can't download the songs
do u have a popup blocker ?
hmm where can i find it? i dont know
do not sing lyrics😡 🤬 🤬
if u have a pop blocker it would be something you set up. its a browser extension
if u do then disable it and refresh the page then try again
worked for me
It seems about the same. The two noisy clips I shared above were both on the Instrumental setting.
does the first part that you're trying to continue from end in silence (or a fadeout)? if so, may need to use the continue from time field
no, just makes sounds and sounds good buit then ends at like 5 sec
mind linking me the first part? can DM if better
can you DM me a link to the song part that won't give you the full song?
sure
This is the worst it will ever be!! 😁 Don’t fight against the nature of the medium, work with it! 🥰✨💖
Indeed! I'm hyped, thrilled, and all-in. Just trying to do my part to contribute to V3's improvement.
🥳🌈
I just had a 3:58 long first part of a song generated (v3 alpha). Whilst it was generating I could listen to it all. After it finished, it was cut to 2 minutes in length. I managed to retrieve the rest from my browser cache. Seems a bit wasteful to lose half of what was generated.
Have just reported it via the site. Last few seconds were glitchy but otherwise it was a nice song.
Are you sure it isn't two distinct clips being mashed together? That is a bug
synthesizer virtuoso, complex melodic, electronica, downtempo, blues, trip hop, sophisticated, Japanese synthwave song. Listen and make your own with Suno.
Single generation that is obviously the two generations meshed together, confirmed by the devs
Hmmn, I think I am hearing the dataset bleed into my generation and I can tell the exact artists it came from 👀
90% of the songs i generated today have come out without lyrics. Tried both v2 and v3, tried changing genre, and changing lyrics. I can't get anything to work. Lost several hundreds of credits.
ok what's your prompt...
just unlucky? cause others clearly can generate songs with lyrics...
[Verse]
bor-isu plays a game of F F A
always at N H F F A
don't have the skill, dont have the A P M
i never learn, i never get good
cant stop playing, starcraft is my drug
[Chorus]
A hundred carriers invading blue,
storming red, yellow don't know what to do
battlecruisers and carriers in the sky,
dark archons feedbacking, all terrans die
lyrics generated by a bot a friend is using
I did get it to work finally
but honestly i must have to generate 40 times before it popped one with words
if i click thumbs up or down on list, page is scrolling to the end of page.
My prompt is being flagged for moderation but I made a song hours ago with similar lyrics. I have the word "butt" and "booty" in the song, and the style is "Alternative, 80s Synth Rock, Dark, Indie, Almost monotone male vocals.". Any idea what I'm doing wrong? It doesn't say what is not allowed, just that I'm flagged...
Has the moderation changed in the last couple hours?
I don't really get it, but I managed to put it through. I had "Dat Booty!!!!" as a line, which it didn't like. But it was fine if I deleted three of the exclamation points and just had "Dat booty!"... the more you know
Possibly. It does suddenly jump at one point but not at the 2:00 mark and seems to be part of the same track.
Hello, reporting a bug. If you delete a track in your library, you won't be able to navigate to the next page unless you refresh. I have ran into this bug before in the library before, but I am not sure if I was moving a generation to the trash that triggered it.
But this time I did observe that it occurred after that condition.
Also not sure if this is a bug or a feature, but you can have duplicate tracks in a playlist. So you could have a playlist with just the same track repeated. Maybe there is a valid use for this, but it seems there shouldn't be duplicate tracks in playlists?
Kalimba is an instrument. Just because some dumb musician decided to use it as its name doesn't make it a copyright infrigement
My Part 5 is Stuck and does Not finish i can get no more Full Song now, what can i do?
Deleted the Stuck Part, works now
use similar instruments, mbira, sansa, or try spelling kalimba wrong (depending how advanced/literal their detection systems are)
I know. Should still be in some whitelist of allowed words though
oh I agree 🙂 its VERY literal and a bit obnoxious, should be able to indicate "Kalimba (instrument)" or something. Not a hard build either
Well, it's not pops and crackles. It sounds more like a vacuum cleaner, like in this track:
https://app.suno.ai/song/1837205e-8984-45fa-af54-cbcc9b7392cf
psychedelic soul, 1972, shoegaze song. Listen and make your own with Suno.
https://app.suno.ai/song/329325ea-f4db-4d89-88df-3939934fa49a
This was the fest of a sample I generated on suno when it was discord exclusive in like 2023, shame they haven't fixed the old v1 clips so you will get a error attempting to load it (yes the link is link only)
I might have made too many parts to a song and now no ai image is shown, (even when creating new songs). Is this a common bug? It's only been like this for the last hour. I have not made a hard refresh or anything much yet.
No, it's a server issue across the board right now
Thanks!
logged into my account on browser and when i went to generate today it had me accept a blank popup on the page? i couldn't read any text but it was stopping me from writing lyrics until then
Created a 9 part song last night, moved part 1 to trash because part 2 had a better intro. At some point part 1 has been delinked (not being highlighted with the rest of the parts) and when I try to get the whole song, it gets stuck spinning and after about 5 minutes shows the error below.
I've tried moving all the parts between the trash and library, I've tried creating the whole song from a previous part, I've just gone in and perma deleted part 1 as I saw in another channel someone said they deleted a part to fix the error, but still failing.
I believe the first part of the song was delinked because I moved it into the trash however I moved other parts from different songs into the trash and none of those have delinked.
also just to check if it was definitely a problem with part 1, i made a diff song with several parts, combined into a full song successfully, then went ahead and perma deleted part 1 of that song then tried to get the whole song again, which ended in this same error
i guess its mostly a problem because it cant get the whole song starting from part 2, it needs part 1 in there 😭
https://app.suno.ai/song/914a5010-c108-4174-9037-7f7535938fa7 lead to a weird bug with this continuation (0:47) https://app.suno.ai/song/9880625b-7ef4-4824-91d9-56aa114baefb
I am going to assume Suno is have some issue with verse 2 and 3 today? Its generating like 3 to 4 secs for song (instrumentals) but saying its 1- 2 minutes long?
I cannot change my email associated with third-party login from time I first registered with this first account. For example, I changed email in Discord and this is not reflected in Suno. Try to add another account - Google - this is not respected and still the same old email appear in Suno and Stripe and everywhere. Cannot even really disconnect from any third-parties. Even the connection is disabled in Google or Discord, discconnected in Suno, when trying to create new account in Suno - these disconnected and permission-removed accounts are associated with the same old Suno account no matter what I do. So I cannot create new account, cannot change email address (for Stripe) in any way and disconnection is not working in the backend (still remember wrong details). Why we can't use just simple email and password to login like normally? And what purpose have the password in 'Security'? Because of these problems I never use third-party logins anywhere. Only in Suno I am forced to use it and this is not even working as it should.
2 full songs but in chorus 1 it cuts the last word every time
songs always never conclude and get cut off at the end (SUNO)
You have to put [ fade out ] or [end song]. your right it wont just end the song
Im using version 3 and like version 2 i have found you can only use so many characters per line and its best to stick with 8 lines only to get all the lyrics.
I'm just having trouble getting lyrics. Even though I provided them, I've gotten dozens of songs with no lyrics in the last few hours.
Did you guys know that it keeps repeating lyrics that it has already sung?
I frequently get missed lines of lyrics, and/or unexpected repetition.
Im using v3
Why did I subscribe, but it still shows that I haven't subscribed on Discord?
Try this: #💬┃general-chat message
@sick smelt.ai Why did I subscribe, but it still shows that I haven't subscribed on Discord?
I registered and paid for my subscription and even before exhausting 1/3 of my credits
I was disconnected and when I reconnected my credit became zero and I was asked to re-subscribe
What should I do to fix it?
Have you made sure that you are logged in with the correct account?
Otherwise you will need to contact billing@suno.ai regarding this matter.
oh my bad thought this was the general chat
When I first started using Suno a few months ago it produced very melodic songs on the first or second try. But now it takes 20 or more tried to get anything decent that does not sound like all the previous songs.
You're not alone in experiencing this; many Suno users occasionally find variations in their music generation results. Despite no recent major changes to the model, the nature of AI means that it can produce varying outcomes even with similar inputs. This is a typical aspect of working with AI-generated content. To enhance your experience, keep experimenting with different prompts and styles. The diversity of inputs can often lead to more varied and satisfying outcomes.
Just tried to deliberately continue a v3 part with v2 to see if it would add any interesting spice. Instrumental only. 3 previous parts. But I changed the style prompt to something else and it seemed to only consider that and completely ignored everything else, acting as if it was part 1.
As far as I know Freestyle is a genre, even if an an artist has that name it would be weird to ban the genre when the genre is more popular than the artist, no?
try this it may work too: An end tag in the lyrics may work best alone as its own clip. Clear the Style Prompt, or add ‘End’ to the style description.
[End]
[Fade Out]
[Fade to End]
When ever I use prompt "rock song" 90% of the time it produces country genre songs.
So I try to make German songs, but at the beginning everything is always completely crazy. 2 weeks ago everything was really good and now there is no more music 😦
When editing song titles, I am sometimes getting a spinner icon over the Ok/Confirm button, so I can't change the song name. It just spins forever
I'm having some issues with v3 streaming today. About 50% of the time, it seems to pull one or two chunks, then stop streaming the clip, even if that's only 12 seconds or so. It's possible it's actually getting cut off that soon, but it feels much more like the generation is continuing server-side and just not streaming to my browser anymore. Reloading doesn't get it to complete, either. It could be my VPN -- I'm about to do some testing -- but since it's intermittent, it's hard to test for sure.
I have issues with playback similar to that without a VPN on. Playback is choppy frequently
Still usable but it will stop for a few seconds. But my internet speed is 1 gigabit per second down
If you're streaming from v3, that's to be expected; generating the next chunk could take longer than playing the previous one, and then you'll have to wait. That's fine; but in my case, it's permanently stopping, never generating/streaming anything more.
Rock is to generic. You need "hard rock", "surf rock", "rockabilly", "rock opera" or something more specific to narrow in on the style you are looking for. Also add some emotional tag like powerful, heartfelt, angst, or something like that.
If you look at the Explore tab, people often use "genre()s, instrument(s), emotion(s)" to describe their songs.
SUNO web starts to generate trash, it ignores all written genres or tempo or whatever you write to tht "Genre window" and 50% chance to ignore lyrics, it just generate a random unclear s**t with Slavic accent
It would really help us in tracking this issue down if you could please post some examples? Thanks.
https://app.suno.ai/song/e8d3bc2a-c814-4954-9a0d-3efef3130039 Does this look at least a little bit like rock? Most of my generations are thrown into the trash.
This is an example of the absurdity of genres, but I apparently deleted those generations where there was a random set of words that did not correspond to my lyrics.
I would be grateful for an answer or at least a hint on how to avoid generation loss in the future
Alternative metal, post grunge, nu metal, nu-metal, male and female vocal song. Listen and make your own with Suno.
Every time I continue the song it only adds 3 seconds
SUNO web is totally broken! It cant generate any rock that harder than POP even if you writing something like "Electro guitar, hardcore rock, electro DISTORTION, epic rock, alternative rock" or any kind of metal genres..... It generates some pop s**t no matter i do.
I just drop all my generations to trash, thx.
ive found if you want harder metal songs. start the track with v2 instrumental and continue it in v3, the generations and a bit more poppy in v3 i do agree
Come on, a month ago everything was fine with the rock genre, but now folk music is generated instead of rock? or some Latin pop crap, this is not normal, I need a working generator but definitely not a crutch.
You go to the store to buy salted potato chips and they give you raw potatoes and say, “Well, you can make chips out of them.”
I hope this will be fixed quickly because it is obvious that something went wrong, the neural network is broken.
Btw, this problem appears in v2, just check the link I attached above
Groovy fast crossover thrash metal hardcore, massive guitars song. Listen and make your own with Suno.
I found this hack useful too, not necessarily for heavier songs, but I feel like v2 has much more variativeness. (I think it's also a natural effect as any AI model evolves for it to need more specifiers to generate variative output.)
Thanks for fixing this stupid bug, I thought my ears were going to bleed from all this pop crap.
Rock music is being generated again, however, I lost a lot of generations thinking that it was just some kind of strange randomness
I seem to do fairly well...
https://app.suno.ai/song/76fde1a4-5561-4366-9ea7-3db1afda8345
melodic instrumental, glam metal, death metal, brutal zombie metal, blues rock, darkwave, dream pop song. Listen and make your own with Suno.
"sido" is an artist name so it can't be used, it also means "have been" in spanish. I would like to use this word.
is this giving a problem in the lyrics?
I tried using it in basic mode, didn't try in custom
That's not what I meant. Is that word in the "style of music", or in the field of your lyrics?
in Song Description
Welcome! For help, check out Suno Help
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
Is anybody else having a terrible time getting lyrics in their songs on V3? Out of the 12 songs I've created this morning, only 2 have come out with lyrics. The lyrics are provided each time. I've tried changing the prompts around, but I'm just not having any luck.
I am getting this issue, I am trying to do industrial metal and getting most of the time Latin pop..
I just created a song and got no lyrics too
I understand the frustration with generating songs without the desired lyrics in V3. There are a few potential reasons for this. Suno's V3 is in a developmental stage and can sometimes struggle with certain prompts and song structures. The length and complexity of the input text, as well as specific metatags, can significantly influence the outcome. To improve lyric generation, you might try providing more detailed inputs in the "style of music" section, experimenting with different styles, and ensuring your lyrics are formatted in a way that's easily interpretable by the AI. Remember, V3 is continuously being updated, so giving detailed feedback on these issues can help the Suno team make improvements. @pulsar stirrup
I believe there is an issue with V2, everything is getting a "pop" makeover, I tried Industrial Metal and Dark Jazz, both are more leaning towards pop in their way, V3 looks like more stable but it always cuts at the wrong time (in metal, jazz is more ok)
The model has not been changed.. So it is more likely your prompts, lyrics, and unlucky
It worked fine few days ago, the style is simple "Industrial Metal", the lyrics are also pretty regular, 8/10 industrial metal songs generations are just Pop songs, I don't think it's luck
I gave it 4 verses of lyrics and it didn't include a single word. My style was "country", which generally contains lyrics. Saying that this is user error on my part is weak sauce.
Feel free to send me some examples and I will look at it after I had my dinner 😉
Sometimes you just have off days
I was almost finished with a song. for the last part, I wanted to generate the outro stanza... suddenly it stopped generating the same vocal, and changed to generating only female vocals or vocals siounding like a boy who just entered puberty. i even entered "manly male vocal" manly male voice as prompts,. regardless it will continue to generate female voicals,.I had thiws situation a few times before. vocals suddely changing, ansd once t was even within the same part which it generated, like started as male, next line went on with female. bt this time i have generatede about 7-8 times. Just losing credits but it doesn't seem to go back o the same vocal.
would that be maybe better if i continue another time? does thazt make a diffrence in the reslts, like when you generate depending on whatever?
It's more that the system is weird and unpredictable and we all really lack actual control in lots of ways, especially if you are from a music sorta background. Suno works backwards where you get the master first and have to try to finesse it to get to where you want
i generated with v2 many heavy metal, nu metal, death metal songs. and i even kept a bunch of them as they were really good. so i think there must be sth else with it generating more pop like results. i mostly used many tag words in the style description btw, almost to the character limit. like, aggressive, slow, guitars, etc etc. just put the mode you like and the intruments yo want heard prominently.
Thanks for the tips, I also generate many, over 100 songs in fact, never had this issue, especially not in this capacity. Thanks again I'll try adding the tag words you suggested
Yeah, i'll just go one step back and retry from there. it was just so good that part, but it seem i'll have to let go..;) I just hope it's not a new Google Gemini Case 😬
I find V3 is broken still. much worse then when it came out. rare intros, cut verses in the middle, skip chorus or double them randomly, almost no instrumental and usually ignore prompts and commands related to the music itself, also the generations inside the song (continues) are too random. really hard to make songs now. I'd say custom mode feels much like random mode in many ways. user control over the outcome is very limited compared to early V3 and V2/V1. it's too creative and doesn't follow instructions well. IMO V3 at the worst state since it came out.
I have been generating songs for a few days now, getting good results. The past 2-3 days I have the same issue - Poppy Female / Effeminate vocals in new generations - Even with the same remix lyrics and prompts (Hardcore/Metal). Many songs also now generate with mixed gibberish (Asian sounding) lyrics, either mixed with my English lyrics or only Gibberish throughout. Something with v2 has definitely changed 😦
No really... The model did not change!
Navigating new technologies like Suno can be both exhilarating and challenging. Your experiences, where results don't always align with expectations, are a common aspect of this journey with AI music generation. Different components like lyrics, metatags, prompts, and music styles can interact in unpredictable ways, leading to unexpected outcomes. While it's easy to attribute these to the AI's shortcomings, it's essential to remember that we're all in the process of adapting to and learning about this new technology.
When you encounter such inconsistencies, consider tweaking your approach. Experiment with different music styles, change up the lyrics, or adjust your prompts. Although this may consume more credits, view it as an educational investment. It's an opportunity to explore, learn, and shape your unique style within this pioneering domain.
Suno stands out in the realm of AI music generation, and the advancements it has made are nothing short of mind-blowing. There's still a lot of room for growth and development, but that's the exciting part. We're ahead of the curve, exploring uncharted territory in AI music. Embrace the challenges and the opportunities to learn and grow with this groundbreaking technology.
Maybe we should just be honest...
There was interference from the Van Allen Belt?
No...eh?
How about quantum decoherence is the issue?
Still no?
I suppose it could be Neutrino overload on the nameserver...yeah that's...not buying it?
How about ya'll just pick one from the list: https://pages.cs.wisc.edu/~ballard/bofh/excuses
There has been no changes in the backend that we are aware of at this time. So given that everything operates on randomness in some form or another I'd say you're just having some bad luck. But if that explanation isn't good enough for you there's a whole list of BOFH approved excuses you can use instead.
Yeah today is another Frustrating day on Suno. Feels like its getting Worse every day.
There is this idea that people have about AI systems that they learn in real time
It uh, isn't like that
It's a whole bunch of data, training on that data, then you drop the model and cross your fingers
There is the interface we interact with it sure yeah, but it really is just good and bad days since it's so unpredictable
Not necessarily a bug, but I would like to report general poor performance in the suno web app. I have a large playlist, hundreds of songs. IWhen I click on a song it takes about 5 seconds to start playing. Multiselect often doesn't work (shift + click).
You can add the same song to the same playlist over and over again and it will just add a copy of the same song instead of saying "song already in playlist" or something along those lines.
Oh sorry, my bad. I thought you were an employee. Nothing to see, here...move along. 😉
"a clip got stuck while generating" the past 5 minutes. anyone else?
same here
yes.. everybody has it... As it says: They are investigating it 🙂
I need a song that doesn't have a vocal, so even if I activate the instrumental domain, I get a song that keeps getting vocal. What should I do?
Try adding "instrumental" to the style prompt. So "something, something, instrumental" with a comma and a space after your existing prompt
This has happened twice now so it's a minor, minor bug. But when I ask Suno for a guitar solo it will generate the start of one, but cut off after an extremely short amount of time. 3 seconds or so. I don't have to spend credits on it because it's refunded, but it seems like that isn't how it's supposed to be. Here is a link to an example:
https://app.suno.ai/song/13407da6-4071-4b06-ab1a-87f18fdc84dd
electric guitar solo, best quality, studio recording, 80s rock, hair metal, rebellious, Detroit, Rock City song. Listen and make your own with Suno.
There is clearly more there but it stops for whatever reason.
I went and found the other song this happened to. Here are the two clips where it happened. The only common thing between these two songs is me asking it for a solo. It took two 3-4~ second clips, one right after the other, to finally generate more.
https://app.suno.ai/song/7c2ec143-b491-4591-8ad2-c1c071b64265
https://app.suno.ai/song/d6e92b30-c608-4181-aa1e-65c024760d46
70s british rock, british hair metal, rebellious, electric guitar song. Listen and make your own with Suno.
70s british rock, british hair metal, rebellious, electric guitar song. Listen and make your own with Suno.
it is not from me
I've almost 2000 credits disapear of my account ????
Accepted terms of service.
In the past using the "Continue from this song" option cost 5 credits. Brand new prompts cost 10. But now it costs 10 credits for either one. Was this a deliberate change or is it a bug?
You have to look at it differently. You get 2 for the price of one. It is designed like this. Just like with other AI services, you get more options.
So, the credits charged for a "Continuation" did change?
As far as I know it was always this way. At least it was already like this in October, and I see you joined in November.. Thus...
For some reason this song isn't allowing playlist additions.
https://app.suno.ai/song/9d9fa678-c105-4a72-8a8f-cf21ec5863f9/
Female vocals, forward in the mix, melodic, singer-songwriter, mid-tempo, intimate song. Listen and make your own with Suno.
I'm on Brave browser for Android.
v3 is producing short clips, when clicking continue it produced only 49 then 53 seconds
I have a couple of 20+ minute epics done
Renaming Full Song doesn’t work via phone?
I clicked Submit. Not sure if it takes a good or how it works.
I'm having the same problem. I have tried making the same song dozens of times and it refuses to finish the lyrics even though it's under the max time allowed. It ranges from 13 seconds to 41 seconds for the same continuation.
Try adding [solo], [instrumental], or more lyrics in @daring junco @oblique pewter
Where would I add those. In the song I'm working on the initial verse 1 and chorus came out fine. Then I started a continuation and it will either make it to the chorus and end with just one line left or it will stop without even finishing the next verse and completely skip the chorus? Thank you for the help!
Just add it under your last word from your lyrics
[verse]
your text
[intstrumental]
I just tried that and it's still doing the same thing, sorry for being such a pain.
Are you using V3? Can you share the part (url) that is giving you the headache?
Yeah. https://app.suno.ai/song/63af4f2d-fe2b-4baf-a292-b789e26a9037 The lyrics might seem mean, a friend and I are just having a dis track contest lol
This is the part that is stuck? And what is the part it continues from?
To test it, I need to do the same as you 😉
Yeah I can't get it to finish that verse and chorus, it'll end at a different part everytime I try. https://app.suno.ai/song/016f6bcd-5138-4c06-8018-84a0987cd1d1 this is what its continuing from.
Alternative, electric guitar, clear vocals song. Listen and make your own with Suno.
why isn't there "style of music" filled in your second part?
Is there a special reason?
I thought that when you continued it would just try to copy the initial style so I thought I didn't need to put in a style. I initially had the same style in there as the first but the same thing was happening
okay let me try
Thank you so much for the help!!
Too much dicking around lol
AI has a mind on his own on your chorus. He likes the first 2 lines the most and repeats it and then doesn't want to sing the last sentence 😂
Yeha, in the first part of the song that came out fine it sang the first part twice too but ended up getting to the second part of it
Sometimes it gets the words all jumbled and I cant even understand it 😂
Okay.. I understand the problem. It is confused because it doesn't sing the chorus right the first time.. and you let him continue.. so he wants to make again the same "mistake" if that makes sense? @oblique pewter
I get what you're saying, Ill try to do the first part again and see if that helps the second part
Just start after the first verse
0:19
Like edit the first part?
How can I get songs into the explore page for others to see?
Did anyone get a response from billing@suno.ai?
I upgraded from Pro to Premier plan 10 days ago. Billing details are fine: $30, paid. I received 2500 credits and still on the Pro plan. I immediately sent an email to billing@suno.ai. 10 days later, no response 😫
I am sorry to hear that 😦
please send a DM to @muted flame I think he can help you with that
Yes.. let me give you what I did with the first part: https://app.suno.ai/song/7d21abb4-54f0-4ea8-9301-7701a2a5cb84
Alternative, electric guitar, clear vocals song. Listen and make your own with Suno.
You can try to continue from that part if you wish
Thank you! How do you edit a song like that tho??
Okayy I see, I was confused because it still showed all the original lyrics in there with an extra chorus, is that how it works?
What I did,
1: I continue the first part after 0.19 seconds, and added in the chorus + instrumental,
2: When I was happy with the second part (the chorus), I made a whole song (the one I gave you) where you again can continue from
3: to demonstrate you and continued it and made a full song of the part two.
https://app.suno.ai/song/75bc8e11-6656-4137-aeb3-e8a13a26266b
Alternative, electric guitar, clear vocals song. Listen and make your own with Suno.
So this song above is part 1, 2 and 3
And you can just open that url and keep continuing for the next parts
Making part "fulls songs" in between give it more memory to remember melody and chorus.. Plus you can hear if you like where it is going
Okayy, thank you so much for the help, you're the best!!!
I achtually like the song.. Good luck with it 😉
Thank you!! We've each made 2 dis tracks for each other so far, it's really funny lol
yeah Suno has so many use cases.. 🙂
Yeah, it has some kinks I have to learn how to get through, but it is really good and I've been having a lot of fun with it
Thank you!!
Getting errors at the moment with V3 gens, images are missing, appearing later as smaller, message say songs are aborted yet are generated and appear with a delay.
Nothing is rendering.
yeah seems servers are spotty atm. I would stand down for now and wait till it is fixed.
same here - all dead. maybe once office hours start and people roll in.
glad to see im not the only one
Hello. My generation cancelled because it was stuck in a loop.
Now when trying to generate a song, it says I should wait for other generations to finish processing before starting a new one.
@olive ocean see my message above #🪲┃bugs-web message
Ok thanks!
seems like now works better
Yes.. back on track!
I get "stuck in generation" every time I want to "get whole song" on the web. Normal generations work fine
Humm, a lot of time when making instrumental, i get lyrics in a bunch of different languages
I think because they are bussy
they only sound like language but it is mostly gibbrish. This tends to happen now and then. You can compare it with ChatGPT who is know to "lie".. It makes up stuff now and then. But the team is always trying to improve things like that. Stay tuned!
I'm not getting this message on the web so I wasn't aware of it 🙈
Thanks
Seems like when using mobile web (iOS) in my case, when attempting to add a track to a playlist, it returns a message like “0 songs were added to x playlist” when I go back to the library and do the same actions again, it usually will add the track to the playlist. So it seems like it’s got something going on where doing that action loop twice gets it to work.
The "Continue From" number resets to default if you switch between the Explore/Create/Library tabs and that seems like some sort of oversight since it didn't do that before.
I'll set a time that I want to continue the clip from, them sometimes go to Library or Explore to get some tags or whatever, then come back, past those tabs and make my changes, only to generate from the wrong time because the 'Continue From' box reset, wasting the generation.
I can't see why, from a usability point ov view, the number would reset like that.
Is it working again?
Yeah seems so.. 🙂
ok ty
Why i'm charged credits for this 5sec clip?
I try to piece together a 3 part song but its also "stuck in generation" everytime i want to get the whole song. Part 1+2 as one song works but when i do the same from part 3 on it will stay stuck. any ideas?
Did you remove a part? Trashed it?
thats all my parts i need from my libary
Could be that you have to try later... ATM we have some interferance with a part of OpenAI. And Suno doesn't have influence on that.
will do ... thank you for letting me know
Weird random jumping with two different songs. Prompt: progressive house sad melodic arpeggio, percussions, epic, big, loud {chords:| F#m | D | A | D |}
it's rare.. it's a job that tried again (it probably took a long time to render too) so you have 2 songs in one. Staff is aware of this bug. But thank you for reporting ❤️
Usability bug: I have 100 tracks in list and listen from top to bottom. Everytime i click down vote the screen scrolls down to the bottom. Really annoying 😦
@mc fixed my issue fast and nicely ❤️
Good to hear!
posted in general but here may be better: getting between 51seconds and 6seconds v3 first generation. is there issue, any suggestions
Was your problem fixed? I have the same one. But i Only have two parts and it gets stuck all the time
Brand new to the service here, but upon first use of the web app every time I generate a track it does so twice with the same title, same prompt, but different music. It deducts credits for both tracks but I have no option to only do one, so Im assuming its a bug. Anyone else have this issue?
You always get two, it's by design
In the generative AI space it's just kind of a thing where you get at least two generations at once
Since you aren't exactly sure what you are going to get
Ahh no worries then. Thanks for the heads up. I'll look into things more. Getting good results though!
are the dev's in this discord if i have issue
Guides are here to help you mostly... and moderators
all v3 songs i generate are now under 60 seconds, is there a setting i need to adjust
To enrich and extend your song in Suno V3, especially when you’re short on lyrics or specific prompts, here are a few steps you can follow:
1. Utilize Metatags: Include tags like [Instrumental], [Break], or [Solo] in your lyrics. This instructs the AI to create instrumental sections, pauses, or solos, adding variety and length to your song.
2. Experiment with Song Structure: Try different combinations of these metatags to see how they affect your song’s structure. For instance, an [Instrumental] tag can introduce a musical interlude, while a [Break] might provide a short pause in the melody.
3. Balance Lyrics and Instrumentals: If your lyrics are limited, balancing them with instrumental sections can create a more complete musical piece without needing extensive text.
4. Creative Continuations: When continuing a song, consider adding these metatags in subsequent parts to maintain the song’s flow and coherence.
5. Feedback and Adjustments: Keep experimenting with different inputs and settings. V3 is in development, and user feedback is crucial for its improvement.
Thank you i will try some of those, i've been using [Verse] [Pre-Chorus] [Chorus] [Verse 2]
Not really a bug, but assuming I have the commercial subscription, is there any Creative Commons license that I could release my songs under (given the TOS)?
sigh V3 Alpha ...continue song still not working and after a week getting worse. Will now not generate a new verse after prompting to do so, it just repeats the same verse of the original clip.
That prompt just gave me a verse and pre-chorus and the 2 clips outputted were exactly the same.
Why would your lyrics be limited? V3 alpha creates brilliant lyrics.
what was the length of the song
was trying to make a ska song and the word skankin gets flagged for moderation lol
Not so much a bug on the app end, but I just paid to renew my subscription, the payment was successful but the website is saying it was unsuccessful now I'm unable to access my membership including using v3 which was the whole reason why I renewed the subscription. I tried emailing billing but it has been 2-3 days and I still haven't heard from anyone in customer service. I've paid for a subscription that I can't use.
Howdy! As of this morning, all of the not-custom-mode songs I make have two outputs with the exact same title and lyrics instead of making two unique ones like it was before.
Screenshots attached from a test I just did.
Here's links:
https://app.suno.ai/song/b54ddaa2-4eee-457f-96a1-769197646a15/
https://app.suno.ai/song/3d311d94-2f79-427c-b553-0920fddb16f4/
Kinda mad about this since I wasted a bunch of credits today because of it. Please advise.
Yes a company that has their clients pay for them to test a system. Happy to pay once it is Beta and I think it will be good. but not happy to pay to be their tester. just my 2 cents.
follow-up to this ended up in #💬┃general-chat , in case anyone was curious: #💬┃general-chat message
@hushed crypt I hear your concerns and would like to address them comprehensively. First, it's essential to remember that AI in music generation, like Suno's V3 Alpha, is an evolving field. If V3 Alpha is not meeting your expectations, you have the option to use the more stable V2 version. Suno aims to be transparent about its capabilities and the challenges faced in this innovative area. Regarding the refund of credits for deleted songs, it's not a feasible option due to the high resource costs involved in creating each piece of music. AI technology, while promising, is still developing, and not every expectation or desired feature is achievable at this moment. Suno values your feedback and strives to improve, but it's crucial to have realistic expectations of what the technology can do today. If these limitations are significantly impacting your experience, it might be worth considering whether to continue using the service at this stage. Suno appreciates your support and patience as they navigate and grow in this pioneering field.
I see the team changed this deliberately. It's not a bug. Thank you for brining this up. I was not informed either. 🙂
I'm sorry to hear about your subscription issue. If you've already tried emailing billing without a response, a direct approach could be more effective. I recommend sending a direct message to @muted flame here on Discord.
It's important to clarify that the recommendation to use [instrumental] tags isn’t about limiting the quality or quantity of lyrics. In fact, V3 alpha does generate brilliant lyrics. The purpose of the [instrumental] tag is to help fill any remaining time in the song when the lyrics you've provided don't fully utilize the available duration. This way, the song can reach its full length more effectively.
If you have a paid subscription with Suno, you generally have the freedom to release your songs under a Creative Commons license. However, it’s crucial to thoroughly review Suno’s Terms of Service to ensure that your choice of Creative Commons license doesn't conflict with any terms, especially regarding attribution and usage rights. Creative Commons offers different types of licenses, so selecting one that aligns with both your goals and Suno's TOS is key for legal compliance and maintaining the integrity of your work.
It's been 1 hour and it's still generating
I am an official guide. And I am trying to help you. But if it is not appreciated, so be it. 🤷♀️
Excuse me, I am sure this problem has already been addressed and said many times, but I would like to ask you what tools you use to 'clean up' the annoying background noise that is added to the vocal from the third part that I plug in, and this happens on every song. The voice gets more and more squeaky and garbled to the point where you can barely make out the words because of the noise.
Man this Guy is angry calm down little Boy. Theres No Point in getting this angry.
Says he's made 5000 songs, but only joined Discord on the 5th of March. K
Are there any that are known to fit with the tos?
Bear in mind that V3 is still "alpha". I expect it to be buggy until such time as it is no longer alpha/beta.
I am not a legal advisor. But I wouldn't expect too much problem tbh. If you are a paid subscriber it's your creation 😉
well, I seemed to remember something about not being allowed to train AI with the output which is a restriction that prevents CC...
But reading the TOS I can't find it. So maybe its OK?
I guess that was a memory from another AI site
please leave me alone
Please take a day to time out and chill. There is no call for being abusive to our Guides. They are there to help. If you insist on being argumenative and abusive then you're going to find less and less people willing to try to help you.
I didn't see your helpful message, so my request for you to leave me alone was uncalled for, I apologize.
TBH I saw a wall of text complaining and stopped reading, but now I see that you were actually trying to be helpful
Unfortunately, you cannot use this prompt if you enter your own lyrics, can you? (like HD lyrics quality, i mean)
hi ! i have a question, im stuck in generation when trying to get whole song, is this normal ?
The individual in question has chosen to leave the Discord rather than endure a simple timeout. They were not banned.
Subsequent behavior in DMs (as seen below) resulted in an agreement among all the helpers here to ban this individual, which was confirmed by the Team before enacted.
🤦♂️ Meant to say "actually find yourself banned" but meh.
is there a better channel to ask a question ? for support
How long have you been waiting for the whole track to generate?
hours
Have you tried more than once to stitch the whole clip?
Did you by any chance remove some of the parts (trash it)?
no its a 2 parts song
On the webapp?
yes
Is this the first time you used Suno?
Hey, feel free to DM me your acc email and someone will take a look
Alright I've had enough. This guy is going to get his wish.
its pretty weird for mods to just put dms in a support channel XD
Maybe this song is corrupted then. It is rare, but does happen. @oblique verge
so ... i lost my credits ?
I like transparency. So I'm a bit of an odd mod.
If we were to just ban the guy after the first time out it'd look as though he was right in some way. By demonstarating the continued abusive behavior it becomes clear *why *we acted when we do ban the guy.
Are you a paid subscriber?
In any case deliberations do take place in private. So I'm off to signal my agreement in that section.
can you DM me the song clip you're trying to get full song of?
sure
I used to work customer service and been yelled at by people in person, its so draining. I feel your pain.
For some reason adding [Guitar solo] after a [chorus] seems to make the guitar solo play after the next instruction you give. So it ends like [chorus] [verse(example)] [Guitar solo] instead of [chorus] [Guitar solo] [verse(example)]
Hey, a bug i found today. If you like a song and dislike afterwards, the like wont disapear, so if you now put away your dislike, you can like it again, so it got 2 likes only from myself, which can stack up infinitly
profit
actually, I found this part:
In no event will you use the Output to compete with Suno, including to create a competing product or service. ```
well, that throws all use in Creative Commons or Foss projects out the window 😐
That's a restriction on the subsciber tier. I'd consider that a bug.
This doesn't necessarily prohibit all Creative Commons licenses but does restrict uses that could directly compete with Suno.
being that creative commons licenses grant all rights or almost all rights, and certainly don't have a clause prohibiting competition with a specific company, it does prohibit Creative Commons licenses
this is the most restrictive cc license: https://creativecommons.org/licenses/by-nc/2.0/deed.en
it has that "no additional restrictions" part
Creative Commons licenses are broad and don't contain clauses about competition with specific companies. While Suno’s TOS doesn't prohibit all Creative Commons licenses, it does impose certain restrictions, particularly around competition, which might limit your ability to use certain types of Creative Commons licenses. But maybe it is best to contact Suno via email to support@suno.ai. I don't want to give you the wrong information.
not to be disrespectful, but are you using ai to formulate these responses? It would be understandable given the volume, but... AI often gives repetative and just false answers
It is not a false answer. I am using help to formulate my answers in correct English.
My answers are purely based on the information I have on Suno.
You can find other Turkish people here: https://discord.com/channels/1069381916492562582/1159208016537718875
After parti için dans etmek istiyorum grup şeklinde ve bunun için yabanci bir şarkı istiyorum macerana gibi hareketler ekleyip
@low cypress teşekkür ediyorum ama şarkı istiyorum ben
Creative Commons licenses grant permissions, some then apply specific restrictions like attribution or non-commercial requirements, and then specify "no additional restrictions". Hence, if suno prohibits suno-competative uses of output, no CC license may be used.
As I said before, maybe it is best to contact Suno via support@suno.ai
I am a guide, I aim to help. Not give legal advice.
I want a song for the after party, how to do it, the moves are a foreign song
I emailed billing@suno.ai a week ago and did not recieve a reply. I tried again just now with support
This hasn’t happened to me before! I can’t use get whole song on one of my clips, it spins forever. Tried three times with the same result.
Okay it’s not just one song. I can’t get any songs. Get whole song is not working at all for me,
It takes many minutes to ‘get whole song’
And now it takes forever, it doesn’t work anymore
Can someone help me?
As far as I can tell no devs are on right now
Ouch
Whole Song is bugging out. The songs aren't stitching together, they're just sitting there.
can confirm this, yesterday it worked like a charm. Today "get whole Song" never comes to a Result
haha
the other user? you mean yourself? mr dj...
Did he seriously just make a new account to troll again?
I don't think they are trolling, this reads like they really believe what they are posting
They just uh, don't know how to articulate themselves without main-charactering themselves
insults people and thinks others got him banned...
Right
Out of touch, can't read the room, classic stuff
~as a paying customer~ myself, I don't like this person 
There my paying customer line negates theirs
I can't get a whole song I created. it's stuck again and again, while other generations work normally. tried 3 times past few minutes and stuck
get whole song isn't working for me at the moment. It's been sitting on two of them for 20 mins
I ain't reading all that
I'm happy for you
or sad that it happened
Ban evasion is a crime.
glass em'
My guess is they are a not-native English speaker, and maybe on the younger side
Always hard to say 'cause they read like a, man like, 14 year old?
lets forget him and move on
But I've seen adults warble away giant blocks of text too idk
Hi. How do i get the song to make a proper intro? Everything I've tried it either stats singing first or jumps straight into a beat. Thanks
Are you using version two or the alpha version three?
V2 on the website. I'm debating whether to buy the pro subscription but I can't get a decent intro on anything I make.
This is the part where we say "v3 is much better at doing what you ask"
Becuase it is
BUT you have to be a paying subscriber to use v3, so
and v3 is in alpha so you might need to throw quite some credits at it till you get what you want. just as a fair warning.
That's odd. To me it's a noted quirk of v3 alpha to not give you an intro unless you jump through some hoops, like using continue from time.
I haven't encountered it in v2 in a while.
Really?
May I ask how you start your lyrics?
it depends heavily on the song style. rock and metal you can get intros in certain ways, pop for example never got a single intro
in v3 i like to start with an instrumental piece and then continue with lyrics at a time i like.
V3 never gave me a single intro in pop, electronic lots, rock fewer, metal alot
It's the only way to fly. And now with that feature where you can start at any point in any take, it makes a lot more sense if you only need like 10 seconds of it
Oh so it's hit and miss then. Atm I'm a free user and the songs are good but not really usable to me without a intro. 50 free credits ain't enough to get an intro lol used 50 just now and not god an intro. 😕
/chip
Suno never isn't hit or miss. You become a lot less worried about credits though when you are a subscriber simply because you have so many vs. a free account
Yeah I am debating subbing to the pro account.
this
just extend
super easy
v3 allowing you to pick a time to continue from is the killer ap of v3 for sre
lol
it saves so much time
A ton yeah
ruins your lyric sheets though
OK so I'm probably gona subscribe to the pro account later. 10k credits is very appealing.
Never have enough credits 
hell yeah
welcme
facts
That is a downside, mostly because I get something I like and you get allllll the garbage I don't want out there in the lyric box
Feels like my laundry is out on the street
word
It's like hey check out this chorus and every iteration of it I've ever come up with
and by the way I'm not mad
please don't put in the paper that I was mad
I'm really annoyed that full song is timing out
I finished this stupid thing an hour ago and I want to share it
Ya'll think we're stupid, eh? Could have said so with less words.
NOTE: Ban evasion is a sin...
I think you're perfect and beautiful
People come to me and they say, they come to me and they say "Hey, what's up with Hepcat" and I turn to them and say "you mean THE Hepcat?" and they nod and then I go "S-tier, no notes"
Timeouts here too, that worked this Morning
and they usually clap or something idk
@tame latch Is there a limit on the length of a song, I'm at around 20min, and now that I'm trying to merge a new part, it just never complete.
Yeah, it was fine until 9:40ish
@viscid garden Yeah ! It's Rock/Metal Progressive... it sounds quite awesome 😛
You get full compositional decay at around 5 mins in. Longest I've ever done is 9 mins because the moderation goes wonky. But it's glorious
With v2 you'd pick up way too much distortion, but v3 does seem capable of 'cleaning' a lot of that out later in the song
But I've not pushed it far past 4~ minutes
But have done a bunch of smaller generations at the front for like, conversations and things
I need to try to do a crazy long experimental piece in v3
Yeah I kinda want to try now too
must resist urge to join in...
You know, when continuing works heeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeh dies
The Get whole song feature doesn't work right now or is it just me?
Same
And also my parts are disapearing like i cant go back up my list of generated parts
the server has some kind of issues. these are not bugs, just side effects of the server having a moment...
I am not having luck getting any prompts to generate
year its a little spotty at the moment... i had luck on the website to generate stuff, but stitching it together took a looong time for the server
Have never used the website, only Discord. Is the website better?
👀 Really? yes yes yes website all the way!!
it uses a newer version with better audio and stuff
GREAT, you saved my day buddy. Cheers and happy Friday
Hey, i kinda have the same problem, trying to get the whole song of a 67 parts song (yeah i might have a lil exagerated on this one lol, almost 60m song i think) How could i do it?
I think there is an issue with the "Get whole song" feature right now, cause I've seen a few people with the same problem.
"Smack it" as lyrics is rejected. "[smack it]" is ok. All the while, "motherf---r" is just fine.
Yes maybe, but i mean i understand that longer songs might be heavy process if they are running again the render the whole track dunno
That should be solved.. Hence I personaly have no problems atm.
Yeah.. I think next time if you do it like this, you should make full songs in between. At least that is what I do. You can still continue from a full song, you know?
Try homophones, worked for me a few times. Smae kit or something
I'll be very happy with :
1/multiples voices sélection
2/ Prompt length longueur than 200 c
3/ keed the perfect seed of a sound created
suno is not working at all
it is for me?
not working for me either
nothing is wrong on our end from what I can see
I'm having zero issues
Me too.. works perfect on my end
I've sent a few in to generate like I normally do but its been generating for around 10 hours without finishing. Did something change?
the create page is just constantly loading but never loads
been like that for 24 hours, not sure what issue could be
it works on microsoft edge but not chrome for me, strange
Well at least you found a solution. It should work on chrome too though.
Hey guys, with the autogen, from the default page, the way it's been working for a while, is that you prompt in something for an idea, and it gives you two different songs with different sets of lyrics. At the moment, what it's doing it giving you two different iterations of the same song. This is new, and I don't like it. Defeats the purpose of using the default autogen. Might as well use custom mode for everything.
They talk about it specifically. It's a deliberate choice and change they made
Autogen was already bordering on useless for songs with lyrics. Now there's no reason to do use it all
It's a bad decision
only if you are in front of a computer and want to type
on mobile autogen is much easier
makes sense
I'm not really a big user of autogen anyway
I use the one in custom mode to fix my meter and show me where the holes are in my song
Dude the V3 have so much bugs
Like it doesnt even says everything you wrote, or hallucinate and say the same things in a row, like wtf? I feel like the v2 is far better than the v3 man
I just hope they'll fix this shit bc its frustates me alot
easy easy...
I think he meant like "easy there", like telling you to chill out lol
Hi, my Library is quite empty, there was something like 10 page when i logout some weeks ago
and now less than one page. I just logged in because my community is asking for an extended version of a previous song.... Need to get it back in my library
It means you were naughty
I wouldn't worry to much about it. This is one we all see every day
Pops up when sex and violence are too explicit. Which is both a lot and not much at all depending on what you're talking about
Ah, I understand, thanks for helping, I'll see what I can do, thank you 😁
no worries
we try as much as we can...even on a Friday night
if we can't fix them, can't fix them for now...
maybe in v4, v5? 🙂
or perhaps even.... drumroll v3 non-alpha
Im having a problem where I've been waiting over an hour for my song to generate and it still hasn't. Any help?
No hury man. Much respect to you and the team, thanks for your hard works!
I believe all of you
@tame latch my work is lost ?
Suno V3 only does 20 seconds .. it siphons all my premium credits Seriously. This is F upped.
Seriously WTF.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Oh fuck this. WTF! 30-40 results 20 seconds flat...this is digital inflation WTF.
Suno hasn’t been working for me, through DM or on a channel. It can generate lyrics but the music never comes and eventually the reply that contains the generated lyrics disappears. I’m on the mobile ver. if that helps
Same. Nothing is coming out in discord DM
I've been having a lot of trouble with overly short generations today. Has something in the model changed?
I haven't seen any issues on my end other than the generations being flagged on me if I use the word Saxophone in the style of music as an instrument I'd like it to use in some instrumentals I'm trying to do. No matter how many times I try to generate a song using Saxophone it keeps flagging for moderation.
my songs arnt generating
Why chirp is not generating?
I am stuck with the same song prompt from yesterday for all the songs I am trying to generate moving forward
So, one of my favorite descriptive terms "GOES HARD!" was recently flagged by the moderation thingy. It didn't even occur to me that this could be a sexual inuendo until Suno brought it up. It seems very unlikely to me that it would pull up anything bad, as I've used it hundreds of times, and it never has. Just wanted you to be aware of it. I am a sad panda now
I have had things from lyrics flagged and completely refused in Portugese myself. 😿
Any AI is not very smart, but I actually can understand how it react to 'goes hard' LOL
It's a bit more complicated. If you get a reason, it's Suno's rule.. if you get NO reason, it's chatGPT who decided on that.
Kasumi above had problem with saxophone - I could understand the AI go weird on someone got the words 'blows the saxophone' in the lyrics.
But not when it's in the style as in that example.
All I got twas the new flagged for moderation, go to the principle's office error
That's ChatGPT
Thank you Daizy, it did provide a reason. So it was Suno that jumped on me that time - it was growl metal, and that genre is a bit graphic at times.
Well if it is not PG13... it can be tricky
Yeah it was not, still surprised it noted it in a less used language. As I do my stuf in Br-portugese only.
I don't know the deeper details.. so I cannot help you with that.. And for the reecord, I find it annoying too.. I cannot create a song with a lyric I wrote as a child 😭
Oh that is annoying!
I use my own lyrics also, and they are always on the dark side - cultural thing you know.
The text is a bit "violent" and now ChatGPT thinks I am pysho I guess..
Oh dear.
As for me, I bet the AI think I am either depressed or stark raving mad. (The latter one is ½ correct anyway so I don't worry.)
The new version of Claude blows ChatGPT out of the water when it comes to lyrics. It's still not perfect but it's miles ahead if you prompt it right.
You can even ask for things like the syllable stress of the lyrics it makes and Suno can somewhat use them with a bit of tweaking for more complex rhyme schemes like polysyllabic .
Do yall need help building a better moderation system? I just finished one that I'm really proud of, and it beats microsoft and mj in testing for accuracy.
That is exactly what I need, there's another where one indeed can indicate where the voice is to use emphasis and change the speed.
I used that one for a talking section in an animated film. But certainly need this for Suno songs also, as this AI have no 'feel' for the words used.
What kind of moderation - on Discord for for the AI?
I have no say in that.. I am just a guide. I know and hear shit.. and that's it.. 😬
No worries.
For the ai
I know next to nothing about discord
I got the same role as Daizy on Haiper and former Moderator on SeaArt. So I wonder.
Ok lets go DM!
cool, hit me up
That's supposed to be your local time.
But it might not be correct if you use VPN.
Sure you're not typing sexophone?
Anyway, I have wondered a long time about that annoying grain and hiss in the background of Suno tracks.
So I separated, and here it is: There's even a cuckoo clock there! LOL
its my local time, its been like 15 minutes and suno didnt generated the music
Oh you should have said so with your post.
Reload your page and try again, else from that I know nuffin.
do you happen to know any software or method to remove the grain and hiss from downloaded suno songs?
thanks a lot! really appreciate it 🙂
I have just started to use it myself, but these initial tests do wonder.
Enough of a waiting list there, so the info repeated in DM for you justDominik.
Chirp
Funktioniert aktuell nicht! Jemand ne Info? Beste Grüße ✌️
Everything works for me.. what do you mean?
yeah it's not working at the moment. But to be honest, most of us are on the WebApp which has V2 and V3 which are way better than Discord chirp who's V1
You shoudl try it: https://www.suno.ai/
I'm also have trouble with very short generations yesterday and today on the web app with V3. I think I need to wait until this is resolved before generating more. Sometimes the generations trigger a refund, but sometimes they don't. But just about all of them are useless the last couple of days.
Thank you!
I personally don't have it. I created 2 songs yesterday with no problems at all. Even made a video for it to educate people and show my workflow. So I assume you are consistently are doing the same. And it could be several things.
- Insufficient Content: If the provided text or lyrics are too brief, Suno might not have enough material to generate a full-length song. Ensure you're providing ample content for Suno to work with.
- Prompt Interpretation: Suno may struggle with certain prompts, including specific instructions on style, genre, or metatags. Review your prompts to ensure they're clear and detailed.
- Song Structure: Sometimes, the AI might interpret the song structure differently than expected. Use structure tags like [Verse], [Chorus], or [Bridge] effectively.
- AI Limitations: As an AI tool, Suno sometimes faces challenges in generating consistent lengths, especially with complex styles or mixed genres. Experiment with simpler styles or genres for more predictable outcomes.
Here the video I made yesterday: https://youtu.be/5uJw5kb05as?si=cEjrUthNYtUN-ooQ
The full song I created in the video, can be found here: https://app.suno.ai/song/33f4b016-c8bb-474d-9fe3-6c05b29ff9df
Discover the world of Suno.ai, your gateway to effortless music creation! Whether you're a seasoned musician seeking inspiration or a novice eager to craft your first melody, Suno is your ideal partner. Perfect for personal use...
There was a change a while ago which changed how downloads work for Suno on mobile browsers. It doesn't download a file, but instead brings up options for which app to share the audio or video with. This is convenient in some cases but an option called "download" really should do what it says.
e.g. here is what I see when trying to download:
That’s too long. Devs always have advices to keep it within 3 to 4 minutes. Above that it will take too many resources and might not be able to render.
I wasn't able to generate a song using the instrument "kalimba" in the style. The error said it is an artist name, but "kalimba" is a type of "mbira" or "thumb piano" that has existed for decades.
https://open.spotify.com/search/kalimba
I have 2 premium accounts. And I think it’s common sense tbh
I am an official guide but unpaid (volunteer)
michael_jackson_popcorn.gif
To be fair it's not obvious that longer generations would cause an issue. I've been here a while now and didn't know, just assumed it might take longer to render.
I want to clarify that as a volunteer guide, my intention is to assist and share insights, not to diminish your concerns. Suno, particularly its V3 Alpha version, is a groundbreaking yet complex technology, and we are all still learning its capabilities and boundaries. The advice about song length limitations is based on general recommendations and experiences within the Suno community. If you feel your concerns are not being adequately addressed here, I encourage you to reach out directly to Suno's official support team. They may be able to provide more comprehensive assistance and address your specific needs as a premium customer. Your feedback is valuable, and Suno takes the experiences of its users seriously to continually improve its service.
And for the record. I am answering you on a Saturday will being in my car. So my answers were short.
And I attacked no one.. Please tone your attitude to me. I don't need this. Feel free to look in my previous answers. A lot of people value my help.
I am doing this for the love for Suno..
Relax, dude.
Jesus.
I, along with many many users here, can assure you @low cypress is indeed an official mod here on Discord. I get that you're upset but there are more ways to go about things than angrily demanding things.
https://suno-ai.notion.site/Docs-Guides-Resources-38e5ba5856d249a89dcea31655f4fb74
"Email us at support@suno.ai if you have questions"
"Please email billing@suno.ai if you are unable to find the answer to your question"
This really seems like a billing question.
Welcome! See Guides and Resources for help using Chirp and Bark. Email us at support@suno.ai if you have questions, comments or feedback.
Curious if there's any intention to either undo this change or offer a way to use the original behavior? Whether or not I renew my subscription is somewhat contingent on this, since this change doubled the cost of everything for me without warning.
I found that very interesting actually. The songs themselves are different, but the prompts are the same. I'm going to try the same prompt here and see what I get.
Oh, I think there's a misunderstanding here. Suno will generate one "prompt": a set of lyrics and style, but two different "generations" of the same prompt. It always works like this.
Regarding the recent change in Suno's song generation with the auto lyric option, it's aimed at standardizing the experience across different modes.
What does "Flagged for Moderation" mean? and how do i know which bit is wrong?
Usually you can figure it out, look for "bad words" or other too suggestive wording. Sometimes, it just flags a perfectly fine prompt though.
ok thanks for the reply. I cant see any bad words, so might just be a bug.
No, it's not retroactive like that. However you might find if you get a subscription and start using V3 Alpha you will make songs you prefer over your old onees pretty quick.
You get a 20% discount on the year plans Suno offers, if you pay for the whole year. That is the only discount Suno offers atm 🙂
Any advice on avoiding the hiss or techniques for filtering it out, post-Suno? I actually have many, but they are only partially effective and they make everything else more difficult. But the hiss is a killer, so I have to.
Edit: Oh, I see you deleted a reply to justDominik about this. Can you DM it to me please?🫶
Your situation sucks. As a fallback, you can download the segments and connect them outside Suno, so at least all is not lost.
He did go from zero to Karen in two comments
People like that come in every now and then and scream and demand things, I think he was just nipped in the bud
This is an alpha version, it's going to be buggy. There are ways to ask for help which doesn't include "I will get you fired"
It was implied.
That's my take on it, I'm not a mod, just a user. My opinion means nothing.
He complained about being "attacked" which absolutely didn't happen.
I agree that getting the full song should either work for longer generations, or users should be informed about this on the site when they exceed a particular length, as not everyone keeps up-to-date via Discord etc. and it was news even to me that this didn't work for longer generations.
There are different ways to react to the information provided by guides/mods though. I think it was taken the wrong way. I've not seen people banned for being reasonable.
Look just pretend you aren't the same person
Clearly the mods know who you are
So if you are here again just pretend you aren't that person. It's easy, a bit of role-play
Take your Ls and just fade
That person getting banned over and over again fully deserves it, on account of spamming DMs to people who aren't even moderators, like me
So again
Just let it go. You know, if you are still here.
Were three times before so you know, odds are good
Huh huh. Just relax, you'll be fine. If it really was chaos then it wouldn't be working just fine and wouldn't have to keep kicking out a single person that doesn't jive with the community, now would it?
It's ok to be wrong, make mistakes.
Let's remember to stick to our rules: stay on topic with bugs and always be respectful, as outlined in Rule 1. If you have concerns or complaints, please direct them to support@suno.ai. Let's refocus on our shared passion for music and keep our discussions constructive. Thanks for your understanding.
Nope I double and triple checked it. I even showed my husband to make sure he saw what I seen in the text before hitting create.
The sax is just too sexy as an instrument 😉
Probably. Lol
Thanks for your reply! This is a reply to your suggestions. 1. The content I'm providing is several lines, consistent as I have been doing for dozens of songs, averaging about 16 lines or so - or 1 [verse] and 1[chorus]. 2. I've been using the same style prompt for several songs with no trouble - Indie Blues. 3. Song structure - I've been using the same prompts with no trouble before the past couple of days [verse 1] [chorus] [verse 2] [bridge]. 4. The clip lengths had been varying, which I'm fine with, ranging from 1:00 to 2:00. The trouble I'm having the past couple of days, is the clips I'm receiving are mostly about 10 seconds or less, give or take. It's a very stark difference, and the results are completely unusable.
here is your song -- I hacked stuff and get it together. Please don't concat too long again...https://app.suno.ai/song/5edbc3d2-52a6-4592-a5e5-ce689dee569c
Chillwave, Synthwave, Retrowave song. Listen and make your own with Suno.
I have problems with the login
@real bane @latent phoenix I tested it, and I am able to login. So I am not sure tbh. I do think you are both from Germany? Do any of you have a VPN to test if there are network disruptions on route? (just a guess, but good to rule out)
had the same problem. use google chrome and it works. greetings from germany
@wind wedge which browser gave you the problems?
opera gx with vpn enabled
Thank you! Maybe that's it.. let's see if they return.
no problem 🙂
Has anyone solved the degradation of sound in V3 Alpha from part to part? Part 1 is crisp and clean, and then by part 3 or 4 every continuation is more static and muddied. I can't seem to fix this by repetitive re-dos. Is this a known consideration, and if so, how does one solve this transition from clear to unlistenable in parts stacking to finish a song?
you can try putting different part 2s or 3s together in video editting software.
Apparently using "get whole song" and continuing from what that creates can help
(rather than continuing from individual parts each time)
But the original generation is bad, so editing it only seals it as being unlistenable.
You said it gets bad by part3.
Trying this now.
Wait, does that help with getting a new voice after an instrumental part?
Every continue results in a loss of clarity. I try to fit as much of the song in Part 1 as possible to avoid the next generation degradation.
No idea. Just heard it mentioned a few times that the quality degrades if you continue from individual parts. Though also it may help to give further background context for it to continue generating from.
This is the bane of my existence. I really struggle with this. But the best I've got so far...
Listen to the noise and choose the "Continue From" timestamp carefully. Think of the noise like an instrument in the mix (I think a lot of the noise in my songs is trying to imitate snare rolls, poorly). Do not pick a timestamp of a time when the noise is in the middle of getting louder -- guaranteed it will continue getting louder in the next part (like a snare roll). Pick a timestamp when the noise is minimal or getting quieter. Good chance it will continue getting quieter in the next part.
But that can really limit your freedom when it comes to choosing continuations. It hurts, and I hope the model gets better.
YES!!! "Bane" is just the right word! I can't really figure it out. It just degrades the longer the song. So are you saying you can continue a song without letting it finish the version? You do, what? Stop it mid part and then choose "continue" and it will ignore the rest of the original part? Or am I missing something?
I will try that! I find that when I continue, the next part will pick up and "finish" the last word or two that got cut off from the natural end to the last part. So you're saying if you choose, say, 29 seconds into a 59 second part, the next "continue" will start up at 29 seconds and ignore the rest of the part?
THE SERVERS IS CANT PLACE /chirp I CANT MAKE MUSICS
Yes and no. It will still try to continue/finish whatever was happening exactly at 29s, which is likely to finish a phrase, etc. But it will ignore whatever was between 29-59s in the original. But that's the thing -- if the noise was slowly getting louder at 29s, then it will continue getting louder in Part 2 (logical continuation of whatever was happening at exactly 29s). So don't pick a continuation time that has 1) loud noise or 2) quiet noise getting louder.
Hello, I found a funny bug
Thank you! I will try this!
Super fancy! Well done! Lots of hard work! I hope to find the same stamina as you. Right now, I just burn credits. It's about 1k credits per finished song -- and there's always a glitch or two even with that!
Thank you for posting this. I will certainly not abuse this information....🤞
I notice the number reverts to the correct display after refreshing.
Yes, I noticed that too, but the bug is funny because it doesn't break anything and doesn't help, but it's a bug
Definitely. I suppose if there has to be a bug this is the best kind.
Done! Thanks for the help!
true
or it is a feature to combat boredom XD
I sincerely hope I am never quite this bored...
I have tried to separate it in lala.ai - what app did you use, as I had no luck with lala, and if you used lala, what instrument did you choose to separate?
Hey I'm trying to make a song for my little daughter , her name is Elodie, it just doesn't let me and keep saying it's a name of an artist!!!
try to do elo-die for now or elo- die
El-o-die
El O Die
If a song is only half way done and stops there (usually at 00:55 s so it still has time) are you able to get credits refunded?
Figured it out but it's annoying ...
sry man
hello?
no
are there any tips for repeating the same chorus? i have the first Chorus which has great backing vocals but they arent there on the second and burnt through about 400 credits so far trying to get it to repeat it. Any help appreciated
you can try to continue from the middle of the first chorus and see...
I did try that and it lost the backing vocals on that one
i hope v3 will fix the static and muddled voice when i continue the song part. I have generated great songs but the voice always get muddled in the end.. https://www.youtube.com/watch?v=VHeXDRL5qtw
Provided to YouTube by DistroKid
Phantom of the Battlefield · Midnight Walker
Phantom of the Battlefield
℗ Midnight AI
Released on: 2024-03-17
Auto-generated by YouTube.
hello, when i try to make a part 2 of a song in the website, it makes like 9 seconds
is it a bug?
what i do is click on continue from this song in the first part, then i put the new lyrics in the part 2, delete the field of style of music and press continue
but it only generate 3 to 9 seconds for the second part, the lyrics lenght is the same
From March 16th, Suno V3 has been easily generating short tracks. The continuations produced using the 'continue' method also fail to seamlessly connect with the previous segment, wasting quite a bit of time and hundreds of credits.
This is the weirdest bug I’ve ever seen on any software. In V3, If you type in glitch or related genres, it will literally glitch. Wtf, how does that even happen? Sometimes the song just doesn’t come out and it refunds your credits, sometimes this song is just weird noises which is not entirely what glitch means as a genre, only somewhat. Sometimes the song is super short, but it doesn’t refund your credits. Makes me wonder what other words do things besides affect the genre.
It's not that weird most AIs are very literal.
my music heve not a voice is only music !!!
10 cedits for nothing
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
and I have no idea what that is https://app.suno.ai/song/e4030084-eef8-4e9e-bc6d-2e7a250b69b6
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
whyyyyy https://app.suno.ai/song/21f1b8fa-0830-4e44-b8b3-a37cfc99f285 its all only music https://app.suno.ai/song/2b44db53-55e2-4aeb-a44b-888a146e908f
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
@hoary terrace because music like this often have a big intro in real life. But if you continue the songs, they usually will pick it up. I took one of you songs, continued it and both start to sing.. One at 15 seconds: https://app.suno.ai/song/7e7a0d34-f071-4c17-a625-c9e168c6548c
And the other on 43 seconds: https://app.suno.ai/song/0a7fd557-bdfb-4ca5-a31a-f52931e7c072
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
hardbass. hardsyle. techno. fast. 180 bpm song. Listen and make your own with Suno.
Thanks
please provide a full screenshot of lyrics and style of music to understand what's going on here.
No.. that is not what I mean 😄
I mean like this:
no
Then you will have to hit "clear" and try to "continue from this song" again
clear, refresh page
can you continue from this song ?
Then I am out of answers. It normally works for me..
i can continue from new songs, but can not from older
oooooooh.. this is an old song?
Yeah.. I heard there are some old songs. you cannot continue. This is a bug, and I don't know if Suno still can fix that.
Sorry man! I wish I could help you there.
OK. thnaks
Yeah I know.. but if you refresh, the count is back to normal 😉
I separated the song track, and then split voice from other sounds. But there's also odd noises to be found in bass and guitar channels for example.
So taken together, part of the mystery is solved.
Lets say you work in a studio, you will hear each instrument sounds clear and defined.
This should not be seen as I am complaining - but an attempt to provide feedback to make Suno better.
But yes I return here to add one funnie, and no idea why or even how Suno did anything like that.
It was not in my lyrics, but at the very end Suno inserted a nonsense sentence. "Te encheu o cor Pocumu".
I have no idea how, and it got nothing to do with context of the previous lyrics either.
...now I have digged deeper, 🔬 and found that Suno in one take of the same lyrics replace Relembre with Recontar at one point.
The AI is obviously trying to be smart, but 'retelling' and 'remembering' are not quite synonyms!
Edit: And the real price winner was found on a track I did not intend to use anyway, but lets translate the lyrics to make this madness entirely clear: "Protect me from despair before your foot masks me once again." 😹
This happened since Suno reshuffled the words of my original lyrics.
When generating lyrics today, I always get the second verse directly after the first. Is this new? Or is the problem on my end? Regards.
V3 sometimes (depending on your lyrics and genre) has problems recognizing what is verse, where to end it or where to start a new ellement. Are you putting in [verse] [chorus] etc? If yes, you can try to add [break] or [instrumental] in between them
I just wrote that as an example, write a song about love: And then [verse1] immediately followed by [verse2]. Wait, let me show you:
No matter if it's v2 or v3, and no matter what I write, it's been happening like this since today. The days before, it was normal.
I had this behavior with V3.. it's normal for me, depending on genre and lyrics
AI can be unpredictable.. but if you follow what I said, it could help.
Okay thank you I'll try 🙂
As for V3 it's said we should change that on the create webpage.
But I cannot see any toggle for that!
are you talking about using V3? And not seeing how?
Indeed, ty for fast response.
that option is only available for paying subscribers. If you are you can toggle it like this:
I had a hunch, no worries - I'll live with that
Oh I was not even suggesting that, and I will remove that post of mine also.
But I've had discussions with several who is in the same situation as myself.
Luckily I can write my own stuf and use MIDI also.
Hello, I think I found an error in the app and I would like to know if someone can fix it.
The free version is supposed to give you 50 credits and if you run out you have to wait 1 day for them to give you the points again but 1 day has passed and they haven't given me the credit yet
Hello,
Today I logged into my Suno account, but in the library there are missing all songs generated on one particular day. It was approximately 1 week ago. I think the titles of the songs were my girlfriend is hungry as a wolf.
Can you please solve this issue for me?
Make sure you are logged in with the right account.
I am 100% sure I am logged in the correct account since it was generated using V3. I don't have more accounts where I have paid subscription
Have you clicked on "library"?
Yes, I can see all songs except for this one day
Last thing I can think of, did you look in your trash? (maybe by accident)
Now I checked. Nothing 😦
Am I the first one experiencing this?
😦 Oh that's sad... Yeah I am practiclly living in Suno tbh.. I never heard this one.
If I provide to you my account email, the date and time of that generation, you cannot take a look in the system history to check where is the problem?
I am just a guide volunteer, so I have no access to back end, but you could send an email to support@suno.ai
Ah, ok 🙂 Thank You
My apologies, now I found it, it was indeed in the library. Only problem I have so many generations, it was lost among them 😄 . Sorry again for inconvinience.
No problem.. Glad you found it.. yeah.. they are working on a better library, but they are still a small team, so stay tuned and enjoy your creations ❤️ @tardy topaz
made music but it's not playing and there's no lyrics for some reason even though it did took my credits off
faced the same now, a few clips just don't have the "continue" enabled for them whatever params I try to set
Almost every song I generate with V3 mentions the genre(s) i used to set the style, within the ai generated lyrics. Eww. Still a lot of overdriven distortion sounds in the mix.
Not sure if this count as a bug, but I want to generate something with the instrument "Kalimba" and if I try to do so, it says that there's an artist called Kalimba... So i'm basically not able to use that kind of instrument. A suggestion at the same time: A new option for choosing instruments : D
Lol, I reported that same thing just yesterday. You can still use other terms for the instrument. I was not successful at getting anything that sounded like a thumb piano.
It seems that just a few days ago I when using simple creation it would generate two versions, now I only seem to get one.
On the site I liked a song I generated, downloaded it but later when I saw it was not present in my library.
site keeps making 2 of the same songs over and over but just with different music and voice
is this AI supposed to recognize every genre? It does not know how to do Frenchcore, anyway
Not exactly supposed to, but it does have the capability to produce most genres. However the words you might need to summon a particular music style or less common genre might take awhile to figure out exactly. So what people do is web search for other words people use to describe genres frenchcore and try those, all together. "techno, speedcore, dirty, fast, offbeat bassline, hardcore" is on the front page of Google for me. (I don't know the genre so that's probably not a great prompt, but as an example.)
It doesnt even know hardcore
Just gives generic EDM
Very dissapointing imho
Not even eurodisco…
I've used it to create Blues music, and honestly, it's better than the real deal! Perhaps it'll take some time for the model to learn from our inputs, but for the genre I've tried, it's truly a shining star! I love it.
What it knows, it knows. Question remains is how it can learn more and correctly do certain genres n styles
you need to learn how to prompt it
Uhh
Literally ”hardcore” does not give hardcore. Literally ”eurodisco” does not give eurodisco
These are well known genres
maybe in europe, try to find a way to say it differently
Lol how?
what makes up those genres, describe to it what you want
i've told it to do a 7/8 time signature and it does it
These aren’t subgenres
Tropical house is a subgenre. It knows it and therefore correctly generates it
But the whole chain down to hardstyle or frenchcore it doesnt understand
i mean you can ignore my suggestion and keep repeating yourself too, that works
Further development and/or additional data is needed
but you may want to include the word techno when trying to prompt it
Hello cptn obvious
FRENCHcore is a subgenre of HARDcore techno.
"hardcore techno 200 bpm" would likely get close to what you're looking for
Maybe read
Speedy hardcore techno gave me slow paced generic EDM
…
did you type what I said? im literally telling you how to get the result you desire but it seems like you just want to be mad
Lol strawman
its AI, you have to word things differently to manipulate the results
My opinion is that it should understand well known genres with the correct words to describe those genres
neat, its not a bug but maybe a suggestion. and they may be well known to you and where you're from, but maybe not other parts of the world
Its very not user-friendly otherwise
It depends on how it’s described to function
Which is why I originally posted here
that's the beauty of it, trying to figure out how to prompt it to get your desired outcome
Not if that is not how it is supposed to function
i bet you still haven't gone and typed in hardcore techno 200 or 220 bpm or whatever, just try it my guy and stop wanting to be mad
its not that deep
Again, the strawman type gaslighting
stay respectful please
LOL
brother i literally just told you what to type, just go try it
He is only trying to help you.. consider this a fair warning
Have a nice day
Anyone else have problems with v3 only making short clips and not the full 2 minutes? I put my lyrics in but only makes like 20 to 39 seconds and doesn't finish the lyrics.
yes same here, but i keep regenerate as its an AI performance, nothing guarantee but it works 90% of the time
ok, submitting this as a bug
This is getting flagged. Asked chatGPT which suno uses (AFAIK) and the prompt is getting flagged. Another user tested it too with the same error. here the lyrics so far.
[Chanting]
We are the Helldivers
Bullets from the sky our spine shivers
crash landing in a steaming pod,
its time we show those bugs their god.
[Chorus]
No matter if bugs or robots,
We are not here to miss shots,
Cant you hear the screams of our comerades,
lets have these insects with our gamma rays
okay, so the last line - gamma rays - is an issue. this should not be an issue honestly...
What kind of style are you using? And which model?
the one on the website, I dont have the option to change the model there. styles were: metal, hard rock
Are you using a free account then? Paying account users should have the option for v3 alpha in a listbox.
I'm seeing the same issue. Huh...
Yes its a free account 🙂
Hmm...having the same issues. Not sure if it's the gamma ray or the helldivers that is causing the issue or if ChatGPT is getting snippy because of the violence. I'm going to link this in the moderator's chat to see if they can follow up and divine what exactly is causing this hiccup.
Since it is probably triggered by ChatGPT, I aaked the source:
The text you provided contains elements that could potentially be flagged due to violent content or themes. Phrases like "Bullets from the sky" and "crash landing" may trigger automated systems designed to detect violent or aggressive language. Additionally, the mention of "gamma rays" could also raise concerns as it suggests the use of potentially lethal weaponry.
Yeah, I've come acustomed to how ChatGPT thinks. This is an issue. It's why I escalated it.
Not all songs can be poppy, nor would we want them to be.
I removed gamma rays, and it worked. So it was deffo that. I even asked ChatGPT itself and that was the ansswer:
Based on the provided song text, there doesn't appear to be any content that violates OpenAI's usage policies or terms. The song revolves around themes of combat and warfare, which are common in many forms of artistic expression, including music. As long as the content remains within the context of fictional storytelling and does not promote harm, discrimination, or illegal activities, it should be acceptable. However, it's always a good idea to review content for compliance with community guidelines and standards. If you have any specific concerns about the content, feel free to ask, and I can provide further guidance.
There does seem to be a very easy way to crash the website client-side, almost consistently. Windows 10, Chrome browser.
I've got an ultra-wide monitor and will use the windows key + directional arrows to throw the window to one side or the other. When on the 'Explore' page, specifically, I can get it to consistently crash when Windows + left or right.
Does not happen on Library or Create, and I suspect it's do to the reactive elements on the Explore page, but I'm not sure.
Granted you guys are still working on it, and it is client side so ??? if it's constant, but perhaps it's helpful idk.
A refresh fixes it right away so minor minor.
We are currently working on resolving this - thank you so much for raising though, that is helpful!
hi i want to use suno ai but that dont work : i do the connection with my discord account and that work but after when i go to create ,the little circle turn and turn and nothing happening ,no creating buttons ,nothing can you help me please
soemeone is here ?
no need to post double..
The player on New/ Explore just skips forward, even if you press the backward button.
also the trashbin has no more page indicators or ways to switch pages... only 1 page visible.
Why does /bark not load voices so I can choose?
Bark is no longer active on this server. Suno is concentration on chirp in WebApp
heres a tip.. spell out the words phonetically and it wont flag 🙂
Please, what is this?
There is no way to ping a Moderator or Team member in this channel O.o
@royal mason
sorry just saw you come online as i typed there was no way to flag the team. just reporting spam
@tame latch .-.
How come the quality of my songs is declining? Is it possible that they take version 3 away from me even though I am a pro user?
no we won't do such a thing...I hate to say it but a lot of it is a matter of luck...also, really depends on what you are generating as well
Hmmm what? The quality of my songs were really good and now they are really bad. Can someone please see why
As said, luck of the draw. Sometimes you have good generations, other times not so good.
There is nothing that has changed.
Well here is what I think is happening. That can only have so many people testing version three so they kicked me off because I can tell a huge difference
That isn't how any of this works though
I don’t think I am on version three even though it says I am
I am not being mean, you have a fundamental misunderstanding of how Suno works
But it happened before too
You know how some days it rains?
In real life I mean
It's that, sometimes you don't like the generations you get
Nothing has changed in any meaningful way, besides you don't like the songs you are getting.
Well how can the songs be really good and then all really bad suddenly to the point I have to report all of them
You aren't going to like the answer, but it's you. You don't like the generations coming out of Suno. If they are low quaility then that's just what batch you've got
It happens to me, it happens to all of us
Sometimes you have a not so good time
It's the nature of how LLMs like Suno work
I should know, I've spend 15,000 credits so far this month
Wait do you work for them?
I do not. If that means you don't trust what I have to say then:
#🪲┃bugs-web message
tony does work for them
and has already answered you
Suno does no such thing
It is a matter of luck
I'm just expanding on his answer
There is nothing wrong with you or Suno, nor are you somehow being secretly put on V2
Okay well I am wondering if they only have a limited amount that can test out version 3. The quality is giving me version 2 now
It keeps being said, that can't actually happen
Why do you think v2 uses less than v3?
If you want to ignore what tony has told you and think he's lying to you then you really can't be helped.
No clue but I can tell a huge difference
I do not doubt that the generations you are getting aren't very good
I am saying they aren't very good because of luck
You already know some generations aren't good
Well I had really good luck for the first half of my monthly credits and now really bad I guess
Yeah, that's all it is.
I truly beleive you, so does tony
BUT it is just the reality of working with something like Suno
I know that isn't a satisfying answer, and it would be nice to go 'ah yeah it's this thing' rather than having someone shrug at you
Well shouldn’t it be more AI skill instead of luck?
Well Suno isn't a living thing
It's trying to predict what you want to have presented to you
But suno is being trained
Not in real time
It doesn't work like that
V3 is the same v3 that was released weeks ago
The training we are doing is for v4
Which will be one giant release all at once
Well all of the vocals are off key now and they weren’t before
You are generating different songs than before, correct?
As in, you have your good songs you like you made before, when generations were good
Im generating the same type of songs but they have gone downhill over time to the point I don’t even know what the vocals are trying to say
Would you like to share your prompt? Maybe I can help guide it in a different direction to what you'd like
Let me give you an example of something good and then really bad
Sure
I made those lyrics
Hard to exactly help you though since I can't see what your prompts are
Right but what was the musical style
Is it the same as what you are trying now?
As blues tends to be
And off key
I have used blues mixed in with other prompts though to guide the song in a more soulsy direction
Doesn't always work but hey

I think I see an issue already
Just on the prompts you are using for your musical style
Okay
Are you wanting like a single singer and single guitar type thing?
Yes
Give me a moment and see what I can do
So here is the bad one
That’s the crap I’m getting
But I have tons of songs like that that sound awful
Thank you
Yeah I understand. I need a few more minutes because irl stuff, BUT
Can I ask you how you view using your credits?
I only ask because the way I use Suno is if I don't like a generation then I simply don't use it. It's garbage, trash, etc
Each song I make that I like is made up of a lot more credits than it took to make it
Some might say 'wasted'
soulful, emotional, female singer, angelic voice, crisp, clean, heartfelt, guitar song. Listen and make your own with Suno.
soulful, emotional, female singer, angelic voice, crisp, clean, heartfelt, guitar
That's the song style
Those are just generated lyrics
Yeah but at first my songs were quality and then they gradually got worse. It happened before too when I paid for a membership and got rid of it and signed up again with this new version but it is starting to happen again
That’s really good
I know you really feel that is what happened
I will try that
Well it’s happened twice
So avoid stuff like "sing it really nice"
and instead just use 'emotional'
Whatever you want Suno to do, get it down to a base word
I will do this and show you result
soulful, emotional, those are good ones for singers
I know. I just want to assure you that it would take far more work for Suno to target people directly, somehow
Than simply Suno to not be cooperating with you on account of it being a weirdo AI system
It does that with everyone
Including me 
dude you don't know the amount of crap that I have to listen through testing things
thx for your help
syr for getting to this late, like @viscid garden said, a lot depends on how you "prompt" it
I can promise you that we didn't change anything on v3 in the past days, and if you are selecting v3, there is no way we ship you a v2 under the hood.
So here it is and you can hear the off key
Keep listening until the end
Well I don’t know if you notice but I guess it is better than some of the others
this doesn't sound bad to me...
Okay
I will find another
Well I will keep generating and let you know when I have an issue
I need to sleep...there will always be issues
Oh sorry
sry man if your luck isn't on your side today 😦
I appreciate the feedback though
but work on your prompts
Okay
do things like clean vocal, acoustic guitar, slow/medium/fast tempo, emotional
Okay
and make your lyrics slightly longer -- right now each line of your verse is very short
good luck
emm maybe we should really do a better job at helping users prompting
probably do some prompt live assistant?
@tame latch enjoy your sleep, but I am down to test these systems, let me at em'
Why does V2 music sound clearer and crisper than V3 which sounds almost as though it was put through a bitcrusher?
In terms of creativeness V3 is better but its a huge downgrade in terms of fidelity for certain genres (maybe all?).
I am a Suno subscriber and would love the sound quality of V2 with the generation ability of V3.
because V3 is in Alpha. Which is a very early stage in development and not finished.
You couldn't pay me to go back to V2
I don't know how you are prompting Jade but I do not share your experience at all.
I'm well aware of what Alpha is. I have tested plenty of alphas. I was just wondering if that was intended since the banner says "Help us test and improve our newest model before the official release." Which has a sort of "this is going to be released soon vibe" rather than alpha vibe.
I'm a producer and engineer and I can hear a distinct sound quality difference. The melodies and ideas are vastly better in V3 but the sound quality is lo fi
That may be intended as to save processing power or something but yeah its a big difference.
Take a look at the last mini poll. The devs are on it. There is a looong list of improvements for v3. it is going to take time though.
and trust me I feel you on the sound quality of v3. but all we can do is being patient 🙂
Thanks. 🙂
Well it's a fantastic tool which I will continue to subscribe to and follow its development.
hmm... right now, in V3, main clips are < 50secs and continuation < 30secs. Technical problems?
It’d be really great if you made a guide or something on how exactly the lyrics are prompted to the underlying model, especially when you do continuations (say you’re cutting off a previous generation after a certain lyric, what should you enter as the continued lyric?) Cutrently there’s a lot of trial and error in the dark involved.
Not sure if this is a bug or oversight but you cannot put songs from 'New' on Explore page directly into Playlists like you can with 'Trending'. You have to click on the song link and then Add to Playlist.
This happens time after time, I write a song which fits within the 2 minutes allocated time easily. Write it as Intro, Verse, Chorus, Bridge, Chorus, Fade, End. Should take at the most 2 minutes (based on the amount of words and song progression). Suno sings almost the whole song and at the last two lines it just stops, after less than a minute and a half. I specifically build the song to be finished fast, and it still doesn't do it.
My normal length songs of 3-4 minutes it sometimes can accomplish with 3-4 parts, but the songs written to accommodate the one-part time slotted it struggles with.
Меня растроила Альфа V3 тем, что иногда когда песня готовится, она может взять и остановиться на слове, или вообще на каком либо месте. Песни не могут кончаться в принципе, это расстраивает. Также я заметил, что иногда даёшь текст (я писатель и поэт) нейросети, даёшь ему полностью данные где припев, где куплет, а нейросеть тупо забивает и делает по своему, хотелось бы чтобы нейросеть контролировалась немного понятиями [Verse] {chorus]. Также разочаровало продолжение песен, первый фрагмент был отличный, уже думаешь что можно продолжить давать нейросети продолжение и внезапно все продолжения даже если ты пишешь текст просто оканчиваются спустя 10-20 секунд.
Естественно за эти эксперименты забираются кредиты.