#📚┃suno-school
1 messages · Page 86 of 1
This happens when you're uploading copyrighted material 🙏
I never actually listened to them before today. Cool sound, lots of record popping
I haven't tested it, but try playing with this, you will know what parts fit and which ones should be something else:
Dream Pop Jazz Fusion, Dark Ambient, Breakbeat Rhythm, Female Vocals, Melancholic Atmosphere, Hypnagogic Echo
Has anyone used a persona to make a song, then use that new song to make a new persona?
And then use that new persona to re-do / remix the original song that the other persona came from?
Yes. Yes. No.
What I did: I made a Persona (Kea) from a V3 song I originally did. took the lyrics from a totally different song I had done in V3 and did a new version in V3.5. Really liked how that song turned out and now I made a new Persona (Kearna) from this new V3.5 version and I am trying to see if I can get a similar or better version to the original V3 (Kea) song I had done.
Trip hop, acid jazz, slow tempo, sultry female vocals, jazzy, drum and bass <-- gave me AMAZING results everytime I did it. I've got 10 songs in the Portishead style
Trip hop, acid jazz, slow tempo, sultry female vocals, jazzy, drum and bass song. Listen and make your own with Suno.
Ancient music mode UNLOCKED
Grimdark ambience chanting, tribal grimdark marching, industrial marching, slow grimdark hell chant song. Listen and make your own with Suno.
Since yesterday I can no longer put my new songs in public status. I can check "public", but once I refresh the browser page, it is private again. I tested with another browser and another PC it's the same. I am in France.
Do these songs start from an audio upload?
is there a link for phonetic sounds humans make?
there have been some odd things happening on the browser based version of the app.
- Credits leaving then coming back (not such a bad thing)
- Songs timing out
- Public option not working
For me, a browser cleaing (cache etc) and restart seemed to work. It also may just be timing and it was a bug the team fixed
yes "cover". but it's not the first time i post cover in public
Is it here pls support from suno live?
If you use "audio upload" and the automoderator detects vocals, you won't be able to set the clips to public.
Is it possible this is the case?
I sent an email to support but I haven't had a response yet. I hope it will be fixed soon
support@suno.com is the best place to contact official support, but the community can try to help with some things 😄
I have same problem i waiting 7days for emil
But i need it fast because tune core need confirm im owner of this music
Just to check, are you asking Suno to send you proof that you're the owner of the songs?
I'm going to test with a song generated without upload. But it's strange because I've already put songs in public with uploads
💪 Case-by-case basis -- the platform changes, so just because something worked previously doesn't mean it's guaranteed to work forever
@charred merlin Yeh they send my email if i pay for subscription i can use my songs for comercial using in thath moment 5music i chose before i had subscription
ok thanks i will do some more tests
Why?
If it helps, there isn't anything Suno can send you -- you need to show TuneCore your receipt for Suno (should have been emailed to the inbox you used to set up the account with) -- just a matter of showing them that you had a paid membership during the times when the songs were generated (you might need to send them the screenshots or suno.com/song pages for the songs)
Duno have same problem somerime
Almost every instance of this that I've seen recently stems from something in the style prompt or lyrics being flagged for moderation. Songs in languages other than English seem to receive false-positives more frequently
@charred merlin I already send my back transfer with suno email about it
Thank anyway
ok I did a test and, it's the songs with an upload that I can't post in public 😭
Yeah, any uploads that are flagged as containing vocals are restricted from being made public -- sometimes it can impact songs without vocals that the AI accidentally mistakes as vocals, like digeridoo noises
You can still share the song link & place the songs in playlists -- it's just that the songs won't show up on your profile page or be searchable through the site.
hi there can someone pls help with this problem that when i try to change 1word it only change the first one that i do but when i try to change the rest of them that do not work it say it the same as the original word and i was the replace section to change the word is there a better way?
how can i make my audio i uploaded appear in a song because they are not
What are your best prompts for getting the best and most realistic sound vocals?
https://suno.com/song/61252c97-6007-43ed-a9b9-e6a28644368e how does this person make the fart sounds?
was replace disabled for uploads today? no matter how much I select in my uploaded clip, it doesnt let me recreate
only for suno version songs. is this a bug? (i realize the screenshot shows I have no prompt. I am aware. Even when I put a prompt, the recreate button does not light up)
this was not the case yesterday, so I ask if a soft patch has been pushed.
thats not cool. if a feature is going to be removed, it should be addressed
I used that method a lot
pro tip
- if you haven't used replace that much, here is some fun little info I just found out that worried me.
I thought replace was disabled for uploads (in my opinion, get rid of the whole 10 second minimum for replace). its not disabled, but your clips need to be more than 10 seconds long even though Suno accepts the upload. Replace won't light up if your sample is too short which doesn't make sense given there is a "make same length as selection" button. I usually have that off to let replace make a whole new song for me rather than extending or covering.
confusing, but problem solved
As of now upload a track of silence for 2 minutes and cover it with all the lyrics and style of your song. The quality of the vocals will be very noticeable. Or, you can play with persona. Persona and Cover make better quality Vocals for whatever reason, enough to start the rumor about v4 before v4 was announced. If you want immediate results, then I recommend doing a cover over a silent 2 minute track.
Hey, I have a question after I have 'wasted' (I have learned a bit, but not that much) over 2000 credits:
I wanted to create a song with various styles from multiple genres, and I managed to combine those for my verses, a pre-chorus and chorus, but whenever I repeat the structure, it really wants to do its own thing, so I have to work with very small snippets (resulting in a lot of 'wasted' credits). Is there any chance to reference the melody or rhythm of a certain passage from which he can be inspired or copy it from? I tried with [repeat mm:ss.ms - mm:ss.ms] but so far it is mostly by chance.
Does someone have a better approach working with V3.5?
There are structures that Suno adheres to in a more consistent way. It has been talked about in this channel. If you look at the instructions it actually tell you what Suno responds best to. If your Vocals get too complicated in ways like too many tags or too much variance in amount of syllables per line and lines in that section. Intro, Verse, Pre-Chorus, Chorus, Verse, Bridge, Chorus, Outro tends to work many times as long as you follow the things I just said.
Oh you got a good answer coming
Well, color me unimpressed by the Persona feature. Instead of creating just a vocal style or whatever, every persona utterly ignores any other lyrics and regenerates the song it was based on for me. What am I doing wrong, if anything?
There's no simple answer. It can be done with lots of effort and various tools, but nothing as simple as a prompt to reference a time like that. Maybe in the future some more simple methods could do it. Better control over song history for example.
I just want the same vocals with a different song. Am I gonna have to do the replace section thing?
You have to make large changes to the lyrics
To the base song? Doesnt that eff up the original song?
If the lyrics resemble the original then it will poke through more often
Changing the style too much does
Lyrics need to be different
Lyrics are different lol
This is the first song that I try to create with that much change in lyrical and rhythm change, so after having achieved a sequence, I was pretty sad to not be able to repeat that, even though the structure keeps the same, only the lyrics change. But I get that I put the system to its limits by not doing only one genre for the whole creation. Thanks for the general advise however!
Once you make the person and select it, it essentially makes a copy of the persona song that sits under the new song and influences it. It is finicky, but some of my best songs came from persona, but it did take a lot of credits
Biggest issue with persona is the speed tends to come out slow quite a bit
To make sure there are no misunderstandings, don't mess with the original
Biggest issue for me is no matter how much I change the lyrics, it constantly uses the same ones from the originating song
I even tried an ENTIRELY different song, no similar lyrics, and it still did it
I find that for whatever reasons it works better at some times than others. Sometimes it will spit out the same song over and over, if I do it twice and get the same song then I just leave it alone for a while
Thank you, I was afraid that could be the answer. I was inspired by the replace section which created similar tags on the title section (probably only for reference to the user). Then I will probably let my experiment rest for a while or until the next month starts and I receive more credits. 😄
https://suno.com/song/44081585-5cc7-4587-88ad-6875803e535e - no silence cover
https://suno.com/song/09460b79-1216-47f2-818c-ceaa6493be46 - silence cover. Notice the belting comparison and the lack of high end noise
Ah, I see. Might have to take a look into the persona system then. Thank you. 🙂
The little feature I keep suggesting that would help here is Anti-Crop (Glue/Fuse/Merge/Join) the ability to take any two Suno clips and combine them into 1 clip. This would let advanced users do some of the fancy song history manipulation to make things like this better.
That is what i have been dying to have
And to be able to crop replacements until the issue of it going too long is fixed
Oh yeah, I definitely was in need of a glue feature which crops out a specific section, not keeping it and removing everything around it. That would have been less credit intensive.
You have to play with it...I'll link my profile, look through it and play a regular song and then play one labeled persona mix, notice the quality difference.
https://suno.com/@sysmai - try those two
Lol nope
https://suno.com/song/2af34cf1-936c-4e96-b736-8b871b95c854
😭 yooo what is happening to my music frrr its still out here saying go to suno.com
Is that your persona?
It seems like the persona itself may be doing that
I know it is a big ask, but I think the Suno team needs to comb through the site and remove all of the songs generated with an unauthorized API, they are becoming a problem.
Are you the one that your songs disappeared while you were watching it happen?
It might be let me try without it, if its the persona thats kind of funny but its surreal like why is it happening
It is happening anytime an unauthorized API is used to create a song. There were many unsanctioned sites and even some apps that were making money off people using an unauthorized API. A protection got built in that makes it say that instead when it recognizes that it came from the unauthorized source and all generations have that mark in it.
In your case it seems like the persona you were using was made with just that.
Nah its still happening without the persona
Are you using an app or the site
Paste that link in here instead of a screenshot please. I want to check something
The exact link that you took a screenshot of
Also, are you US side or EU side?
I use ChatGPT to rewrite the lyrics phonetically.
Ok I'm going to do other things now, if anyone comes back and responds to me, please mention me so i get the notification.
I guess "trim" would be to cut off the ends and "splice" would be to "crop out" the middle. I see it labeled differently here and there.
You're right. hope that comes soon. I'd use it on many many songs I've done already.
I will say, I've been using "replace" as much as it gets results, to choose a 10 second piece , replace with a space bar click and run them. It cuts a second or two usually and blends the two ends, often, not always. Do a few times and it helps a little, instead of doing nothing. So that's always a positive.
I’m the one who accidentally deleted his account when the mobile app got set up and none of my credits were showing up or songs. I deleted the mobile account fearing it would corrupt my pro account. And it deleted everything
So at the most I want my old account restored if possible. At the least I’d like my credits back and the pro account status back because I paid for it so I would own the license to my songs and produced several.
Or a refund.
But I love using Suno, and want my songs back ✊🤘
Sent request to customer service 6 am or so yesterday, just gotta give em time to get to it I’m sure.
I have receipts for my payment and I know my old account name too
So if they need evidence, there it is.
Just contacted Suno yesterday. I've got a reply five minutes after, saying that my mail was received and will be answered as soon as possible. Today I got a reply (a bit more than 24 hours later). Just to give you an example of a current time frame. 🙂
Earlier this morning we'd talked about contacting support@suno.com so they could look into the matter -- have they responded yet?
What does suno mean by "A clip got stuck while generating" and how can I fix it
When did you send your first email to them? Just saw you'd reached out yesterday morning, still early in terms of a support response. From everything I've seen they usually get back to people within 48-72 hours, though being proactive with followups after that point can help if there were any oversights
Recently, we've seen it happen most frequently because something in the style prompt or lyrics is flagging the automoderation
https://help.suno.com/en/articles/3198209
We support a welcoming community for anyone to make music, but inappropriate content may get flagged before you can generate a song. At the moment, so...
Hm ok so it's too vulgar or sum?
I posted a link in the comment above yours that talks about some of the reasons a song could get blocked while generating
I've noticed that translated lyrics sometimes might be blocked as false positives -- perhaps something is lost in translation?
Yea, they are known for being Quick to reply. I'll let them take they have more experience
I've found it helpful to ask https://chat.com which of the lyrics might trigger an oversensitive automoderation system, and asking it to rewrite those sections -- sometimes it can help to figure out which words are being blocked -- sometimes it might seem innocent in our language, but it is considered a slur without us realizing 👀
Yeah, I got assaulted by Woodhouse for using a perfectly harmless German term in the German chat. He came over me with vengeance. 👀 Didn't know what hit me. 😂
👀
"oh shit"?
Woodhouse mag die Deutschen nicht, ich denke. 😛
(I think Woodhouse doesn't like the Germans)
Yeah, in Germany there' s an expression for being touched emotionally by something that translates to "have we-e we-e in one's eyes". Like that: 🥹 Hopefully, it's okay to explain it here. ^^
I have a better feeling thanks to all your replies.
btw -- I love the German language so much -- I was trying to describe someone not from Germany who would have programmed those Woodhouse rules, so I stacked together 2 appropriate sounding German-words that I knew like they were Lego, and it ended up being the exact term I was looking for.
Compound words are fun.
I didn’t see a response acknowledging I sent it to them right after though, lemme make sure I have the right customer service email
Support@suno.com right?
Yeah. There are some really long words. I'm not German but born and raised here. My first language was Spanish, then German. But it's my strongest language, I'd say. Although I can't write lyrics in German. Those are always either English or Spanish. 🙂
Yeah. Quick channel search confirms it
Ok I can keep sitting tight.
Spanish and German are such dynamic languages.
I enjoyed learning Spanish and haven’t tried to learn German but I’m really interested in the annunciation and linguistic rules of Deutsch
My problem is about Credits so a I wrote to billing@suno.com.
I don’t want to duplicate work, so I’ll stay in my lane with support.
Time will tell.
They go to the same place in the long run afaik 🙏 Just helps for organization purposes.
Thanks again everyone who has responded to me.
If this is fixed I’ll release my lofi album or something sooner
Suno made me realize as a songwriter who lacks the capacity to execute my songs that I have hope.
So I appreciate the tool and the community around it.
I completely empathize with you in this situation -- I saw the comment earlier where you showed a lot of tenacity and seemed committed to recreate any content, which seemed like a remarkable display of determination in the face of something that would leave most people raging -- I absolutely understand the urge to rage, but absolutely was caught off guard and impressed by the gumption.
Whatever happens, I'm genuinely excited to hear your next release 😄
Is Suno down - from a creation, extension, and editing perspective?
The page loads, but I just get the spinning wheels of insanity whenever I try to do anything. It's been that way since Sunday.
Not a system-wide issue -- some basic troubleshooting would be to try visiting Suno.com from a private browser window without any extensions, clearing the browser history & cached data, etc, installing a new browser, etc -- essentially just trying to make your computer act like it's the first time it's seeing Suno.com.
imho -- easiest method is just checking if the same issue happens on a diferent device (friends computer, mobile browser if you usually use desktop, etc)
If the issue persists, you'd need to contact support@suno.com so they could look into the matter for you 💪
Does anyone know how long it takes for Suno support to get back to you?
no 😭
Yeah, okay, I can see what you mean. It sounds different, more natural in a way.
Has it been more than 72 hours since you've contacted them?
afaik, the email support is active from Monday - Friday, <t:1731463200:t> - <t:1731448800:t> (your local time), and though I don't think they've outright stated their turnaround times, it usually seems to fall within a few business days -- any delays after that would seem reasonable to send a followup email
Right. You will need to test it for yourself. You will notice much more than just Vocals. It seems like it uses more instruments and harmonies hit more natural.
Thanks for showing by your examples!
Np
They responded and said they can’t help me.
How terrible.
It wasn’t clear that the mobile account was linked to my main account.
The deletion of my main account was inadvertent.
Even in the app it showed no distinct connection, had me set up a new name etc.
But that’s the response.
I asked if they could give me pro again but it sounds like I might be hosed.
I’ll update if I get a response
¯_(ツ)_/¯
I think you should make it clear that your account was not deleted but likely that the account ID linked/associated with your email from your original account was overwritten when you used the app which forcefully linked it as a new account instead of logging in to your old account
(Assuming that was the case because you are not the first to experience that issue)
Oh, my bad 🫂
It wasn’t clear my account was linked
In the app I had none of my library or credits.
I deleted the account the app spun up in an effort to preserve those.
It did the opposite.
It’s a bad experience.
I’ll have to spend another hundred dollars to get the pro status back.
You can see it at the end of this clip — btw, I’m TERRIFIED of this screen on iOS now.
And all the songs I made are gone.
But yeah, definitely ask them about getting credit back or ask if there is anything else that can be done. Provide your original receipt and hope they provide you the paid credits on your new account
But that is absolute bull nonetheless, ugh. Your sentiments were on track though
(this is the reason why many companies have like a 7 day leeway before deletion which gets canceled if you login to your account within that duration).
I told them I had all that, but I have a feeling they got my money and don’t care.
Sorry that might have been a bit less than optimistic
Trying to maintain optimism.
Honestly, I would chalk it up with the person likely being tired after a long day and near the end of the shift.
Support is usually great about crediting these things back so long as the account was not doing something illegal, so definitely try asking again - or maybe ask billing?
I suppose you can possibly talk to Mike B, the Community Manager, about his opinion of what can be done. Unfortunately, he is out of the office for like a week or so
This was my response.
If that doesn’t make my case then I am left without any options.
Must be the universe speaking to me.
good evening folks,
first to say really enjoy suno, today traing a bit with pictures, but cant extend them.
I’m running into a frustrating issue with Suno AI and could use some help. Here’s what’s going on: I’ve been creating songs from images, and while the initial 30-second tracks are great, I’m having trouble extending them. When I click the button to optimize the lyrics for a longer version, I keep getting an error message: “An error occurred. A system error occurred.” 😩 Interestingly, this only happens with songs generated from images. Regularly created songs extend just fine. Here’s what I’ve tried:
(pro user)
Refreshing the page
Logging out and back in
Switching browsers (firefox and chrome)
Has anyone else experienced this? Is it a known bug with the image-to-song feature? Any tips or workarounds would be greatly appreciated! Thanks in advance for your help!
It's rectangular?
Ahh. Yea, as long as a voice isn't detected, it'll persona. I can't find a workaround for when a voice is detected, tho.
Guys, how do you manage to correct tempo drift? Any "auto" tempo correction software or something that helps do that automatically?
It depends on if you need speed and automation or precision and natural-sounding. DAWs like Cubase, Ableton, FL Studio have tempo correction built-in that can automate the tempo. If you need a more granular accuracy and natural-sounding Melodyne is the way to go even though it does not have automatic adjustments.
Has anyone noticed quality and prompt adherence issues in the past serval hours?
Note: I'm using V.3
Correction: V3.5
There are a couple ways that can help control the drift. One is change the Tempo even if you don't change the Tempo. You change the beat or give it a break this can get AI to check its timing and reset the beat.
This will not remove the drift, but it will help control it. The way that works best costs more credits but you extend from the point of where the drift starts.
For me the past almost month now has one song come out not adhering at all and the other doing a fairly decent job. Still one should not be the byproduct of the crap not gone into the second song. However, if it has been doing it to both for you have you been working in the same song?
hi there can someone pls help with this problem that when i try to change 1word it only change the first one that i do but when i try to change the rest of them that do not work it say it the same as the original word and i was the replace section to change the word is there a better way?
in terms of cover? its extremely difficult to replace words using cover you may get it right once in like 30 generations
Excuse me, is there a model of China opera style and opera music in it, and how can we accurately trigger the China opera singing method of opera?
Have you tried using style tags like "traditional Chinese opera", "Mandarin opera", etc.
You may have better luck asking people in https://discord.com/channels/1069381916492562582/1154780167047217233
Of course, I tried a lot, but there was no opera singing.
Not sure if this helps:
Classical Chinese Opera song. Listen and make your own with Suno.
Classical Chinese Opera song. Listen and make your own with Suno.
I couldn't do any better... @fallow pollen it came out a Chinese folk song. https://suno.com/song/98d09bc8-fc8f-4ee3-be75-c6b9780bc2a9
Traditional classic Peking opera, dynasty, folk, Jinghu, Erhu, Pipa, Dizi, Luo, Chinese folk, non-metric song. Listen and make your own with Suno.
I don't know anything about chinese opera but sounds pretty good to me.
Sounds good to me, too.
Was it you that shares the love of BSG when Kara Remembers?
I get confused. So many conversations...lol.
It was someone in here
not sure what bsg is so.. probably not me 😉
Yea I didn't think so. I just remember it was someone that is in here a decent amount...lol. either way I got a song I uploaded that I made a long time ago, The beginning is supposed to sound like the part of Battlestar Galactica where Kara Remembers. So far it hasn't gotten kicked back, I'd hope not since it is a derivative of a derivative.
aah. bsg. i should of put that together.
It is one of those shows that didn't get its proper due because of the original that was terrible. I made fun of my wife when we first met for watching it. Even though I am a Star Wars nut and I know a decent amount about Star Trek as well. Eventually we watched together. Best series I have ever seen and now I have a BLU-RAY box set...lol.
You have to know the way the story played out to appreciate the song. Bear McCreary the genius Composer made a version of All Along the Watchtower. Which was based on Hendrix's version of Bob Dylan's original. Regardless, it brought the show to the most epic suspense moment so perfect it was insane.
i love when I tell suno "grunge" twice, and what I get isn't grunge..
not really complaining about what I got though, its just funny https://suno.com/song/2a56c625-a371-45d7-b8fd-d36f587dda17
Yea that is closer to Sam Hunt than grunge and that is a looong way away...lol
lol i know, its so strange
the closer i try to get to nutshell the further i get from it
You try emo and melancholic? I find Suno doesn't put things in the right place so those descriptions would tell me something like Smashing Pumpkins, but I got a Smashing Pumpkins sounding song from a totally different prompt that should not have produced that style...lol
every time i try emo i get a vocals I don't like, to high picked.
well when I think grunge, when I go for grunge, I want things like alice in chains nutshell (But thats not a good example because lance does sing high in that.) but think like pearl jam jeremy, collective soul, stp, and yes, smashing pumpkins too
and of course, nirvana, but not trying to mimic nirvana really
temple of the dog - hunger strike is a good one too
Last two weeks have been like this
Well temple of the dog is mashing up things you named already. Missing some Soundgarden...lol
yeah, that too
So man in the box with the off key harmonies or close to Jeremy I imagine we aren't going to call Yellow Ledbetter grunge...lol
but i mentioned temple of the dog and specificly hunger strike because of how its sung, and how the instruments play so far behind the vocals in the start
Yea nice groove
Let me check the list...lol
Grunge Americana, Southern Rock, Garage Grunge, Alternative Rock
Try that and list the specific instruments in lower case
wow okay, I found a song on suno today that had close to what I want, couldn't duplicate it though, it seems to be gone now, heres the tags the song used, I copied though down thankfully
blues slow bass-heavy,Heavy Grunge,Alice in chain, aic,slow,dark,melodic,layne style vocals, male vocals,professional studio recording
and Ill try those real quick
You may want to add dark and slow and all that since Americana tends to pick the beat up in my experience. Never tried it with grunge. I just took the entire list from the explore page and fed it into chatgpt so i can get quick starters for songs I want to sound like a song. I give it the song and it spits out stuff from that list. Even though they don't always match but at least its a start
seems more country then grunge when I try to slow it down
Take americana out and put Melancholy in front
all though, the song I am working on does kind of sound country
Melancholy Grunge, .....
yeah, changing grunge americana to melancholy grunge was a huge difference
no longer sounds country
I'm playing now, I want to head in the stp or Alice in chained direction.
You know what I mean.,,autocorrect....blah
lol
not the best song to try this with, but this is grunge... https://suno.com/song/fdfada5b-5258-4e50-8c30-37fded8f98fd
Melancholy Grunge , Southern Rock, Garage Grunge, Alternative Rock, dark, slow, song. Listen and make your own with Suno.
This is getting closer https://suno.com/song/b4bdb2ac-da8c-4b5e-934b-558d1d70b533
I just got suno to tell a story of a man in a box...lol
Lol not quite man in the box
Well no, but it was the direction of the style.
The lyrics weren't even close...lol.
I think one of the hardest things for me, is actually writing like that, grunge lyrics are a bit different, take nutshell as an example, beautiful and powerfull song, but lyricly its pretty simple
So well, I created a persona who I'd like to sing a song that I have with a lesser/weaker singer. Unfortunately the cover option doesn't let me choose a persona to sing the song. This is a feature I am truly missing now that I can't combine singers I like with my older songs, at least not with my Android. Or does this work on Desktop and/or in some other way?
As far as i know you cant persona right now with uploads
Ok take this one on https://suno.com/song/10245115-3e19-47b3-8a93-96a3214432bc you can't deny that gets waaay closer
I don't think it is active for uploads even if it shows it. I have seen both
It shows enabled for me now, but yesterday it didn't show up at all
Woa woa woa.. the vocals in that.. what is that song.
Yup
anyone else having issues editing song details tonight?
Cough cough
Im trying to pin point that song
Nope just finished one actually
I was having issues yesterday, kept saying issue with image when i wasnt changing image
exactly my issue
Oh the band it sounds like? I'd say some pearl jam mixed with sound garden. I would like to hear that voice grit, I'm gonna make a persona off it. It has to scratch at highs
Its going to bother me till i figure that song, that sounds like a very specific song
Let's see what Google sees it as maybe it is close enough
Well, I wouldn't use with uploads now: I have two Suno songs in which from one I'd take the singer, "persona", to sing on a cover on another Suno song that would need the voice of my persona. But that isn't seemingly possible, unfortunately...
Which part isn't possible, the upload to persona or the persona in general
Its not even grunge im thinking of.. guns and roses maybe?
It has some bon Jovi in it just not the same vocals
Yea guns and roses for sure
I'm definitely making the persona. Now I just need James Hetfield
https://suno.com/song/7aeebba2-e06f-4584-a36a-90abfeb83548 well this one is loaded with shinedown sound not Vocals so much, probably a more eloquent Kurt
I got close to the Alice and chains harmony, I'll put it in between Alice and chains and nirvana
Put "aic" as a tag, im curious
K
Ive seen that done a few times, just not sure if its actualy effecting anything, also seen "alice in chain" no s be accepted
What the hell is that
Oh
It definitely dragged it that way
Almost some seether in that one
Oh yeah, i can hear it
It only works with my persona, without it then it goes basic
Aight. Sleep time. Holla
Night
Hi everyone.
does work the cover or persona option with you?
Since days, I get the circle of death and then Suno chancelled the creation process.
i can't create cover it says you are not subscribed but i am and i have 2800 credit i think its bug i refreshed the page but still the same
(For me) a new problem. Songs end before 4.00 minutes in the middle of the song. When I extendm they restart and still the end does not gets added.
yesterday I had the same issue
for the extended section, within lyrics, add [outro] and maybe 2 to 6 lines of lyrics underneath
What is the little red dot mean
u didnt clicked on it or its new
If the song ends prematurely you can either extend it or use to crap function to crop out the empty space and then publish it
No I definitely clicked on it and it's definitely not new it's been liked a few times
i just generated one and after i clicked on it the dot dissapeard. idk what could be then
I'll click on it two or three times and a red dot persists
Ok. I decided to find elements of prompts which would give that Peking Opera. Up to 1:15 it sounds relevant. After which, it continued a modern pop style. But the idea is there -- Suno may simply not be familiar with Peking Opera / Beijing Opera, so other prompts are needed to get that non-metric, high pitched, sounds. So , again, this is not accurate. Only interesting enough to make it public. 😀. (First run gen of 4 pages of tests prompts- no edits)
Anyone faced this: suno singing is clipping little annoying?
And only fix isolating voice and back track, then remix? 😦
3rd partu sof?
It isn't possible to have my persona to sing – meaning singing on a cover – on the song that I would like it to sing on.
How can I send a ticket to ask them about some problems
Anyone faced (a bit) clipped singing voice on suno outputs?
I'm looking for assistance from the techies of technical discussion. I'm trying to be really fancy and transition a main vocal into an background vocal / instrumental track but I can't get the transition consistent, nor am I capable of keeping the original vocal as a background vocal / instrumental whenever I continue the extensions. here's what I've currently got: https://suno.com/song/ebd85473-a6bd-4e07-820a-06cbc607ad13
hello, is the feedback link for the replace section still available? i have some issues to list
The small red dot beside your outputs indicates It's a new generation that you have not listened to yet.
Is it technically possible to set the upload of vocals to 3-4 minutes? Is it possible that Suno adopts my voice 1:1 and doesn't always change it? Is it possible that with Remix Suno takes my voice and doesn't just change it to AI? Is it planned or possible to create 1-5 voice profiles of yourself or other singers?
I was playing with Personas. Created a persona out of song. Used it to sing updated lyrics. But it is singing old lyrics instead 😑 I was under impression Personas should only capture music styles and vocal style?
Yup -- #📢┃announcements message
can I upload 3-4 minutes?
Nope, 1m for free users, 2m for paid
can Suno adopt my voice 1:1 and not change it?
Upload allows Suno to extend from a clip, but this is a real-time process that studies the clip you provided & tries to continue from it -- if you've provided a strong enough musical context, it will be able to continue from that clip.
This is not the same as Vocal Cloning.
possible to create voice profiles?
Not of yourself, since this feature isn't available with audio uploads, but if you enjoy a vocalist from one of your Suno tracks, paid users can "Get Persona" from that clip to re-use that same instrumental / vocal style on other tracks 💪
Okay thanks. I‘m a paid user. If I have create a Song with my Vocals. Can I use this Song for persona and then I have my own Voice? You know what I mean?
Did your vocals make it onto Suno by-way-of Audio Upload? If so, you can't make a persona of it.
ah okay thanks.
V4 sounds great. When will the Patchnotes for V4 come?
is the crop feature only made to cut what we want to keep? i was hoping i could use it to cut some seconds of a very good generation i managed to have, but with a very few unwanted things
Hopefully soon! Make sure you're subscribed to #🚀┃updates and #📢┃announcements so you're the first of your friends to know!
Yes -- with Crop you highlight the sections you would like turned into a new song -- if there are unwanted segments, you may need to use the "Replace Section" feature, or edit the song externally (may not be too difficult to edit the audio, now that in-browser audio-workspaces and stem-splitting is available for free 🙂 )
bruh...
still waiting for my prize after winning in the "Contest feat. Flosstradamus" .. No answear on support mail.. So whats going on?
Hey everyone! I have another slight silly questions.
Is it possible to get audio hallucinations on purpose? I would like to get something "spacey".
I just edited and changed the words in my lyrics, but 'create' keeps creating the old lyrics. I tried creating a whole different track to clear it, but it seems to be clinging on to the old lyrics regardless. How can I clear it down and force it it use new lyrics? (I
I think #🌴┃creator-programs is the channel for the cash-prize contests -- they'd likely be able to suggest who to get in touch with 🙏
Absolutely! You'll see lots of weird symbols in people's lyrics -- oftentimes Suno will see that something has been provided, but not know how to verbalize it, so it might come out weird.
I'll never get tired of this hallucinatory song from furrybeans https://suno.com/song/197cffc1-da6e-4fb6-afe7-362b62f0c84b
Ah okay! Thank you! :3
Help! Why do one of my songs have a GREEN "v3"?
^- answered this in general chat, it's a visual bug
yo, probably someone asked somewhere but I can't find any solution: I wanna create an instrumental cover for a song. How do? if I click cover and choose "Instrumental", it still has the singer singing
It's not an uncommon "hallucination" for Suno to imagine lyrics should fit somewhere that you don't have lyrics presented, even with the "instrumental" toggle turned on.
Direct answer is that while this kinda stuff can happen regardless with generative AI, the biggest impact on what Suno spits out is what we use as input. Roundabout answer is that you could use extensions to reroll the sections of a song that you like before the lyrics come in -- keeping only the sections you love & getting rid of anything you don't like while choosing the extension points.
Couldn't find any info about languages. Can I make a song in Japanese, for instance?
Yes as demonstrated by the homepage when you change from English Trending to Japanese. Or look up the style tag Japanese https://suno.com/style/Japanese
Just enter your lyrics as Kanji. Katakana should work as well as Romaji.
For certain stylistic outputs, use the keywords Japanese in the style prompt or associated Japanese styles (e.g. J-Pop, J-Rock)
Yo guys, I got a question and I'm maybe out of topic but :
I made some songs for a "viking album", and sometimes I replaced sections and made custom edits on my laptop on editing softwares directly because I liked only some parts of the edits (idk if it's clear) but I would like to know if there was a way to reupload my final version on Suno after all to make a playlist to share it to public ? I know we can actually only upload 2 minutes songs.
You answered the question at the end there -- we can only upload 2m clips to Suno.
Alright, too bad then, I won't be able to remaster my files for the V4 except by looking for every pieces I putted together lol, too sad but ty
Externally editing a clip to reupload is a great way to improve the quality of the extensions, but anything beyond the 2m mark is lost, as far as Suno is concerned -- you can ask Suno to regenerate the closing segments, or the finished song will have to "live outside"
bet ty
If it counts for anything, v4 can remaster the songs with next-to-no touchups and does a decent job 👀
Yes I hope so, I will dig deeper once it's out and try to reassemble everything :3
I've had a lot of fun going back through songs that I thought were more or less unsalvageable & letting the Remaster feature work some magic on it -- suddenly they're actually worth releasing 👀
So excited for people to get their hands on it 😄
Yeeeeaaaa, you're hyping me so much but i'm kind of pissed off to hear about the v4 once I pasted like 2 weeks in a row digging the software and make outside edits and try to mastering them from other softs and everything ... so unfair lol
The work of an audio engineer never rests -- and the time spent working on earlier tracks isn't lost if you learned ANYTHING during that time
Hell, if anything now you'll have 3 different versions of the song to release if it really resonates with you -- the "original", YOUR remaster, and Suno's remaster
hahaha didn't think about that
Follow #📢┃announcements and #🚀┃updates so you'll learn at the same time as us alpha testers 😅
and do you know if with the v4 we could upload songs longer than 2 minutes ?
There hasn't been any changelogs announced, so anything I'm commenting on would just be the early-access state of the model. ATM, everything works the same internally, just the music sounds better
thx
I still have my problem. I used a particular verse with certain phrasing. When I change the wording, each new creation still creates the OLD wrong wording. Even is I create a whole new song, a surprise me and let it create. Then paste that (new, correctly worded) bit in, and hit Create, it creates it with the old WRONG wording.
This song seems higher quality to me, but both are really good.
Still waiting on a response to my second appeal for customer service to restore me.
Hi, why i cant see the generate channel on my discord?
That's been phased out a year ago.
Q: Anyone have any success with excluding pianos and especially arpeggiated pianos? I add -pianos and -arpeggios and write [Interlude WITH NO &^%&* PIANOS] and 3.5 is like, Hey buddy, no problem. Enjoy some of these pianos i've delightfully arpeggiated.
Hello friends, not sure if this was previously addressed, few questions here: 1) how exactly will V4 "remaster" songs? 2) Will it have more options on splitting stems then currently? 3) Is there a way to prompt the female vocal to sing in low to mid-range in a pop song, without involving "personas"? Thanks
looks like I have an opposite problem, I would prompt "arpegiated piano" and it's like "hey enjoy piano with whole chords" 😂
Have you tried using the negative prompt section? Suno doesn't understand "negative prompting" in the lyrics/style section by default, so saying "no drums" is literally the same as asking for drums -- this is what the "Exclude Styles" section is intended for.
I wish they would add an include section for things to include. For genres that don't usually include piano for instance, Suno's include section would figure out a way to fit it in.
Is it a good idea to remaster V4 tracks when fully complete? Or is it specifically designed for older tracks produced with older models?
-
Using image gen as comparison, if "cover" was like using reference material, "remaster" is closer to img-2-img -- 2m30 in? 2m30 out.
-
No updates regarding stem splitting, but they called it v1 of stem splitting when they announced it in #📢┃announcements or #🚀┃updates, so with that naming convention I'd wager it's only a matter of time before that gets a some love.
-
There MIGHT be a way to help suggest a lower vocal register for a female vocalist, but it's more a matter of experimentation within the prompts & styles you're using opposed to knowng "the right way" to prompt it. Arguably, the easiest way would be to use a persona that captures that vocal range already, but that requires HAVING that persona to begin with 😅
"Style of Music" IS the "Include" section -- but it's possible to overload it to the point that it doesn't respect the stuff you're asking for
I've been remastering all sorts of stuff -- remastering remasters to see what minor changes it makes to the melody (not a viable method of changing lyrics, from what I've seen)
anyone else notice this while making generations that it sometimes switches off of v3.5 and onto v3
I do. I do, so so hard. I tried minuses too earlier, and found how badly that worked. But yeah, here are the prompts for this (the minuses are in the exclude section) https://suno.com/song/e37afd15-75bc-4b86-af6e-1ace12ee6360
atmospheric drone, ethereal, hypnotic, dark energy, slow, dysphoric, foreboding, granular song. Listen and make your own with Suno.
Have you played around with v4 covers, though? 👀
It's absolutely been a breath of fresh air, but "going back under" with covers is blowing my mind just the same as it did when it first came out.
Some features are only available on v3.5 -- covers, personas, etc
did anyone feel like the quality has been higher for the past 3 days?
i use the same prompts alot and feel like something "changed" and im getting higher quality output
okay -- this is a cute change 🙂
is there an idea when this will be available?
It was something that WAS available that they disabled when they were made aware of it -- I wouldn't imagine it would be reversed any time in the immediate future.
That said, there's a typeform survey in the Personas #📢┃announcements or #🚀┃updates page if you wanted to give feedback about Personas to let them know how much you want that ability.
@charred merlin Do I understand that alpha users are already playing with v4? 🤩
I guess you can't say much because of some NDAs
That clip was generated on the fly with v4 😛
#🗣️┃chill-chat message trying to get over my reticence to share my music publicly, in order to advertise v4
Man, I'd be cranking out lofi tracks left an right rn if I had my credentials
This is so inopportune, I have a small break between contracted gigs, all my excess creative energy was going to go into this now botched project. 😐
The response I got was so... detached and clinical.
"Sorry to hear that (assigned random number here), unfortunately, (copy/pasted script here)."
I'm just hoping someone will care.
Because someone has to care.
Or else I'm hosed.
Why would the mobile app not show any clear connection to my main app, and make it appear that all my balances were zeroed out? That's why I deleted it, to save my stuff.
Seems asinine.
From my perspective I feel as if I was stolen from.
Anyway, other people have probably got better stuff to worry about, nevermind me.
Getting flashbacks of applecare -- positioning potential data loss is like a soft-skill in it's own.
I'd hope that the email followup to their response would act the same, but I wonder if asking for the case to be escalated could help 🤷♂️
I'll try to escalate it as much as possible, but I'll be going for an alternative service and kissing this place goodbye if I am told nothing can be done enough times.
You guys really buried those gold nuggets under the off-topic channel 😛 v4 sounds like a big improvement. Better harmonies, more rich sound - especially voices and acoustic instruments. I understand if they are reluctant to release it and just want to polish it while they can. But it's all about understanding and taking it for what it is and not what people want it to be. Hissing or not, it sounds much better already and I can't wait!
in trash using browser:
Trying to delete stuff I'm not needing, but can't with these few song and keep getting the error to delete the persona. Moved to trash, but it's not deleted. Bug?
If you've made a persona from one of those clips, you need to delete that persona before you're allowed to remove the clip.
well....that stinks....They'll just have to live in trash.
On the suno.com/persona page, you can click the 3-dot interaction menu to send the persona to the trash
I know it doesn't help, but I do feel your pain. A lot of us do. When both iOS and Android decided to move to have apps sign in through iOS or Google instead of making new accounts they were overwriting original accounts. The overwrite. The most absolute way of getting rid of something that has no backup. I lost years of work on so many apps. It was insane. All the money spent gone, all the notes and writing and everything gone. The worst part was that they forced the new sign in without notice for most people. You couldn't continue to use the app without the automatic attempt at having you sign in using the SSO-ish account, talk about a PR and everyone nightmare.
checking once more, if anyone is able to help with this
Have you tried (...insert background vocal lines...) with the ( )'s and ...'s?
Their isn't a true way to ensure they will become backup Vocals just inference
It may be a second set of vocals like a scream or whisper, or it may become backup. What I actually found is that using things like cover and persona, do it more often automatically in specific parts.
(...) tends to bring the second vocal whether it's backup or whatever, and {....} tends to bring effects or a third set of vocals depending on the style and lyric structure. At the same time, it also can give Suno permission to skip those words at times if it would mess up the flow. So make sure you have a solid scheme.
Like always you can pull it out and run it through a DAW and drop the volume then layer it and automate the vocals
After listening to what you are doing i can safely say that what you are attempting is not an easy undertaking, you may have to either figure out something new, maybe there is someone out there that figured it out, there is always going to be the accidental ones that just work because it just happened to work on that Instrumental. But getting even vocals to overlap can be hit or miss. To bring an Instrumental on top of vocals seems kind of against the nature of Suno. The other way around, no problem...lol
Suno as it is now
Best bet would be to use an outside tool and upload the audio, it will be much more likely to work if you give it a decent sized example first.
Do you have any examples you can share from Spotify or Youtube, etc?
I do not, sorry.
Maybe you can guide in the lyrics phase more then the style. Things like this might work
[chorus]
[echoed vocal fragments weave through instruments]
[verse]
[distant vocals]
[bridge]
[vocal elements dissolve into ambient textures]
What prompts do you use to create vocal harmonies?
Is this a question you are asking the channel or are you asking to help the current question above?
Just trying to understand what you are trying to achieve...
Something like this maybe?
https://suno.com/song/d9cad2e7-8d70-4af0-97b3-98231a390d0c
You know you can remove the [] in style and have more characters for descriptions right? I assume you know this since you don't have the tags around all of them
From experience, having around tags has better outcomes 😉
Not in my experience, but no one really knows. I can tell you for sure that capital first letters weight the words heavier in style.
I think it also depends on the genres you are trying to work on. For my case for example turkish-instruments work better then [turkish instruments]
But [slow lyrics] is better then slow-lyrics 🤷♀️
I'd say this is more of an echo kinda thing, I'm trying to keep the echo going as a continuous instrumental in a way
Hey, there is nothing i can say to prove otherwise, I do find that dashes in general do not work that well except for a few styles. What I mean is main style is caps: Epic Orchestra, Alternative, violin, cello, guitar, dulcimer
Oh, in that case try to keep up with words in paranthesis like I did yesterday(day)(day)(day)
If you add "enough" this breaks Suno, and you'll start to hear more repeats throughout the generated music.
Add [Echo] after that
For instrumentals, an example from me turkish melodies, turkish melodies, mezzo-soprano, choir, epic symphonic, eastern symphonic, emotional, melodic, rythmic 😄
...1000s of credits on these 😛
Sorry I meant: Alternative Rock, Mezzo-Soprano Female Vocals, distorted guitar....
Does anyone have any ideas on how to get a truly acoustic rock song? I'm finding that drums are added in on most generations, even if I put drums as a negative genre. I was wanting to go for just guitar and vocals
Yup
Experimental tends to give a minimal version and minimalistic does as well, you can also put acoustic and unplugged in style
Try using style names that are used with acoustic versions even if that style isn't what you are using. Do it later in lowercase letters further down the prompt
On top of minimalistic, you probably need the *-driven style here. Like acoustic rock, guitar-driven, vocal-driven etc... Keep the drums in negative, maybe add drum-set as well... Suno is mostly experimentation.
Actually that is better. Acoustic-Driven
And core, but never tried it with a sound style. But Genre core attaches to almost everything
interseting. I'll give some of thes a shot
Do Acoustic-Driven first it should help, and put the instrument in lower case, acoustic guitar
You can add the tag in lyrics as well
Does anyone what I can do to fix the error: "A clip got stuck generating, No credits were taken from your account and you can try again".
I get this when trying to generate a song from Custom, with my own lyrics. I've been getting it for about an hour or 2 now on 3 different devices, 2 different wifi connections and through data... I've deleted my browser stuff like cache etc,
Is that error caused by something I'm doing or is this a bug...?
Almost every time someone has had that issue it was the lyrics content causing the issue
Yep. There is a word or sentence Suno is not liking...
Profanity, things that talk about murder stuff like that
Ah damn, I'm having lyrics for a game related song... Might be that
I'm not saying you can't use it, just things to test removing
It can't mention copyright things
One thing I do mostly is put the lyrics in ChatGPT, and ask, "if I shared this with Suno what can be a problem with moderation?"... ChatGPT knows what Suno is.
Oh thats smart
Suno uses nanoGPT so including that can help with errors in the way things are worded. Also use the web search to get your results since Suno is ever changing.
T his work well in some instances. Just pepper some ( vocals cues ) of the words or sounds you want among and between the lines of well metered lyrics. Suno doesn't like going off-meter so it has, with me, sung the parenthetical in the background. But it has also just skipped some words or a line in order to fit the (vocal cue) in.
Not at all perfect. It's a crap shoot, but for me, fun to see what Suno will do.
{...} I find is more likely to be skipped than (...), but it is funny that it is always still inference. Since you can put [....] around lyrics as long as there are enough lyrics in them to be inferred. I think you were part of that conversation.
Suno uses nanoGPT
Can I ask where you're citing this from? I'd be fascinated to understand more about how Suno works internally so I can give better advice 🙂
GitHub, We spoke about this. Bark is not an LLM, Suno uses nanoGPT for prompts on a simplifed level since it isn't based on a huge dataset like Large Language Models, it is good to know the limitations of nanoGPT
we spoke about this
I'd beg to differ -- we've never spoken about nanoGPT.
Are you saying Bark uses nanoGPT? Or are you talking about a Suno github page that you're getting this information from?
For the record -- I'm not saying you're wrong, I'm genuinely asking to be provided with the citation so I can read up for myself
Ahh -- I see, the Bark github -- saying for certain that Suno uses something because Bark used it when they were being open-sourced might not be entirely reflective of the current state of Suno -- it MIGHT BE, but that's not public knowledge.
no, we didn't talk about nanoGPT but we did have the discussion about Bark not being an LLM that is when Jonathan I believe came in and confirmed that it wasn't an LLM but it is like an LLM in a way. No I will have to find the place I found the nanoGPT part it is further in. It is its own component
looking
I DM'd you a couple days ago after my old professor that teaches all this stuff replied. I was going to share what she had to say about it, but I don't want to quote her so I could give you this information personally.
where is this from?
Well color me jealous...so what secret did you use to test v4
I late-edited my post -- I see that it's on the Bark github page, but it may not be accurate to say that since Bark used nanoGPT back when it was open-sourced that it means for certain Suno uses it on the closed-source side of things.
They may use it, but that's not public information, so it's slightly misleading to infer with certainty how Suno works under the hood -- I understand that you're studying this, but I keep a degree of "who knows" until they actually show under the hood.
Would that not mean that unless told otherwise you should go with what knowledge you have available. I can say it like this. Once I started talking to Suno in a nanoGPT way my stuff has improved quite a bit. So, they either still use it or something like it.
Sounds good
To be clear I do realize that I said since Suno uses nanoGPT, and I am sorry to all of Earth for saying something that can't be proven or disproven by me. I should have said since Suno's earlier work with Bark included nanoGPT it could be a good idea to include that in a way to reform your prompt. I really don't see how it could hurt.
Hopefully to my credit, I spend too much time teaching chatgpt how to write a song. But in the end I get good ideas.
And the occassional killer hook
It's nice how those instructions can create lyrics that sound pretty good when you throw them in Suno. Nice suggestion. Thanks.
Ver 4.0? You got that? I bet it sounds killer. I dont d/l audio files on my phone but if it's a suno link I'll give a listen. By the title it sounds like a Country Song?
Sounds great. I'm gonna use that intro type some time. 😁
Just for edification, when I clicked, it wanted to set it up a file on my phone. So.....
But great you got a v4 song under your belt and thanks for the chatgpt suggestion.
💪 i can do with 3.5 or I'll have to wait.
Can anyone remaster this song to v4? https://suno.com/song/24e2a5a9-655b-461d-980d-449c1b0f808e
Female voice, Christmas happy hardcore, Christmas Goa trance, Christmas saxophone, adult contemporary, groove, one take song. Listen and make your own with Suno.
Okay so I have been getting this specific error whenever I try to make a cover lately, anyone know why?
Recently, we've seen it happen most frequently because something in the style prompt or lyrics is flagging the automoderation
https://help.suno.com/en/articles/3198209
We support a welcoming community for anyone to make music, but inappropriate content may get flagged before you can generate a song. At the moment, so...
Oh okay thank you
Can you edit the prompts in a persona or do you just have to save a new one if you've made tweaks after the fact?
I'm sorry, but I'M SO ANGRY! What the hell, I spend all my credits just to get bugged(?) songs? I tried to create song with male voice, but... ALL THE SONGS GOT GIRL VOICE?😡 Whyyyyyy???😭😭😭
cause girls just wanna have fun
Here. I tried to make it clear, but... but??? Male vocal, male voice, boy group, what else I need to write, so suno can understand me?😭😭😭
For one of my songs I literally had to spam suno with (female voice) (female voice) many times to finally get it to do what I wanted.
Ohhh....😵💫 this is crazy. I think Suno's developers should fix it, or do something about it! My anger↗️↗️↗️
I don't think that I've ever actually run into this problem, tho I know there are plenty of users who have. The only times I seem to run into gender issues is when trying to get a duet.
Can you share a link to the track that you're struggling with? Maybe someone can help.
It autopopulates the style section, but you should be able to change it after selecting the persona & before generating the song -- but it might not respond to styles that vary drastically from the original style prompting.
Sure! Here it is.... I tried different genres, but it didn't solve my problem.
https://suno.com/song/89779ff4-71af-4228-b138-f819012468de
https://suno.com/song/b0725090-da76-493d-9001-9bf31a75c12d
https://suno.com/song/1714e772-1ead-42e5-9294-d6c4b4f85834
Soul, r&b, k-pop, male vocal, man voice, boy group song. Listen and make your own with Suno.
K-pop, clear male vocal, soul, r&b, trap, dreamwave, fantasy pop, moderato, 3/4 waltz rhythm, 90 bmp song. Listen and make your own with Suno.
K-pop, male vocal, soul, r&b, trap, dreamwave, fantasy pop song. Listen and make your own with Suno.
'seems like it's either k-pop or just the korean lyrics in general that's pushing Suno towards female vocals. You can increase your chances of getting male singers if you make that the first item in the style field... and maybe even reinforce it later in the style field as well as with tags within the lyrics themselves. https://suno.com/song/47e08f43-be34-40f5-ba07-b85b7c3e3592
Male Vocals, Soul, R&B, K-Pop song. Listen and make your own with Suno.
Have you tried using the "Exclude Style" / "Negative Prompt" section to specify what you don't want to hear?
WOAAAHHH howwww??? (Song that you got is good, I like it)
Well... I tried it too! Here. No "k-pop" at first place on genres. I think, this is bug. Or my account is cursed.😵💫
I don't have exclude style, I'm on free account (sorry, I'm poor).😥
What is negative prompt? I saw someone putting "-style" in genres, this is it?
the Minus symbol is a visual representation, of whatever is used in the Exclude Styles section -- unfortunately adding the symbol before prompts in the Style Prompt section doesn't work the same 😦
So I tested it and got one song that was batty and another one that sounded dead on the original persona just tweaked a little. Gonna play with it a little more.
Thanks for responding bro.
Anything for dark-Kermit -- admittedly I'm still wrestling with getting "The perfect persona", as it's easy for it to simply reuse the original songs almost beat-for-beat -- I know that it's still a beta feature like Covers, but I wonder if its similar hit-or-miss like Covers, because some personas are just absolute GOLD
So what I did was set a simple genre style (punk) and then loaded it down with vocal specific prompts and drop d tuning. Negative prompt out vocal distortion and distorted voices. Saved that, and I can basically change the genre and some of the vocal prompts and get really good voice and instrumental sound consistently.
I still can't get a banging ass punk bass sound tho. Sadly.
Like it shows up but it always seems to be kind of mixed down to much for my taste.
🤔
Has anyone messed with mixing and production prompts at all?
I haven't found anything that seems to respond as you'd expect, if you're talking about -6db vocals or dolby atmos, etc
Yeah, a LOT of users add stuff like that. I've experimented with it a lot (as have many others) and haven't found anything consistent.
plenty of "Oooooh! That worked!!" moments, only to be deflated when it fails on the next 5 renders 😄
like chord callouts... I've had as much luck getting a track to use chords that I write out on a post-it note as I have with chords called out as tags in the lyrics.
Here is another go, I was trying to see if adding [Male vocals] to each song structure worked (and/or adding extra Male keyword to the style). Still, it came out on the second generations...
https://suno.com/song/b5f582fe-3048-451e-926e-d3dc2a3570b5
Maybe this one too? Hmm https://suno.com/song/a28caacc-7e9f-4a37-8816-6cce19a902c2
Honestly, these earlier results, I cannot tell if this was anhydrogenous at the start or if l got duet'd:
Hi! This is technically not a technical question, but I feel I get different answers depending on who I ask: Do I own the rights to my songs or not? Yes, I pay for a subscription.
Nvm! I found the following under Rights & Ownership in the help in Suno:
I have the Pro plan so I understand this as me owning my music. Phew!
My concern came from the claims made in this video: https://www.youtube.com/watch?v=LrkAORPiaEA
Howdy! Today we're going to look at what may be the newest nightmare in this late-stage big tech hellscape we call the music industry. May the odds be ever in your favor.
New FREE Instrument ► https://bit.ly/vtvolts
Become a Patron ► https://venustheory.com/patrons
Samples and Presets ► https://venustheory.com/store
My Awesome Partners:
🔊 Fre...
...would be interesting to know what happens if, while having a subscription, you extend a song that was made using a free account. Am I correct in assuming that it would default to the most restrictive rule?
(just out of curiosity of course. I haven't actually done that...)
It's actually covered by Suno's help page:
https://help.suno.com/en/articles/2871105
There are several factors when determining rights or ownership to an extension. First, let's discuss the basics of ownership relating to songs made in...
This satisfies my curiosity. Thanks!
(The video I posted is 1 day old and he scared me more than being informative. Reading the rights section in Suno's help made me relax)
His point is a bit flawed, preying on the whole anti-AI crowd for views. And fear mongering of course.
Basically, his point comes down to people distributing a derivative work and registering it with Content ID.
You do not need AI to do that, much easier to do it manually with changing pitch, speed, etc. if someone wants to target you. For AI, unless specifically uploaded, it would be a coincidence unless the artist did not register their creation with Content ID.
Regardless, it is something you can dispute with YouTube. Although I know YT claimed they could not do anything, if you have the documented proof it is yours, they will change their tune (esp. if you DO have copyright registered). But, yeah, likely better going after the distributor the person owned.
Lastly, AI companies do not have copyright over the Outputs. They have ownership. For copyright, specifically copyright protection, that needs to be registered through the Intellectual Property Office of each governmental institution (e.g. US Copyright Office).
At the moment, by USCO laws, AI Outputs are not eligible for copyright protection due to insufficient human involvement.
If you want to know more, I can link to my writeup about it or you can look into the case with the comic book Zarya of the Dawn (invoved Image Generators) and their copyright registration
"insufficient human involvement"? Who determined that? A judge? I spent 15 hours on my Halloween song. lol
Yep, literally
So they thought you click a button and done. lol
(I mean, you can do that, but well...)
Also, doesn't paying for the credits spent account for anything?
Basically the artist lost her protection against her character likeness (as they were generated by using AI) but retained protection for the series name, plot/story, dialoguse, character names, etc.
I think she also generated the comic panels so she has no protection for the individual scenes but the arrangement of the comic book IS protected (because that involved "sufficient effort" by a human) which means if someone redistribute her comic book as-is, she can sue them.
One of the earlier court cases involving copyright and AI generations which is why you can find news about it
Copyright is stupid anyway. I copy the appearance of the world into my brain when I look. And copy it back out when I draw it. lol
Having a sub grants you the rights for the pieces in the first place. Although, at the moment, it cannot be copyright protected in most - if not all - jurisdiction, that does not mean it forbids you from distributing it (which you needs the rights over the product for, and will need to provide proof to distributors). I think Suno Help page can cover the specifics of that
Definitely. Isn't it something like copyright expires like 70 years after the owner's death? All thanks to Walt Disney for having it extended so long. It used to be more reasonable in the past
I heard something like that as well once.
But at the same time, it is great to have the protection in place when you put in so much time and effort
Sure. I don't disagree. (Or rather, I don't have my list of arguments prepared)
I went into this in more detail during the piratebay case. But that's ages ago.
Dolby Atmos on Udio makes more of a stereo pan effect, so I suppose that's why people have tried with Suno.
Hmm. Hearing some artifacting in what I think are supposed to be harmonies.
#1256301881966198805 message Here's a link to when we tested it out -- I recall Jon Fly mentioning from his tests at stacking dolby and similar prompts, that he found dolby atmos, country sounded more like the type of country music that you'd expect to find in a film score, but it was subjective.
Goes to show that everything we use as input has SOME kind of impact on the song, it just might not always be as straightforward as we hope.
@hollow hare Suno should start marketing AI music as a synthesizer. The ultimate synthesizer. Or the ultimate sample pack. Laws develop way slower than technology. Before commercial flight was introduced, the boundaries of a property were extending infinitely high, making flight rather impractical.
Well, they are already getting sued, so it doesn't matter heh
But the effort level required for the creation is seriously something the court does look at to see if the copyright protection needs to be stripped away.
(That human clause is the reason why the person who owned the camera and developed the photo for that selfie from the "charismatic ape" that stole his camera could not get it copyrighted. He did not take the picture and it was not intended for the ape to take that selfie)
I remember that case. lol
Or I read about it.
What about the fact that DAW techno producers have used random generation to generate rythms since ages? (no need to respond, as there are usually other elements in the song).
Solution: Make generating a Suno song require 50 clicks. 😛
But thanks for your help. Appreciated.
We know there the likelihood of copyright being Forbidden from AI creations is not going to happen. MAFIAA is looking into that technology. IMO, they will likely just raise the bar to restrict who, or rather, how much (or hours) someone needs to spend in order to get the protection...)
Or how companies can "legally" use copyrighted materials via licenses... well, anyways, that is something likely that would happen in the future)
Yeah. In 25 years when the judiciary catches up. 😂
It's one thing, that the AI sometimes ignores the [End] Metatag, but usually it puts either some random instrumental stuff or repetitive stuff from the lyrics in there... This for e.g. is something interesting I have never seen so far (is that even a language or just glibberish?). Starts at 01:10. Is that worth to report directly out of Suno to further improve Suno or is that not needed?
You can report it, it is technically a bug, but this is pretty common. Referred to as a hallucination or sometimes "SimLish" because it's like the fake language in the Sims game.
Hallucinating occurs pretty often, but I think that applies most likely to jump accross the lyrics in a not intended matter. I only got those glibberish textures in "Instrumental" tracks so far, but not in an actual song.
Yeah, in an instrumental the entire song can be a hallucination if Suno gets confused and thinks it should have lyrics... and then there aren't any in the prompt. If you like this result you can force it by adding "vocals" or "vocals, vocals" to your instrumental style prompt.
It's a bug alright, but many feel it's a good bug. Many of us actively try to induce it to achieve some crazy effects.
This one (My Halloween entry) is based on this hallucination trick where I started with a hallucination and then rendered (and rerendered) pieces bit for bit for hours because Suno constantly wanted its own suffering to end. In this way I could achieve some pretty dark wrapping effects (throughout, but especially toward the end) (total 13 stitched pieces, 15 hours work): https://suno.com/song/01838338-8452-437a-91ef-d3f0e1a8cfe0
So I and many others feel this bug is a feature.
It's fun to cover a track with hallucinated Sunoese vocals using English lyrics that match the flow of the Sunoese.
This is a cover of one that had all Sunoese gibberish vocals. I matched up the english lyrics https://suno.com/song/0a0b70a8-f066-4cfd-b3af-cfd47cae1637
Dark Playful Guitar, Soulful Operetta, Emotive, Dynamic Expressive Female Emoting, Soft, Complex Frosted Flakes song. Listen and make your own with Suno.
Heyo, has anyone else had issues with 'Persona' songs not showing up in their designated Persona?
I have three songs made with this Persona, and aside from the original, one song says it has the Persona while the other doesn't anymore, and only the original song out of the three is showing up in the actual Persona area.
So far, only the designation and classification of the Persona songs are the only things I've found that are wrong with it. Otherwise, the feature works amazing with actually making the songs. ^.^
Hello., guys) I want to say huge thanks for the amazing tool SUNO )))) One problem I have, just want to share, maybe I will find the answer, so I have created a song and the song has a lot of noise, I tried MIXEA (a tool for mastering and mixing) but still I can hear very strong noise. any suggestion on how I can change it? https://suno.com/song/923348d5-1419-41dd-8746-7d9cecd3fa20
style is a fusion of smooth jazz, funk, and R&B. His music features soulful piano melodies, saxophone and violine song. Listen and make your own with Suno.
Hi, while I wish I worked for Suno no one in this channel does. However we do tend to feel the same way. Suno is incredible. The noise you are talking about is the washy flanger type high end sound right?
Great song by the way
The noise in your case is very low by comparison. I am not sure you can get much cleaner than that using suno. You could put it through a compressor or de-hisser, however, I would be afraid that you would lose that brightness your song has, it really doesn't have much noise. I can show you some songs of mine that are loaded with noise.
In many cases I would tell you to not do long generations, but that won't get you any clearer than what I just listened to.
Thank you for trying.🙏🏻 Well, I was creating other songs in another days, with just "male vocal" on style prompt, and it was good and understandable for suno. That's something with that night. I don't know what happened... and why. It worked earlier, but not today.
I am seeking a way to add a type of music genre that is not on the program. Who can I speak to about this?
Hello, I'm pretty new in suno, I have question, when i extend song, AI doesn't pass 2 min, no matter what song I extended, it's a bug or nerft to song extend?(<- that doesn't have sense but I can be wrong)
Someone know why it happens?
More in short:
Extended part don't pass 2 min at any song, meanwhile not long time ago, it could extended 2 min+ and it means I could use whole lyrics
I am very frustrated with your slow responses. I paid for credits, but when creating videos, your brand is still added to my videos, and the music quality is very poor. It doesn't fully use my lyrics, and I have to create multiple times to get a complete result, which consumes too many credits. I request that you review my account immediately.
Please contact support@suno.com
We are a community-ran server.
But that sounds normal to me
hello, why i buy credits but video create have text "SUNO" and music do not create all lyric
Can you provide the URL to the song please?
Contemporary Worship Music,
American voice,
Piano,
Acoustic Guitar,
Bass Guitar,
Drums,
Electric Guitar,
Violin,
Synth P song. Listen and make your own with Suno.
Do you use VPN? If so, what service and ehat region do you connect to (or connected to when it appeared).
Do you use the Suno app too?
i do not use VPN and not app. I use web
Okay, do you use any API system or anything else you think may have compromised your account?
no
i create music from https://suno.com/create
Okay thank you, I will send it to the team that investigates this issue (you are not the only one to get it).
It might take some time since the investigator I know is on vacation but sending an email to support@suno.com with the answers to the questions I asked should get it checked faster
Can you summarize it for me and ask me to fix it because I really need it, I have 2 accounts with the same error.
Can you provide the song or account links to the other accounts?
Iirc, the team actually has to go into the account settings and change some values or something
(I got your email, thanks, do not want it displayed publically here because of email spammers)
ok, thanks
Anything else you think might be important?
please fix it fast so i can work
Unfortunately nothing I can do on my side. I will send the information collected to the investigation team (saves you almost a week) but how long they take depends on how many people they have to investigate with the same problem.
(Like I said, you are not the only one affected by this problem).
The questions I asked will also save you an additional day.
Hopefully the problem will get solved soon 🙂
They can only investigate the accounts given to them so it will be just the one you told me about since you didn't tell me the other one
thank you
You're welcome 🙂
Model v3.5 has always had an extension time limit of 2 minutes. This has not changed.
Part 1 - can be up to 4 minutes
Part 2 and above - can be up to 2 minutes
There is one thing this is different for: audio uploads. The first extension of an audio upload is limited to 4 minutes minus the length of the upload. So for example if you upload 30 seconds the limit for the first extension is 3:30.
https://suno.com/song/7dca0b96-061e-42fc-991d-f2bbe3940e64 I generated a piece of music, but why don't the lyrics follow my instructions? and always adding "suno content" ? anyone can help?
mellow 1980s ballad in C major, intro, Smooth melody, moderate tempo. Use synthesizers and strings. THE END. song. Listen and make your own with Suno.
Can someone explain to me why Suno is so dead set on repeating Verse 2?
Hello, yesterday I created a dozen or so songs on Suno, why don't I have them available today?
Oh, then how some of my later extended songs had 2 min+?
If I will find them, I will send screenshoot
Aslo, even if it had 2 min max, even there, I need extended like 3/4 times, but after 3th, whole vibe, theme is dead, or still I need extended, beacuse AI had trubble to fit whole lyrics that cut in 4 min part
That's the problem I have
I believe that v3.5 extends of uploads can be up to 4 mins minus the length of the upload.
If you want to keep the vibe/theme of earlier extensions, use Get Whole Song to merge them back into a full track and then continue to extend from that merged track. When you do an extension, Suno is only able to read back as far as two minutes or to the beginning of the extension that you're extending from (whichever comes first). If you don't merge those segments back together between extensions, Suno won't be able to use them as a reference.
I created 2x2 today and no songs. Is there any Suno issue?
a bit more on that... I've had tracks with 12+ extensions where Suno has used bits from the very first extension in the very last one, but only because I did Get Whole Song merges in between.
I mean, I've tried extending from 4 min + X min extended aka Whole Song, but AI have problems to fit while cut part of music Intro that 2 min, and always I need to extend like one line of text that was cut, or even less
And that kinda is annoying to be honest, beacuse i need like 4+ extension before I can fit whole lyrics
While like month ago (not sure), i could fit whole lyrics in 4 min part and extended part, and that's all, not like now
But I guess it's now reality that you need 3+ extension instead 1
Hello, I will be DMing you some questions to investigate this matter. Please try to answer those questions to the best of your ability
i need help sometime but not always even in the same formatting suno is singing my bracketed instructional tabs...am i doing something wrong and/or how should i be approaching this so the instructiosn in the lyricsa are not sung
The AI likes it simple and direct. Rule of thumb is to use like three keywords at most per metatag.
Also make sure the metatag is using [brackets]. The (parenthesis) is used for background/backup singer vocals
If you give orders that the AI does not understand in the [], it increases the odds of saying it out loud
awesome ty kindly
Did suno get quieter in it's output?
Not that I've noticed
I checked my audio setup and everything else is much louder where suno used to be relatively close. I feel like I'm really cranking up the volume to hear it now.
Check that this is at 100% -- it drops off QUICKLY
That's the first thing I did,
Beyond that, I'd have to side with Toker 🤷♂️ The audio levels certainly vary between songs, and sometimes the choice of style prompt can play a big role, but I haven't noticed a change regarding volume.
Before I say that, the audio mixing on v4 has been improved -- really satisfying getting a guitar tone with real depth, and sometimes it does "pop" a bit more.
I don't have V4, as far as I can see. but, that's what I am noticing with my generations. The metal genre got a real boost to it's quality it seems.
The v4 comment was just a side comment to note that if anything, things are improving in other channels, but I still use v3.5 regularly and feel nothing has degraded.
It doesn't seem so squashed together. A much better stereo sound ad balance. Not 2 days ago or so it was the usualy very loud squashed mix.
v4 goes so hard, it's making a metalhead out of me 😛 All I know for certain is that the tracks I've made with the new model have phenomenal song structure and make me so excited to create more 😄
I want it.
Don't know if there is anyone on the Suno team here. Maybe that should be a direct email but with using the replace section tool, would there ever be a way to have an insert audio tool within the replace section?
Is there a way to opt-in for the v4 beta?
Not that I know of -- you can email support@suno.com if you want to ask the team directly, I suppose?
I'm hoping v4 will help a ton with my style of music I enjoy writing and producing with Suno
Anyone else have problems when trying to delete a persona or song related to a persona?
While the staff are here, and the support lead has been increasingly involved, Discord isn't an official support channel.
support@suno.com is the direct email if you need to contact them, or #💡┃suggestions-web if you have ideas for Suno (they DO check that channel)
I can send a persona to the trash, but cant really delete it, so cant delete songs of that persona
Thanks! I'll throw this question in there.
👀 You seem to know quite a bit about V4
The red role does mean something 🙂
Fair. That didn't even click, lol.
danger
...yeah. Personas can only be hidden in the basement atm.
Seems like it's hidden from your view unless you choose to restore it -- so "sent to the trash" is more like an Oscar the Grouch kinda deal, where they HAVE been sent to the trash, but they just sort of stay there and set up shop.
Anyone have any issues writing a song about God? When I try it never generates just keeps spinning, but when I try a song without it. It generates almost automatic
Vocal Tags: A Hit or Miss Experience!
I've spent considerable time trying to control vocal delivery, aiming for lyrics to be sung exactly as I envision. However, no Vocal Tags seem to work consistently across styles or genres.
Key Observations:
Vocal Tag Variability: Vocal Tags perform unpredictably, with effects varying across music styles.
Tag Format: There's no standardised way to mark Vocal Tags, though brackets [ ] are generally preferred.
Text Graphics: Special characters like ~~ for vocal runs don’t work reliably.
Suno follows a basic formula of song sections (Intro, Verse, Chorus, etc.), yet the lack of a foundational set of Vocal Tags for directing lyrical delivery limits creative control. Having a standard, effective set of Vocal Tags would significantly enhance the songwriting process!
If a reliable and consistent Vocal Tag List existed, it would be a game-changer for music generation—whoever had it would have a winning advantage. Since Suno hasn't released such a list, it seems one may not exist!
Could you give an example or two of a genre and style with the type of vocal tag and delivery you were trying? I'd like to have a go.
I've tried to add the list but now I'm being told I need to pay for Nitro thing!
I don't have any experience with prompting religious songs, but I know that sometimes there's a delay in service, regardless of what's being prompted.
I'm currently experiencing that trying to generate an instrumental indie song, so take from that what you may. 🤷♂️
You should only get prompted to pay for nitro if your message exceeds a certain number of characters.
Following up on the vocal tags thing, maybe your own #1252267444773326848 thread might be a better place to organize a lot of ideas all in 1 location for others to see 🙏 💪
When I plug in any other lyrics it seems to work just fine
- Vocal Style Tags
[male vocals] / [female vocals]: Sets the vocal style to male or female.
[baritone]: Low, deep male voice.
[whisper]: Soft, hushed vocal delivery.
[belt]: Powerful, full-voice singing.
[falsetto]: High, light vocal tone.
[scream vocals]: Adds intense, screaming vocals.
[unclean vocals]: Rough, gritty vocal style.
[harsh vocals]: Aggressive, rough singing.
[hoarse vocals]: Raspy vocal quality.
[choir vocals]: Adds a choir-like vocal effect.
[vocal harmony]: Adds harmonized background vocals.
[assertive vocals]: Strong, confident vocal delivery.
[clear vocals]: Ensures vocals are distinct and easy to understand.
[vocal sample]: Uses a sample vocal at the start.
[vocal refrain]: Repeats a vocal line, often in the chorus.
Here are just a few tried over and over by friends here on Discord. What I found was that it isn't easy to keep track of anything here, people have tried the same tags over and over and over again!
We need a folder on the left hand side for Vocal Tags. That only contain working Tags, stating what the Tag does and the Style of Music, basic Genre, in which it works.
What about the genre you've tried these with? Also, I presume you are trying these in the lyrics section?
I'm asking because all these tags can be 'utilized' within their respective genre's and with some being cross-genre used.
Albeit, you might have to get lucky on the roll of the die with some. Like the vocal chop. I get it in some hiphop and more often if I put 'boombap' in my style prompt.
You can kind of 'trick' suno into doing a vocal chop by typing out the lyrics in a chopped manner.
would be a great pin
https://discord.com/channels/1069381916492562582/1154587937611190354 you can also give suggestions here.
[Theatrical Vocals]
[Sultry Vocals]
[Breathy Vocals]
[Deep Scratchy Vocals]
Just a few more to add to the list that I've personally used or consistently use.
Congrats on the early access, @polar spade 
This is a good idea like the site that is dated but was a good idea and I still use to remind myself of specific things. It reminded me about narration tags (Good for Instrumentals):
I have used almost everyone of these with decent adherence:
[Story Intro]
[Character Theme]?? <- not too sure about this one
[Plot Development]
[Climactic Build]
[Narrative Resolution]
[Plot Twist] <- This one is good for getting Suno to completely swap instruments
[Surprise Ending]
[Key Change]
Did you apply? Or was it more of a knowing people thing? Oh and congrats
🤷
I've only done one Song About God and it was just now. Didn't have any issues rendering it: https://suno.com/song/4b4b321d-68e0-4693-8540-cdfda60d277a
Playful, Flamenco Guitar, Comedy, Dark, Sarcastic song. Listen and make your own with Suno.
System Error!!!
Ok. So Zelda then, "It's a secret to everyone" lol
Are you receiving a popup message when trying to generate song?
Yes
Thought, it is too long (1400 characters) for model 3
Don't think so
Messaged you
That is what I was going to say. Most of the time my errors in v3 come from that, if that isn't the issue, then lyrics content?
1400 charactors too long?
but model 3.5 no error
I'll let another in here answer that, I would say yes but I don't remember, you would think it would be half of 3.5 since it was from 2 to 4 min, but I had issues with even less. I am not too sure about v3, my experience is limited. I just recently found out from a mod in here that the Sitar is in v3 when I could not get it to show up in v3.5 no matter what I did
@charred merlin You know about v3 can you answer @neat bear
Yes, took some lines off, then no error 😮
Yea...the issue I had is that it showed 3000 char even though I knew v3 didn't accept that many, and I couldn't remember the limit...lol
What count did you go to to stop the error?
1200?
I think it's due to your lyrics "Hello God its me". Suno is probably associating those with the song Hello (God it's me) by Myracle Holloway.
If V4 releases this week then the wait isn’t really long. I think I won’t apply for beta in this case
Nice catch
Does it release this week?
@neat bear -- thank you for confirming a UI bug -- v3 is currently showing a character limit of 3000 -- this is the limit for v3.5
v3 has a limit of 1250 characters 🙏
Matt Vid pro kind of said in his last video it does
Ahh...I gotcha. Well here is hoping.
The fix turns out to be as simple as misspelling it like "Hello Gahd, it's mee again!"
Playful, Flamenco Guitar, Comedy, Dark, Sarcastic, Acoustic Rap song. Listen and make your own with Suno.
That wouldn't render with the original text, but as soon as I misspelled it... rendered.
Also, God is a Cat
Mainstream Pop, Contemporary R&B, Trap Pop, Neo-Soul, Electropop, Experimental, Very High and low female Vocals song. Listen and make your own with Suno.
So it was hitting the copyright or whatever you want to call it, the protection during generation. I'm interested what the parameters actually are for that. I haven't ran into that issue yet, except for my Metroid shot out.
That's what it seems to be in my testing, tho they're still having problems getting the whole song to render.
What you just described @worn flicker happened to me last night, I was messing around and it just won't generate, will just keep throwing errors but sometimes. It actually will do it but only once.
Yea, I have hit the errors but it was never due to lyrics. Length yes, content no. I am truly interested, but I imagine the dataset for that is possibly more enormous that the datasets for creating everything else. Music done been around for...ever
To clarify the Matt vid pro thing, he is someone who got early access, and he said expect a video on Suno V4 quite soon on his Tuesday's video.
Thank you for the clarification. ❤️
yeah I was reimagining KISS songs and it did this on Love Gun, I stole your love. songs
basically was just adding the lyrics and making new music. some interesting results for sure.
Never try Hard Luck Woman, it's always terrible. LOL
Any time I am playing with an inspired song, I use ChatGPT to get me started with ideas on how to take it to a similar area but down a different road.
i've done that for the stuff I wanted a Portishead vibe to, using the prompts I posted here. The results are amazing.
My song "Wrecked" actually originated from, no judgements, the Backstreet Boys' song "Incomplete"
That was a happy accident when I was seeing what could go on with many Vocals.
no judgement here my friend. Inspiration comes from all
i mean ...I'D never do it LOL
Aww how you gonna do me like that,,,lol...*slumps head
Jk
Hahaa
For real, that is one song from them that I actually like. The vocal handoffs are incredible and the piano resonates beautifully
I have to say I don't think I know it. I have not listened to the radio in over 30yrsd
That's about right...lol it isn't a new song. Maybe 20
yeah it's been a while. and i've not watched regular TV in almost as long.
My wife watches regular TV so I see it by association, but I don't know what channels exist on the television any more
Yeah same. I've been streaming or back back in the day downloading what I wanted.
TMD back in the day? F-Serves on IRC
yah that was super back in the day but yes. I remember it well then Pirate bay, the EZ TV. Now I just stream.
i mean with YouTube and the streaming services, you don't need to hunt for anything anymore, its all there.
Yup
this is where you and I go "Back in mah day...." so all the youngins in here can be like "WAAAAAAAAHT??"
Lol...yup...I feel an ache in my back when strange things that I didn't think about happened. When my kid came up and said I dialed this number for the DMV and it is making this weird noise. He had never heard a busy signal before. Crazy the things not thought of, and even funnier that the DMV would be the last place that has one line.
Lol
hahahahahahahaha
My brother showed my neice a rotary phone and she was like "what is it?
its sooo funny. man I'm old
Lol, yea...why is there a bungie cord attached to that phone and why can it wrap around the kitchen and living room 3 times and why is Grandma wrapping the walls with it?
hahahah yep. man I miss some of that stuff I do but I am so happy with where technology is now. I really can't wait for the day when I can play along with AI on a song in real time and we just colaborate. I keep telling friends who are musicians its gonna happen and they think I am crazy...oh and btw...ALL of them hate suno.
they all think they will be out of a job soon and I'm like "man that is very old school thinking. if anything it's going to free you to be more creative"
I get it on both sides. The good thing is that I have yet to see any AI be truly original, like I haven't heard Steven Tyler or anything close.
A secret Menudo fan here. Hahaa.. well, all the kids I was "uncle" to.... I had to listen to lots of boy bands.
I feel like I came close Sysmai, when my album Afterdark sessions is out I'll point to you what I mean
Hey. So now I write kpop. So who knew.
@polar spade Cocteau Twins fan here. and...I won't lie...Bee Gees
Oh yea. Cocteau Twins. I wore that groove out.
I put out a Kpop album as my first AI colab...I did a press release and Kpop fans were NOT happy. they were like "how could you do this?"
I like the way kpop song flow green name person, but I don't know enough about it to use it properly
if you guys look for Luna Flare - Auroura Rhytym on spotify, that's my release.
That is a good green though
it's a pretty good album if I say so myself.
See that is why. I don't want to offend anyone.
Hahaha... oh, I'm so sorry man. I guess there a vibe? I best make sure myself then.
My death metal album is out too and my DnB album but the DnB album is mostly tracks I wrote in 2015 that I never finished. Suno helped me get the job done.
Is it a public Playlist, I'll give a listen.
@polar spade in forums when I promoted it I got blased "we don't want your AI garbage, Kpop is about the artist"
yes it is public, the album was released thru CD baby for distro
same with the Death Metal album and my DnB stuff.
I wanted to give the Kpop and Death Metal albums away for free but they won't let me.
Oh...oh... yes. That much is true. The whole kpop demographic... and pop.. country... etc.. is feeling a belonging to the artist, having a crush, getting DMs, joining a fan group....
Again both sides. I AI generated songs will almost instantly saturate the genre. I don't want that, they should just have dedicated areas/channels for these things. I know some things already do.
agreed, I think if they did that, it wouldn't piss as many people off.
So writing a good song without the artist is counterproductive so far. It's only good to sneak in streaming radio airplay lists.
Now that I say, writers deserve the same amount of credit and money in that area
Not tour money, but a cut
hm,mm
If you think they are good, hit the kpop producers websites and see if you can submit to them.
Gosh, there's one that has, like, a submission contest yearly I think. I should find that company again. 🤔
I may just do that. I want to get my DnB and Kpop music on games like Chunithm
it would 100% fit
Start with YouTube and find jpop boy groups or kpop... check the video owner for the company names and often many have website links. Each producers have about a dozen band/groups.... they may be exclusive to a writer(s), but always try.
I just might dasvon, thanks for the tip
ok I gotta get ready for work. talk to you all later
Does anyone know how you do that the singer sings without music in the beginning of the song?
It can be trial and error to actually get it. When you hear the long ones it is probably made first and then extended with a style
[Spoken Word] xxxxxxxxxxxxxx
[Intro] [Spoken Word]
Xxxxxxxxx, ccxxxxxx, ddd...
Stuff like that
Ok! Have tried that. But the song starts with instrument and song together anyway
Don't forget about v3. There are things in v3 that I have not been able to get in v3.5, also others maybe starting with an audio upload
Only ger error when i try v3
Character limit is 1250
Anyone know about the reverb in v4? I mainly do metalcore. I can share an example if needed but it sounds like something from an arcade.
When does v4 come?
Sings. A cappella? So , singing a melody directly at the beginning without any intro or accompaniment. That may be a tough one.
And without harmonies too...
How long of an intro solo voice singing were you needing?
Yea, I was only talking about the start point, I have never pulled off a vocal solo, closest I got was using Sung Poetry in the style
Is v4 allowed to be discussed?
So , are you free or a paid sub?
Paid! It works with v3 but I want to do a persona on the song, that don't work with v3
I don't know if a 3.5 cover of a 3.0 song will allow a persona after that? Probably not.
Maybe I'll try. Let's see......no I couldn't access it at this time. 😒
Can you download the audio from v3 then upload and extend on 3.5? Never tried but I don't see why not
is anyone else having trubble creating custom songs ? for me it loads for a long time , then give me "a clip got stuck generating"
Lyric content is the reason most of the time
aha, cannot be explicit ?
Depends, I have had it and seen it strait up say explicit words, so I am not sure exactly what is or isnt allowed.
But, a good bet is that is the issue
Try without
alright, well im going to try, but when i had some sort of text that not allowed like names of artist etc, it straight up says so 🙂 now it just says "a clip got stuck generating" then nothing more
Are you under an NDA? If so, that might let you know what you’re restricted from doing.
Otherwise, users are chatting about everything & everything we have access to 🙂
you where right, i switched some of the explicit lyrics and now it works
I have seen people say both, and it still was lyrics. I think of the errors like web errors. It may say this one time and another for a different time but it is the same error
ah i see, feels abit random on what texts that are strickly "forbidden" made other song with some texts that is allowed
It makes errors for sure. It is hit or miss. I have nothing to back this up, but just being in here seeing people come in for different things during different times, it would be a safe bet that more safety is being implemented. Rightly so...lol. there is no way Suno would be able to go on without controversy. I mean even chat bots get blamed for things whether they did or didn't do it...lol
haha true, no i can see why they would need that 🙂 but its kinda common in present day that they use more and more explicit text in song 🙂
I just heard a sample of v4.. the new voice update is a game changer, I hope we can go back to older songs and update them with the voice and still retain what we originally had with style...
I have used Cover with no instruction many times, but not with different versions. Ohhh is that what you mean? I think I figured out what you were saying while typing this message and yet I am still typing.
yeah like...take a song I did which is already very very good and have the new voice redo just the vocals in the same style and format just updated so it sounds more real.
if we can get that....and a way to knock out bleeds and ghost noise in tracks without covers or extending (because these do change it a bit) that would be killer.
Yes. Agreed. I don't know enough about cover and the difference in compatibility, but I saw chatgpt move to switch models mid conversation, but audio vs text...lol. no telling how much more difficult it would be.
The only way I've managed to add lyircs as part of the intro and then download the stems. It does mean a bit of messing aound in a DAW.
But as Sysmai said, you can try prompting in the intro. [Spoken] and [Narration] have worked sometimes. I've not tried any of the other suggestions but I'll certainly give them a go.
Don't forget, you can try to metatag an epilogue so it might finish up the song with a singer solo, as it often does anyway. Ultimately your wanting to create a persona? If I recall, you can't persona from a stem?
Hi I recently started using Suno. Need help as I have created a song where I missed a verse. Please tell me how do I insert the verse (2nd verse) in the song that I have created? I tried Extend, Crop and Replace but none of these meet my requirement.
Is it possible to exclude rap from the styles but use a [rap] tag to trigger it in specific parts of the song ?
👀🫀 🫨
Free member? Paid subscriber?
And welcome!! I'm a user member as well. It's a fun app.
Sure. I've put just [rap] as a tag just before the rap text. Works for me.
I see on one song I put [spoken rap] and it worked. Probably redundant though to use both words.
we need to put a request in (feedback) for "delete persona" it makes no sense why deleting it, is not an option 🤔
How do you guys already have v4?
Hi, I'm trying to create a track that loops so I can use it as main menu/lobby music for a video game. How do I achieve the looping? thanks! if there is a guide, happy to read it.
Can anybody help me? i have problem with my billing
You'll need to contact billing@suno.com -- Discord isn't an official support channel as of yet 🙏
ok just dunno what todo(((((((((
The community can try to provide technical assistance, but usually the error messages related to billing are a matter of payments not going through,even when there's a sufficient balance -- unfortunately there isn't anything the other users can assist with in regards to these errors, and the dev team don't offer first-contact support through Discord (though they sometimes choose to follow up through Discord if it's beneficial for both parties), but the first step should be to contact either Suno's billing department or your card issuer 🙏
What have been the most intriguing aspects of the update so far? Any specific dimensions where it blows v3.5 out of the water? Can't wait to experiment with it haha
The suggestion ive heard is to find a spot in your generation that would work if looped and crop at that point. But the better way is to prep the song as much as you want, then download and trim in a DAW to your liking.
At least, that seems like two easier techniques.
You can reply to the invite email if you haven't been invited into the user-feedback discord group 💪
thanks. noob question -- what do you mean by prep in this case?
Thank you! That was exactly what I was hoping to hear 🙂
#🗣️┃chill-chat message Dynamic range is phenomenal, lyrical adherence is through the roof (of course I'm using an instrumental song to demonstrate 😅 )
a
I just mean to use all the tools you need to get the song sounding best BEFORE you download. You can't bring the song back up into Suno once taken off the app.
You might like a less "produced" sound and go ahead and d/l and do everything off-site. Up to you.
Damn, that's really exciting! Clarity is still really solid through the crescendo at the end - that's when I feel I lose it a lot of the time with the current version. Cover art is great btw, well done with that haha
ah of course. thanks! and.. is there specific DAW tools that are designed for novices/broadly used?
Bandlab 
Free 4-stem splitting, AI mastering, in-browser audio editing, royalty-free samples, AI music gen to get song ideas started -- just ridiculously artist-forward
excellent
Is it a good idea to request beta access, or are we really close to the release? I mean Matt vid pro said with a slight pause quite soon like it meant really soon. Ok, i'm doing this same kind of thing again I did with V3... If V4 will be released this week, someone can DM me the most horrific Suno song ever with noises so loud, like dubstep or whatever, that will make my head spin. I mean I had it happening with V1, after listening for it's dubstep I was like going crazy because it sounded so repetitive.
have you noticed anything intriguing about arrangements inV4?
I don't know who Matt Vid Pro is, but the discord-user-feedback team, alpha team, devs, and mods are saying they have no idea.
The head of social media was quoted as saying "not a matter of weeks" when asked if it was days, weeks, months, etc
As for whether you should apply for beta access, it looks like the "early access" team has evolved into the "user-feedback" team -- make of that as you will -- either way, don't make any major decisions based on whether or not v4 will be released today, tomorrow, or the next day -- it'll be here sooner than you know it 💪
If you were around to appreciate the degree of structure that v3.5 brought to Suno (warts and all), that kind of describes how some things just seem to "click".
There's things I wish were different, just like the v3x series, but I'm absolutely loving it 🙂
it seems like when i go from prompting one style to another it takes a bit of time to get that style "out of it's head" not sure if i'm trippin.
Do you have examples? I've been meaning to look into this, but I never get results like it, because there's always SOME kind of reasonable venn diagram overlap, or it's when I'm using a beta feature like covers or personas or something
you're not tripping🫨
It would be ideal if it were going from making metal songs, then making classical songs -- that way we could see a "cleanly prompted" classical song with obvious metal influences
Yo so support got back to me but yo this watermark issue is frustrating fr
Anyone else try to generate music only for the lyrics to be "Dont use this fake app go to suno.com/ios" or something
makes no sense cuz i use https://suno.com/create
@noble sequoia -- they've explained that they're working to resolve the issue for you, and will give you credits to compensate you for your trouble.
They made this change and were able to prevent a LARGE number of people from stealing their service -- it's unfortunate when someone innocent is caught up in the matter, so they're working to rectify their mistake.
If you have any information to share with them, like they asked in the last sentence, that could help them speed up the investigation to figure out WHY this security protection is mistakenly flagging your generations.
It mentions that some of the regular members, same as you, have used Suno but have been affected by Suno attempts to stop other sites from hacking into the site. So that makes sense, though certainly frustrating when we get caught in the flack.
It's good that those others can be recognized and thwarted.
Are you up now? When all is clear I'll listen to your songs. What's your style you write?
Hey, is it possible to change the language but keep the song?
There is no "translation" mechanism.
There is a "replace" tool for paid subscribers. You choose a section, correct or replace lyrics and click replace to have the new lyrics inserted into an existing melody/arrangement.
How accurate it would be if you replaced a section at a time, I'm not sure. But that would be really fun to try. It's on my list now. I'll (attempt to ) change some Japanese song to English.
You can "reuse prompt" and simply replace one language song for another. Trying to keep the physical syllables and arrangement the same, you might get a nice sounding song nonetheless.
"Cover" will change some newly typed in words occasionally, but I have not attempted to change all the lyrics using "cover" as that was not what it was intended for.
Replace works wonderful. I had to replace a word that was mispronounced and the result was perfect.
{"errors":[{"message":"Unauthorized request","long_message":"You are not authorized to perform this request","code":"authorization_invalid"}],"clerk_trace_id":"c35d593bb3728a004dbe0aaa19f90841"} somebody help
What are you trying to do?
I guess cover feature has no support for 4 minute songs?
It doesn't let me access suno from my cell phone, yes from the page, they blocked my IP, it was after uploading sounds of turtles ....haha
Yes it was amazing they were going to give credits, but hmmm
I also told them I don't use any VPN or filtering. I told them I tried on multiple browsers and got the same issue
Ohhh I see now. I guess I will just have to wait, but wow yea it does make sense why they had to do it
Good day all, hey anyone having problems where the new persona feature cuts songs short? I'm trying to use one but it keeps cutting my songs off at 3:26ish...yet there's one more verse?
It doesn't let me access suno from my cell phone, yes from the page, they blocked my IP, it was after uploading sounds of turtles ....haha
In the event that they have blocked me for uploading turtle sounds in their situation, which seems ridiculous to me, how long do I have to wait for them to lift the block?
That seems very unusual... Wonder if it's due to an IP change and you got someone else's blocked IP? If you put your phone into airplane mode and take it out of that again maybe it'll change IP? Just a guess.
Otherwise e-mail support@suno.com and see if they can figure out what's wrong. If something is blocked it'd need them to unblock it.
Thanks
The strangest thing even is the IP of my cell phone, because I can access the accounts from the PC
Are you using VPN on the phone?
The other reason could be the browser too. Maybe delete your cookies and try again
Sorry if this is the wrong channel for this, but I am not getting credit for my referals. I have two referals who used my link, have made way more than 10 songs, and neither of them generated the bonus credits for either party involved.
Please contact support@suno.com - they should take a few business days to get back to you about it.
thank you
You're welcome. Just provide as much details as you can such as your account @handle (and maybe your referrals too)
It's no fun being the one caught in the middle, right? Well. Glad it's just a glitch and you'll be up and running with new credits soon.
Is your cell phone iPhone, Android, Windows, Google, Huawei or other?
OK. You might have downloaded a different companies app which looked like Suno or said SUNO. But there is not an APP for Android as yet. It's coming soon, but not yet available.
That's the only APP which you may pre-register to receive once it is released.
https://suno.com/android for the shortcut to that page
It is possible you are logged into the suno.com/create website on your PC and have something different loaded on your phone. With this information, you can review and make a decision how to move forward. At any time, I'm sure the folks at suno@support.com would be happy to help as well.
Thanks
What's the airplane mode thing? Click into airplane mode then click OFF airplane mode?
Yes.
OK. Cool. I wasn't sure. If working now and you understand there is no released Android app, then I'll leave it there. Have a great day!! 👍 Hang around and catch the chat.
Thanks mate
I can’t get personas to even render a song… I chose a persona from a prior song and saved it… but when I go to create all I ever get is a “Clip stuck during generation” error
Are you using any lyrics that might stuck to moderation?
What do you mean? It’s just a basic four line song with a repeating chorus
Think of similar to like Little Drummer Boy
Hmm..Try re-creating the persona. Probably there is a problem.
The song that you created it from is not deleted completely right?
No, in fact it is one of my personal favorite creations. I’m not so much worried about the sound as I love the vocal bit.
As long as it's not deleted, it should work. Are you getting the error whatever you try to do?
Yes. I’ve tried two or three different songs that I made personas and it wont work no matter what I try.
I just created one to try, and not having any problems. Maybe it would be better if you asked support about the problem. It's hard to decipher from that error message what's going on.
You've made separate persona's from 3 different and previous songs you had generated? Just trying to be clear.
- How long ago had you generated the older song that you made this persona from?
Yep. I have a few from different songs. An acoustic/ballad, another from something like capella, and from a rock one.
Am i right @vague cargo was the one having an issue?
Ah sorry, I tapped the wrong "reply" 😄
Could your lyrics be hitting the 3k cap
That's happened to me when I've used the feature, lyrics that get to close to the cap cause issues
@weary wharf Nice seeing you in here man. Not sure if you remember me, but you help answered my questions in your workflow chat. Thanks man.
I just got an email from support saying that V4 will be coming to paid users pretty soon first. Was kinda hoping to join the User feedback/beta team but ah well I suppose.
Yes, I made personas from three different songs that had similar vocals and none of them will generate to a new song when I try to use them. (Sorry for late reply, I’m driving right now so typing is a no go).
Sounds like it does not even need mastering. Pretty clear.
I'd still touch it a little to bring the bass out a little more though 😄
Did they improve lyrics generation with V4 too, or did you make this by writing your own or using outside AI?
Just putting it out there. If the persona is or was created from a song generated weeks ago, it's always possible 1) the old lyrics , prompts, or even theme may now trigger the auto-moderator which has been more sensitive. (There are older songs presently sitting on Suno which can not be re-generated) and so 2) using those songs with no changes may simply not gen now. 3) Any personas made from those songs also may not work.
That's just a thought. It's pretty sensitive and will give an error for some words or prompts which are perfectly fine, too. That's why support should know.
Most seem to either emailing support@suno.com or do some control tests. A) Use a public persona on YOUR new lyrics. B) Use your persona on a simple phrase like "this is a test". C) Use half your lyrics on one gen. Then the other half on another. See if both get blocked, an error, or just don't generate.
Claude (paid Sonnet) version, is really good with lyric generation. If you can give it a nice detailed description of what you want, and give a template on how to construct verses, choruses, bridges, etc, it generates some good lyrics.
And there is always pasting lyrics into chatgpt and requesting gpt to review lyrics for any issues they may have for posting into Suno. A good way to test any lyrics for spelling , grammar and censor triggering language or theme.
It's a little work to help rule out the easy stuff. Then on to suspect other issues. Either way, support can sniff it out pretty quickly.
Good stuff. Solid intro and big finish outro.
So I guess with V4, the skipping lyrics haloosinations are way down? In 3.5 sometimes there'd be repeating choruses and verses I guess to try to fit it into the four minutes if there wasn't enough there. The way I see it, if I only give the model enough lyrics for a two minute song, make a two minute song.
Like I said, it did not require too much alteration. Just a little compression, bringing the mids and lows, and some less stereo 😄
V4 will be fun.
Hello.
Does anyone know if it is possible to produce a slow instrumental with only bagpipes and bugles?
I'm sure you can get something by style prompting things like : adagio, ballad, aria, slow . I'm sure others can throw in some ideas.
You'll probably get a lot of Irish sounding outputs. Do you have a genre in mind for this? Having bugles and bagpipes together might be hard.
@lavish surge @weary wharf oh yea. I was focused on the wrong thing. May even throw in "new age" as a prompt and it might use both instruments. Can try french horn, flugelhorn, coronet and uilleann pipes, also
I'll give that a shot tomorrow. Thank you.
Yes, I'm in the Air Force, not that matters. I'm trying to create a melancholy instrumental, featuring the pipes only, then a bugle call at the end.
You can start with bagpipes only and in the style try march or marching band maybe. And replace with bugle for an extend call.
But from experience, with bagpipes, you might get a guitar or piano in the background.
Also enable "exclude styles" to make sure you do not get rock or pop songs.
That helps a lot, thank you so much.
Ahhhh.... OK. Then I would extend back from the end and create s finale section prompting forBugle, military bugle, cadence bugle, anthemic bugle, inspirational and so on...
https://suno.com/song/1947d02c-9cf5-456c-8b7a-01041e27fc45. So here has a French horn slightly in the beginning... but I guess this is like you're going for? Of course, ignoring the noise at the end...
That is beautiful and just what I was thinking of!
@weary wharf <<< he inspired me with his advice. Yea. You'll have to do some extends to get the ending. Ok. Good luck. @lavish surge oh and that was with Custom lyrics ON , and Instrumental OFF. Revised: didn't get bugle , I'm sure it's easy somehow, I just didn't get bugle in a small time.... but got brass! https://suno.com/song/d17fa2b9-d533-4201-acb6-1ea95a9d31d6
Question for my more experienced folks, is there a good way I can input my own melody when generating a certain song?
Thank you immensely.
Hi guys! I am new here. Is there any known workaround for the issue, where Recreate Section ignores timestamp brackets in both same length mode and free length mode?
There is also a huge issue for me with Recreate Section grabbing the lyrics from far outside the selection and trying to jam it where they were never asked for. It is a huge issue because it happens like 5 out of 6 times I try to recreate a section.
Also, is there some kind of syntax to mark the comments for music within lyrics, like when I want to add a solo? The model often tries to sing my comments when I use asterisks. The same question for back vocals, since the round brackets work like 50% of the time.
And lastly, is there a way to try and remove a portion from the middle of the song and cure the cut together? Just deleting the lyrics does not work at all for this, it seems.
I am sorry for a lot of questions that might have been answered before, but I am just desperate, since I've wasted like 1000 credits on takes that would be suitable, if they were not heavily bugged
If you have a song of your own, upload the audio -- up to 2 minutes for me -- if there are vocals, you cannot make your songs public fyi.
If needed you can use Adobe Audition and create a remix and condense the audio to 2 minutes, and use as a "cover".
Or you can extend from the upload from any second. Then it's up to your style definitions.
PS: I do this all the time, the outcomes are great with covers. Extend tends to put noise, but from time to time, you get some really good sounding outcomes.
The upload / audio (record) feature is located on most create pages. Upload or record a file up to 60 seconds for free users.
I can answer one part quickly, [....] tend to work best for instructions
It may not sequence your melody perfectly, so try to have a sample of verse, one of bridge, one of chorus all in 60 seconds if youre free user. Then use lyrics to show proper arrangement. Kind of a shot in the dark. But suno does OK to integrate the concepts together and match the melody with lyrics.
Otherwise you might get a likeable variation of your song. I get a few of those and often better than what I wrote.
For re-create section, make sure that your song is well synced with the lyrics. Listen the audio and correct the song lyrics, including any [break] or [interlude] sections as well.
While you are selecting the range, the selected lyrics should match, else, Suno might be seeing something else.
Not from "edit > song details". That do not sync. But from the "edit song details" on the left of the "replace" view.
| And lastly, is there a way to try and remove a portion from the middle of the song and cure the cut together?
No, there is no way of doing this currently. You can re-try extending from the place you want, and merge manually in a DAW or any multitrack audio editor.
The sync seems to be in order for me, at least when I am listening to the selection, the displayed lyrics are exactly what is being sung. Can the instruction marked with asterisks ruin its perception of the contents?
I am a little bit stumped here, since it happens a lot, except for the song where there were no instructions at all
But I have imagined a lot of probable causes by now 🙂
If this screen shows the correct lyrics, it should be ok. But it's an AI, you might need to burn some credits to get the best.
Try using [...], to define instructions, works best. Example in the screenshot, like [atmospheric break].
I do like this mostly for instructions
[audience goes wild]
[audience joins with the chorus!]
time slips through my hands!
each moment takes flight!
Asterisks were/are also used for sound effects, since Suno is inferring what you most likely mean then it will look at the content and infer the meaning. However, tags are tried and true, they do get sung at times but less likely, also placement of tags. I find that if you can avoid it, try to not put tags in the lyrics. I say this as a person that puts tags in the lyrics quite often, but I know there may be a decent amount of fixing to be done when I am doing something so specific. I try to keep most tags on the side or just under the section title, or at the end of the line. Certain tags get sung more often for some reason. I don't know why. [Building Intensity] or even [Building] has been sung more often for me than any other tag.
I realize I gave more than you asked but why not..lol
Is there a reason to mark sections? Should I do it with [...] also?
I will try and edit all the asterisks out, maybe I will be able to fix this this way. There has been a lot of effort invested in this song, so I am commited 🙂
Yeah, this screen matches and matches what is sung originally
I've seen it both ways honestly, it seems more based on structure, but I always use the section tags for my own sanity if nothing else
Here is an example for you for instructions. You can listen and see what is captured by Suno, and what is not better via examples -- the song in the screenshots.
https://suno.com/song/4ebfc40b-3b50-4835-8e6d-5fb7df5b9957
Consistency is also another key, if you deviate too much with instructions, they either bleed into the lyrics, or passed completely.
Extending is also really really helpful for getting the best out of instructions.
[live concert recording], [crowd goes wild], audience, audience-vocals, [make sure audience sings as well] song. Listen and make your own with Suno.
There are some really good guides in #1252267444773326848 as well for instructions and structure building.
I say do what you find to work best and adjust, I don't know if other people have editted their lyrics after the fact and that is why they dont have labels, but personally I learned about using the section tags by using Suno's generate lyrics button. It did it that way, so I figure...why fight the thing you are trying to get working...lol.
When I am trying to figure out what is causing the repeats I have a couple tactics. First, while I don't know if sections tags truly matter, I do know that if you have them and they are ordered wrong by accident it can cause issues. So, that is an argument as to why not to use the section tags. I'm stubborn, I use the tags, plus it allows me to put the Tempo Change and vocal changes next to the section title and be way less likely to be sung.
When you have songs like that keep wanting to repeat, but you want to keep the structure. Put effect symbols around the last few words or even the entire line you get the point. See what happens. (...) or {...}. Many times you will get different effects the () tends to either bring in another vocal or make an effect on the current vocals or nothing, but the whole point of it is that when you use those symbols you are also giving suno more permission to skip the lyrics that would mess up its flow. I use that to find any offending parts. Then adjust based on what was skipped. That is just my add to something already explained. Also {....} tend to get skipped more often than (...)
I would put this stuff in workflows but I am not sure if my advice is workflow worthy...lol. mine are more tidbits.
@worn flicker @weary wharf You guys helped me a ton! I have no idea why the asterisks were causing the weird looping, but since I've edited them to [] the instructions are now really implemented almost every attempt and the bugs reduced in frequency from "almost always" to "1/10 times"
You are real nerve savers, because I started to burn out at a horrifying pace before this.
I only wish most of these basic instructions were on the site itself somewhere
AI is test and learn experience... It's hard for the Suno team to test these as best as the users.
So, this channel and here we are 😄
Clip got stuck
Np.
Lyrics contents. See if you have any copyright things or profanity. Check that first. See if it fixes.
I can only tell you that a ton of people have come in here with the same issue and it was multiple different errors yet the Lyrics ended up being the thing doing it. Maybe someone else has other ideas...sorry I couldn't offer more assistance.
We don't work for Suno, I don't know if you are aware of that we are just trying to help. I hope the best for you.
I know thanks again
Is anyone having an issue with getting "a clip got stuck" errors? Can't get anything to generate.
We've noticed that message is coming up in place of the "flagged for moderation" message -- either something in the style prompts, lyrics, or perhaps a particular persona (if used) could be the trigger.
What are you using as input?
Try the lyrics without the persona -- I've noticed that the inclusion of a persona could be the issue sometimes -- not sure if that's because the song the persona was based on might be triggering something with a seemingly reworked automoderation system -- a lot of "I dunno"s at this point, but we've been trying to make sense of these weirdnesses
It was doing it last night too though when I was trying a different, less edgy song as well. That one it let me make a bunch of times then just kind of said no after awhile.
The "clip stuck" error is one of about three reasons, it seems. But a major one has been the censor filter has been more sensitive to content of various measures which would go against the Terms of Service. Nearly everyone says if they let Suno write the lyrics, the problem clears up. It's is suggested to try to see if any clears the problem: • • use other lyrics
• test just a small generic phrase like "this is a test"
• Contact support@suno.com and link them your old song if you're working off a previous one.
• paste lyrics into chatgpt and requesting it review for content that may be an issue with Suno
• try any of the additional tests offered 🫴
The other reasons may be caused by recent updates, a greater influx of new IPhone owner members as well as a roll out of an improved version of Suno.
I kind of figured it was glitchy because of new features and stuff. Seems like it happens Everytime they have introduced a new beta feature. It gets wonky for about a week.
But ok cool thanks for the tips I'll report back when I have time to play with it again
Sounds like " different and less edgy" can at least set a benchmark. But using chatgpt may be a quick source for reviewing lyrics immediately after an error message.
I'm just a member here, but I'm always curious what the fix is so I can be aware of what to do.
Word. I may just have to see if I can take that particular song somewhere else to get worked on (not willing to butcher the lyrics for TOS). That's if it doesn't let me run it through suno in the next few days.
I get it. A song is a sculpture and your personal artwork. It's harder than you think, right.
As for anyone hearing the "pg-13" version, they will definitely still be impressed since they wont know the difference.
I've got some songs that just end up on paper or I record with my keyboard or guitar cause I know they won't make it online. Lol. So I write version B so I can get the music.
Keep it real, then, mate. Good luck. Let us know.
Maybe the upcoming maintenence is a sign of v4 rollout
Yeah I feel you. I have a box of notebooks with songs, bits of never finished songs, poems, short stories, all kinds of stuff. Have you released anything that was made with your keyboard and guitar skills? I used to play bass years ago, never have time to now because work kids and wife.
anyone else get timeouts on video regeneration?
I don't call myself a musician, so I write a song, working out the progression, and few fill ins. Before I'd send a worktape and let someone demo, then I'd teach myself how to play the song I wrote. Haha.
Lol hell yeah. I went the "just pick the thing up and make it make the sound" route. Worked out for the most part.
Hello, I'm trying to reach Nedi, Senior Research Scientist. This is Damian Kramer
Well my argument is are people who use computers to create art, artists? Even a songwriter is still an artist and if you can create music from your own original writing and edits, I'd still say your a musician.
But I have used stems or apps to remove vocals and added my own vocals lol
Has anyone upgraded their subscription using the proration? I'm finding it confusing how they calculate the rebate. Does your month of premium begin on the date you upgrade when you select proration? (As opposed to waiting until your renewal date.)
I think...they take the amount off your bill based on credits you had left, something of that nature. I did it recently and I was not upset by the result, I had instant access to my credits with the proper money in exchange.
Unless I was the dude that released No Man's Sky after showing that preview I would never give myself or team an hour and broadcast it. It is almost a karma thing...lol
lol i know, some one had to do it, so i thought, hey, why not me
Oh I am sure the next problem is all of us addicted people are going to keep refreshing and overload the login server...lol
Which is 3rd party so that says something
i may or may not be pressing the refresh every 60 seconds right now
Lol. I definitely have not started hitting refresh at 10:58
that and we all all trying to log in at the same time
Yeah that is what I meant
A single node comes back up Everyone: "I'm in! But... not?" Every dev ever: "Wait, no! Not yet! There are more nodes restarting!" 😂
All us addicts
what else am I supposed to do
I run my singer-songwriter framework every Monday and Thursday night... of course maintenance had to be scheduled for a Thursday night 😝
I used to use Daz 3D for drawing references hah
visual novel
Could just be me, but... that reminds me of Life Is Strange, stylistically. Which is a good thing!
its one of the few "sfw" images I am working on that I could show, lmao
what ever pays the mortgage right?
I got to get rid of all those freaking stock reference photos. Now that takes some jobs away
So easy just to generate
It does look like life is strange and maybe a little Rain
no shit, with ai, you can have a stock photo of wht ever you want in about 10 seconds
Yep and with the controls you just pose them like persona
yep, to bad i still aint found a good path from render to ai enhance