#๐โsuno-school
1 messages ยท Page 41 of 1
where can i use the created song in suno with premier plan subscription?
Welp... I'm guessing the lyrics I have written up are being interpreted as being some sort of darkwave pop song, because even putting in JUST "medieval, melancholic" as the style, I keep getting something that has a dark synthwave pop vibe with very modern instruments. I'm giving up for now. I'll attempt attacking it later, when I feel like burning through another 100 credits...
Any ideas which plugins use to make it sound good?
Can you elaborate on what you mean by "sound good"?
Remove hissing and robotic distortion
In second half esp
seems like V3.5 often tries to extend song lenght to 4:00 even if text fits for 0:50(-2:10 with chorus repeat)
lmao-- it might actually work
hahah
Mind you, this was at the VERY end of a musical expression, but 3/4 ended exactly as I would have predicted they would. The one-off is a reasonable extension, but I wonder how well this actually works in other situations
that 1:00 clip must be special
It's not without it's charm.
Not as effective, yet one did work, tho it ends as abruptly as if it had gone over the limit. I did use it from the 12 second mark of a song, tho... so that's not unimpressive. It gave me a 15 second song after combining. 
... Getting 3.5 to stop after 15 seconds is noteworthy, for sure. โ๏ธ ๐
Hrm, not a tidy resolution, it just sort of stopped. But still interesting that it led to the generation being cut short & refunded
Yea. I'm just surprised that it did it
Hrrm. Is this the same 3.5 we've been using?
I sometimes leave the lyrics box blank when I want to see what Suno will do on its own. This time, it apparently decided that it was done with the song:
yes. lol.
Is there a prompt for Breathy Vocals that i can hear
Has anyone noticed a gradual decline in sound quality from v3.5 since it's release or am I just crazy?
Have you tried "Breathy Vocals"?
it can have noise in the tracks I noticed
yeah I'm getting some high pitched noise as well
it's not ideal that is for sure
I've tried to remove it with Audacity (but it's my first time using it so I don't really know what I'm doing) but it seems like it's too integrated with the music to do so
https://suno.com/playlist/ca526d7f-8abb-482a-b398-68c7bcca7981 some of these might work
Suno does seem to respond decently to Breathy Vocals . I just pasted your text for the style ๐ https://suno.com/song/e126d0a0-97ae-4187-a2c0-72e7873cd70f
Breathy Vocals that i can hear song. Listen and make your own with Suno.
The audio quality varies, but should not degrade since the release of a model (e.g. v3.5) because the model itself does not change and that's basically what determines the audio that generates. The current models all have some degree of noise in them unfortunately, it can vary by prompt/genre or simply be random.
Any more breathy and she'll pass out https://suno.com/song/8009cafa-a479-4b08-ba4b-3b7f3908e0a2
Eerie, Haunted, Ominous, Breathy Vocals, Harmonics, Feedback, Rhodes Piano song. Listen and make your own with Suno.
Thanks, I guess I just got some really lucky generations when I first started using suno!
That plus when you first start using it the reaction tends to be "omg wow this is the most amazing thing ever" and then a bit later on you go "hold on, why does this sound so bad?" - I went back to my earlier v3 songs and realised that the sound quality is actually worse than I thought it was at the time.
Yeah there's probably a fair amount of that too, as I was blown away when I started using this and couldn't believe it was real!
It's still early days, pretty exciting times to be experiencing this even with its limitations.
Would you think in futurein suno we be able to set style % . like for example dnb 50% vocal 30% jazz 80% .... and so on ๐พ
Is there a way to change some of the lyrics but keep the same melody, tune, etc? I'm used to Udio so there are a lot of things I feel like I am missing.
so when you use the inspiration prompt is it legit to put an artist as the inspiration, i.e. Beattles Inspiration, etc.. ?
a capella scat nigun, nigun scat, a capella doowop nigun scat, a capella song. Listen and make your own with Suno.
for a capella people
the core concept is one of my only "rules" of AI, as its also applicable to humans
namely: if you want comprehension, you need to restate things in several different ways in order to smooth out inconsistency in your target
tangentially, this is why you'll often see in my prompts totally incompatible genres, essentially, since I often use this to help temper overfit tendencies in say, 3.5, to certain types of progression
but a capella is quite easy in 3.5 surprisingly, I just saw a reddit post and figured I'd give it a shot.
One message removed from a suspended account.
https://suno.com/song/7f9c84ea-33fb-4bd4-ad11-f697083a44a9
and example of a capella with lyrics
a capella scat nigun, nigun scat, a capella doowop nigun scat, a capella harmony, harmonizing scat nigun a capella song. Listen and make your own with Suno.
A song i just created went to 7:04 before it ended. then when i go back and play it, it stops at 4:00
hello.. quick question. If I make a track public, does that mean it is no longer mine even if the lyrics are original?
One message removed from a suspended account.
One message removed from a suspended account.
One message removed from a suspended account.
One message removed from a suspended account.
This appears to be a v3 extension (looking at the length and going by the audio quality.)
v3 has a tendency to get quite noisy, but you can work around it by extending from a time just before the quality gets worse - or try it with v3.5 (again, extending from before the quality degrades.) Once the quality starts to go bad, extending from a time after that will carry the bad quality across into the new clip.
Making a song public just means other people can listen to it. It doesn't take away any rights you have over the lyrics (and, if the song is made during a paid subscription, the song is also still yours.)
One message removed from a suspended account.
One message removed from a suspended account.
One message removed from a suspended account.
One message removed from a suspended account.
if u make a song public it'll still belong to u, and it will show that it was made by u
The website went down I'm in Brazil I wanna know what happened I think it got error 525, or error 524
A known issue with Cloud Flare atm. Try refreshing again in a minute or two, or better yet, use a VPN since it is likely a DNS issue: https://discord.com/channels/1069381916492562582/1252384524319326310
it's like that here too
are there any tags to use to force a song to end earlier? or am I doomed to have to download and edit?
Extend where you want to end with the extension set to v3.0 when using ending tags (e.g. [end] in lyric box and "end, short" as the style tag) as 3.0 is much more reliable for ending than 3.5
Awesome, thank you. I didn't think to try an older version. Gave up after 3 extension attempts burned the full 2 minutes when the only content was [end]
Hi there!
I have a problem... I've create a really good song, but need to change a few words, because it's playing in a non propriate way... How can I do this? When I reuse prompt - it gives me a new song, but I just need to change one word in that song which I already have... is it possible to do this?
Not unless it's at the end of the song (or you don't mind redoing anything that is after the word)
will it be possible in future maybe? it's really strange that it's impossible to change... you get almost the perfect song, but need to change only one word in text and it's impossible...
It might be possible, but I'm not apart of the team so idk
But it happens, that's just what Suno is like
so i tryed to make a song (extended) off a sound clip i uploaded but i got this message, when trying to make it public
follow up question ( i am on subscription) am i allowed to up load songs to YT that are not public on Suno?
If an uploaded clip was detected to have a voice (or got flagged as having one with a false positive), public sharing is disabled in case the voice belong to a third party and the clip was used without their consent.
You can still share the link though and download the video or audio (just remember you can be held liable for any consequences from the audio clip content used so be sure to do your due diligence)
i should be fine since it was just me yelling Shooo wiii into the mic for like 6 sec right?
If the song is not yours, no. You would be basically stealing (as the original creator would have at least the lyrics copyrighted)
Covered in the FAQ (Suno's Help page).
The free plan songs can be posted the songs on free social media (e.g. YouTube) but you must credit Suno (e.g. "Made using Suno.com v3.5" or variants as simple as "Made with Suno" in the description). You cannot monetize it either.
Only songs generated during the subscription plan can be commercialized.
https://media.discordapp.net/attachments/1102676217687195679/1248122915774533693/image0.jpg
Yep, that would be fine since it is your own voice thus your own consent and material
thank you
No you are not crazy, I feel the same
Maybe it will help the support with the mixing check of a professional mixing and mastering engineer... Here feedback to my New Track for mixing this Track
[08:47]
Here's a report that Mix Check Studio has made about your mixed track.
Your track seems a bit wide in the stereo field. This could be an artistic choice, however if this is not intentional, here is a tip on how to ensure your stereo field is balanced: Be mindful of stereo enhancers, as overdoing it can result in a โhollowโ mix. Play around with the settings until you reach a wide yet balanced result. Try widening specific frequencies or instead of applying stereo enhancers on the whole song, experiment with using stereo width to enhance certain parts of your song. For example, you can use a wider stereo image during the chorus and a more narrow stereo image during verses.
It looks like your mix might be too loud to be sent for mastering. Here's a tip on how you can improve that: If you have applied any limiting on the master channel during the mixing process, consider removing or reducing it to create more headroom. While limiting can be useful for previewing how your mix will sound after mastering, it's best to leave this processing to the mastering engineer for optimal results.
Remember that these are just suggestions based on our analysis and shouldn't be considered the final say. Always listen to your mixes/masters critically and trust your ears.
an here is the Feedback to Mastering the Track....
Here's a report that Mix Check Studio has made about your mastered track.
-
Your track seems a bit wide in the stereo field. This could be an artistic choice, however if this is not intentional, here is a tip on how to ensure your stereo field is balanced: Be mindful of stereo enhancers, as overdoing it can result in a โhollowโ mix. Play around with the settings until you reach a wide yet balanced result. Try widening specific frequencies or instead of applying stereo enhancers on the whole song, experiment with using stereo width to enhance certain parts of your song. For example, you can use a wider stereo image during the chorus and a more narrow stereo image during verses.
-
It looks like your mastered track will be played 1.0 dB louder on music streaming services such as Spotify, Youtube and Tidal. It will be played 1.0 db quieter on Apple Music. Music streaming services do this in an attempt to maintain uniform loudness for every track across their platform.
Remember that these are just suggestions based on our analysis and shouldn't be considered the final say. Always listen to your mixes/masters critically and trust your ears.
I'm noticing that the audio quality seems to get worse every time you extend a song. Any way to combat that?
https://suno.com/song/0ae696fb-f5ad-4bc0-ab71-5f90a847265e the analyse is from this Song
german schlager dance song. Listen and make your own with Suno.
I usually monitor with headphones, but leave on my studio monitors. I can hear that your song is a bit unbalanced with the balance over to the right a bit, but it's not unacceptable.
Can I publish video files from Suno with my songs at my social media? Shouldnโt have any problem with copyright?
Would anyone know if there is an easier way to organize yor music than just playlists? Even just a 'bulk delete' option? is that possible?
There are some tips and tricks for this, e.g. changing the point you extend from before any distortion kicks in. But there's a certain loss of quality you can't get away from. It's frutsrating, as sometimes you need to extend to finish a song with a big chorus or solo and you can tell the audio is rougher than earlier in the song. Although it's possible that not everyone will notice this, especially depending on the headphones they're using. Speaking of which...
I haven't seen this mentioned before, might be obvious to some, but a tip from me: when generating new clips, listen on the very best headphones you have. Sometimes I've worked on a song late at night with some mid-range ear buds and think a track sounds good, but when listening back the next day on my best headphones I can pick out annoying distortions and artefacts in the background.
Toppers ambient house, melodic breakdowns, novation bass station, oscillation, pads, euphoric chord sequence, driving song. Listen and make your own with Suno.
You already posted this a minute before, in the #๐ฌโgeneral-chat channel and I'm not sure why it's being posted here as well?
You can extend the song and find the new section with the right lyrics and edit the song outside of suno.
It's technical.
Can't reach everyone in one channel, especially since technical people are in general channels less frequently.
This is "technical discussion"
A Function is a technical discussion.
Technical discussion between people that use Suno.
It's not a channel for just posting songs about technical things (unless the song is related in some way to a discussion that is taking place at the time, to demonstrate a technique for example.)
๐ฆ oh mb
is there any prompt to get better sound quality ?
i did use the extend feature and the result has a pretty poor quality
thanks for any return
Anyone know if Suno v4 will include some vocal improvements to bring this on par with Udio. The musical creativity of Suno is far superior compared to Udio but the vocals are absolute garbage in comparison. Im just blowing through credits now to finally get a track where the vocals dont sound like they were recorded with a cheap microphone from the 1980s. Ive heard people say Suno V3 alpha had crystal clear vocals but the full launch of v3 turned them into these whispy and distorted vocals. Surely this is a priority for them but have the devs spoken about it at all yet?
Anyone know how to tell Suno to make sound effects? An example that I attempted, but it did not work. My example: "Doors are slowly creaking, footsteps on the floor,
Every time I close my eyes, thereโs something more.
Caught in this illusion, trapped within the night,
Searching for an exit, but thereโs no end in sight. [sfx: creaking door] [sfx: light footsteps]"
v3 alpha (no longer available) is often held in high regard with regard to its vocal quality, compared to the released v3.
Use v3.5 for better vocals (it is much closer to v3 alpha's vocal quality and is partly why this model even exists.)
No idea as to what improvements to quality are coming with v4, the devs have been very quiet about this. I expect they are aiming to be a strong competitor though and quality is an issue that people talk about very often here. So it's probably high on their priority list.
fingers crossed it is something they tackle with v4. ive been creating a verse/chorus with suno and then importing it to udio and using inpainting to fix the vocals and complete the song there but its annoying that it needs to be done. Suno is so close to perfection right now, just the vocals letting it down
I've experimented with doing this as well (though with "extend" to make it re-perform part of the song). It seems to work better than trying to use a dedicated service for replacing the vocals.
Is there a user guide of some kind on Suno about adding sound effects/sfx? I know Suno can do it. Suno did it unintentionally to one of the songs and I love it. It would be great to know how I can tell Suno to do it.
I haven't had much luck with sound effects. I think there's a gunshot sound I got once with [gunshot fx] or something like that but I've cycled through animal sounds, doors, footsteps etc. and none of that worked.
the unintentionally sfx was an undead evil laughter with reverb and breathless. very creepy and i like it for a scary song. But the song that it happened on was an quirkcore genre.
I've heard a legit helicopter in someone else's track. I've had laughter, beeps & dial tones, crashes like pots and pans toppling over, a cat fight, gongs, and a few others. None repeatable except maybe for the beeps and even with those - they went where Suno wanted to put them. 
Experimental Samples, Audio Samples, Humor, Polyrhytmic Glitch, SFX, Acapella, Computer SFX, Sound Effects, song. Listen and make your own with Suno.
Oh yeah just remembered there was a tornado warning siren once in someones song.
THAT would be cool
well... depending upon the rest of the tune. heh
If I recall, the helicopter was in one with nothing in the lyrics box and a very short story in the style box. ๐
Helicopter: #๐ชฒโbugs-web message
You could try to do extend and start from 00:05 into the song or right before the first work you want to change
are they your songs?
I got free samples of air-raid siren and remixed it in audcity to pan
then used it as an attention / song starter....
https://suno.com/song/75f1279b-46a6-4a6d-906b-206c9179b8ca air-raid siren upload into song. that was being referenced.
Alternative Rock, Heavy Metal, Hip Hop, Electronica, Pop elements, lamenting, deep mellow tone, Cinematic, Melodic, song. Listen and make your own with Suno.
Noice! ๐
Almost coaxed one out of Suno. B+ for effort! https://suno.com/song/a3214686-5ef8-4178-8aba-b6f9d37a9be6
Sound Effects, Horror Score, Air Raid Siren, Eerie, Ominous, Air Raid SFX, Horror Score, Siren, Not Dave song. Listen and make your own with Suno.
How do I play musical notes? What would be the tags to play the desired composition of notes, the improvisation, the chords, the scales, the elements of the song??
Scenario - I start on free trial, generate a song and I like it, so I decide to commence a paid plan - can I keep the song created during the trial as mine? If not, then what if I extended the song, is the revised song then mine?
You'd need to contact support@suno.com and explain that you created a song on the free plan that you would like to acquire full rights to.
As far as I am aware, extending a song from the free plan doesn't change this. There is a hidden watermark in the song - no details available on how this works or the info it contains but this will likely continue into any extensions you create, so if it identifies the song as a "free" song, extending it may still retain that same watermark.
click a track and then hold shift to select more tracks, then click the three dots on the right and move to trash, and it'll delete all selected tracks
Has this been discussed elsewhere? I wasn't aware there was a process for requesting songs generated by free accounts, I figured they were just the property of Suno moving forward.
...still would love any official comment on extending free songs / songs generated by others, though.
The question comes up from time to time. They do take a while to respond. I am aware of at least one occasion where the request was granted but I don't know the details of what it involved (e.g. if it cost anything.)
I know they're still looking for a community manager, but it would be helpful for guides such as yourself to have a dedicated team member to contact for clarifications on that kind of stuff, so some stuff isn't left to conjecture or educated guesses.
Thank you, I think an important factor as well would probably be if the song had made public or not
Being set to public only effects whether it's searchable within the social media side of Suno.com
Oh my understanding is that if you make a song public, even with your own lyrics, it is now in the public domain and others can use it?
If you have a paid account, you still have general commercial ownership over the music you create, regardless of whether it's set as public or private.
Unless it's been discussed and I've missed it, there hasn't been any official comment regarding ownership or splits of any kind if you extend someone else's song.
I saw someone ask if they create a song with a free account, then later extend it with a paid account, and the primary focus of the response was making sure that the person wasn't making use of multiple accounts, as that would violate ToS.
So whether you're extending a free song, or someone else's hard work, I don't know one way or another.
Yes but what if you create a song on a free plan and make it public, that would set it to the public domain I believe?
Nope, not public domain. Suno owns it.
Public / Private relates to whether the song or playlist can be found from the Search tab.
Ok thanks for your help
Suffice to say, if you have a paid account and are uncomfortable with people being allowed to make modifications to your song & have a third party sign off on "their ownership", keep your songs set as Private and treat it like the master recording.
Perhaps they should give the free account people an option to sign up and gain ownership of the song they just created for a short period, that would be more in the spirit of a trial I think and also good marketing, showing confidence in their product
The only advantage to making your song public is if you 1) care about popularity within that social media platform, 2) want to actively seek collaboration for the sake of seeing your song reimagined through other people's ears, or 3) are trying to get a piece of that Summer of Suno cash.
There's already a MASSIVE problem with bots taking advantage of the free service, that would just tank signups from people looking to monetize the songs...
IMO this, or a button to "buy" a song from the free plan for a number of credits, after subscribing.
Yep sounds good, sign up to buy this song for xx credits would be a great option.
Create free account > create and "purchase" as many songs as base credits will allow per day > repeat on as many free accounts as possible
Hey, maybe a stupid question, but using a screenreader and am not sure, what do those 3 numbers on my profile state? Like there's one after the other, but they don't have any text, just numbers and probably some graphics, can anyone tell me pls?
There is no official API at this time.
Any unofficial APIs are basically hacks and it's against the Terms of Service to use them.
Also consider that these services may require login credentials and payment, which could be mis-used.
This sounds like the three numbers are:
Followers, Likes, and Plays
Amazing, thanks for that!
This is a math rock style song, can you give me some advice?https://suno.com/song/d6a6d8ae-bfe7-4746-ad46-2e5461869dd6
emotional complex drums midwestern emo prominent bass math rock dynamic song. Listen and make your own with Suno.
Actually I think I mixed the numbers up. It goes:
Followers
Plays
Likes
In that order
Haha thanks, been thinking since if you were right the first time, I'd have more likes than plays ๐
Who owns the rights to Instrumental music? Is it possible?
Are you asking if composers are able to copyright their compositions? It's absolutely possible to copyright music.
In terms of Suno, being instrumental or including lyrics doesn't change the ownership, as long as it was generated using a paid plan.
Composer and the performer.
So you can perform fur elise and try and sell it as the performer
People can't use your performance without your permission
The date for copyright on that music is long passed.
Tho probably varies from country to country
why cant I get a violin to play? I'm losing it over here.
Ive tried putting violin in the instructions and style and even both several times and still no violin
Hello .since yesterday all my songs are muted. Only after I download the file I can play it back. It does not work on neither phone or desktop. Can someone help me?
What instruments currently run on ()[] tags? all ?
Does SUNO have plans to generate a midi file in addition to WAV?
guys i can't find how to do sound effect it doesn't work like i would like. for example i would like to do a music without any rythm at the start but only the sound of waves and so i write this : [Intro]
[Sound of waves crashing on the shore, gradually building] but it ignore it and start music
it would be great to be able to download the music with or without the voices
Already in the suggestion forum https://discord.com/channels/1069381916492562582/1253172964728377455 - vote to show the devs how much people would like that ability
is there a way to limit the time of a song so that it doesn't just cut suddenly
Well, you can use ending tags to cut it off (using 2.0 or 3.0, 3.5 is not great at ending).
What you want though is use Extend to continue the song:
https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
If you did not know:
v3.0 base time = 2 min max
v3.0 extension = 1 min max each
v3.5 base time = 4 min max
v3.5 extension = 2 min max each
excuse me but, before saying its a bug, did anyone know why one of my song got removed without i remove it my self ?
my bad, it goes worst
semi-related to this thread back some days ago, today i tried to use 3rd party stem separation on v3 (not clean though) and fed a Golden Seed voice as an audio upload - 6 generations in and v3.5 still doesn't quite capture the crunchy nature of the golden seed v3 voice - no examples unless i method that actually seems good
Hello , after I upload a clip and I use it to create a song, say parts 1 , 2 and 3.. with number 1 being the original clip I uploaded. How do I kill number 1 and just get to 2 and 3 as the final song. when I choose "get whole song" it always includes the clip.
Website hasn't been functioning for me all morning. I'm on my phone and it really doesn't work good on mobile
It won't even just play music, yet other streaming services are 100
Is it possible to make.
that my points are not reset every month? Because I may not come in for several months. But I want to get all my points. And they do not accumulate, but simply disappear. Although I purchased the annual max subscription and it's depressing me 
"Use it or lose it"-- it states at the bottom of the subscriptions page that credits don't carry over from month to month. Like most subscription based services, you don't get any reimbursement for not having used the service that you paid for. That's like calling Netflix and asking to get some months for free because you were paying for it but didn't watch anything for a couple months.
Instead of paying month to month and choosing to only pay for the service when you want to use it, you paid for the entire year's subscription worth at once to get a cheaper rate for buying multiple months in bulk. It doesn't mean that you're entitled to accumulate 12 months worth of credits, unfortunately.
i paid 4$ for 500 credits but still 0. Please fix
Contact billing@suno.com for any billing associated issues.
And I really do not trust opening foreign PDF fiels. And you should not be providing one of it contains your financial info
WAY to easy to embed nasty stuff in a .pdf.
Is there no one to help with tech issues here?
Are you referring to the question you asked 5 minutes ago?
If I downgrade my subscription, will you refund the difference? Or not?
I believe you get a pro-rated return for the months you didn't use, but I'm not on staff. You'll have to read the fine print from the subscription page, or email billing@suno.com for more information. โค๏ธ
Why need pay for Netflix if everything is in English ๐
Well, technically this is a channel to talk about technical aspects.
Technical issue is more of a #๐ชฒโbugs-web thing.
General recommendations for this issue are:
- Try clearing your cache or perform a cache-bypass reload ( #๐โsuno-school message ) in case it was a cache issue
- Try using another browser to see if it is a browser issue
- Try using a VPN if it is a ISP or DNS issue
Question, I upgraded to pro but the credits never were added to my acount? i have the receipt and everything
Reload the page first if you haven't done so yet in case it is a visual glitch.
If you still did not receive your credits, contact billing@suno.com
Yeah, i've reloaded, cleared cache, etc. and nothing. I'll contact billing now
wasn't sure if this was a common issue
Thank you for the advice. I pay as a professional customer so thought with all the admin, team and mods would answer. It is my fist time having to ask for help here so I am not aware of what channel is for what.
was hoping it was even a visual glitch and tried to generate and got nothing
For future reference, all help and advice here is offered on a volunteer basis.
They haven't hired a community manager to spearhead things yet, and the team is largely focused with the day-to-day operations of making Suno as best as it can be (though they occasionally interact, mostly in the Bugs section if it's a breaking issue). The moderators are responsible for keeping the server in order, but technical advice largely falls onto the shoulders of the users and the few dedicated guides.
Much appreciated
Are there any audio technicians online?
The fadeout in the bassline at the very end of the song has an asymmetrical waveform, and it causes weird audio artefacts. I've already tried to fix the DC offset, so it's not a matter of that-- the rest of the song is nicely centered on 0, but it's just this weird section.
Before I started taking wild guesses & throwing different filter passes at it, I figured I'd check if there's any tribal wisdom that could be passed on โค๏ธ Thanks for reading & thanks in advance for any help or insight anyone can share ๐
I wouldn't expect the asymmetry to cause any audible effect. I've had audio recordings before where the offset is totally off in one direction.
The issue is the notable dropoff just to the left of the highlighted section, you can audibly catch the dropoff in amplitude ๐ฆ
Ohh right so you've just highlighted the bit that is problematic, and that's before any edits were made to it. Hmm.
Is it just bass? No other instruments? And a single note?
I fixed it as best I could with some quick & dirty envelope edited, but the issue with that is that it edits both sides of the waveform equally, so while I'm able to fill in the bottom section, it pushes the higher end louder than I'd want and makes it susceptible to clipping.
I'm not really familiar with the envelope tool to know if I can edit just one half of the waveform? Oh my god that would be so much easier now that I'm typing it out. There has to be a way to do that.
Yeah, just bass, single note that swells and fades, but I'd love to edit what's already there instead of having to somehow match the timbre.
There's other instruments, but this is just the bass stem
God damnit I love this hamster
I'd try copying what comes after it (plus a little extra for padding), put it on a new track underneath, move it under the problematic part, slap a "fade in" and "fade out" at either side of it and adjust the level to try to patch it over.
(if it's just the bass stem)
I had a glitchy note in a Suno generation a while ago that annoyed me so I ended up doing a transplant of it with another note nearby
If nothing else comes from my poking and prodding (beyond a marginally expanded understanding of some of these tools), you've given me the idea to copy that section, invert it, overlay those tracks, and simply pretend like nothing ever happened.
I don't mess much with the effects in Audacity, I take a bit of a caveman approach with it usually and just slam bits of audio all over the place until it sounds reasonable ๐
Presuming invert works the way I anticipate, but I'll burn that bridge when I get to it.
You might not even need to invert
Oh yeah, any self respecting audio engineer would have torn their face off watching me use the envelope tool.
If you've done a stem split then you at least won't be causing damage to the other parts and I think a long bass note is probably going to be less tricky to patch up
I'm on such an adrenaline high I'll figure it out by the time I go to sleep.
I really botched my first song when I was getting back into the swing of mixing & mastering, but was getting enough positive feedback that I released it against my better judgement.
I decided to take another stab at it, and even just punching some of the drum accents with minor amplification takes it from good to great.
Really neat that Suno is helping me feel like a musician again beyond the act of taking a collaborative role along side it to create the songs.
It's both distracting and inspiring me ๐
Hi! I want one word of my verse to be pronounced in German, while the rest to stay in English.
Example: "like some [German:Kinder] surprise" ?Anyone knows how to single out that one word and make the ai pronounce it in another language?
You could spell it out phonetically (kinn derr), or less ideally, create an extension specifically for the separate language.
Success, a couple low filter passes to level everything out, then some normalization & using the draw tool to fix any hiccups
Heck, I tend to write English words phonetically when I care how it's delivered. Especially since I love British pronunciations on several words or just to make them rhyme better. Phonetic spellings are where it's at.
@charred merlin ๐ซก you rock! Always helping others โค๏ธ
Ugh, now I'm just wrestling with trying to remove little "pops" and "cracks" introduced by the draw tool. Repair seems to work SOMETIMES, Click Removal doesn't seem to help because it doesn't detect the "cracks" as readily as a click track.
Any idea how I can purchase membership again? My membership is valid until the 29th but I have 5 credits left now. I would like to buy the subscription again earlier but I can't find the option. ๐ฆ
you need to buy credits if you want to continue using suno until it refreshes
Yeah, but is over price...
it is what it is, AFAIK you are allowed to use one free account to use the 50 free credits per day until that is implemented for paying members that run out of credits
Extra credits don't expire. If you're a pro member, you can upgrade your membership to premium, or you can wait until the 29th around the time you signed up.
I'm not sure if the ToS has any issues with you purchasing multiple plans... if not, it would just be a matter of signing out of your current account, creating a secondary account & following the normal subscription methods for that account.
I got a technical question tho, I have massive issues uploading to samples to suno atm, is this widespread or only a eu problem right now?
upload speeds are super slow, like 0,5mbps
#๐ชฒโbugs-web would probably be a better place, as it's not something anyone can really help with
true, thanks mate
Fix IPad custom playlist navigation
I dont get free credits. I dont know why. I payed from 3 months, when credits are going, i cannot use for free suno.
50 credits * 30 days == 1500 credits.
When you sign up for pro, you get 2500 credits-- all of your monthly credits, plus a bonus.
you are allowed to use one free acount next to your paid account to use the free credits you get each day
I'm pretty sure making multiple accounts to take advantage of free credits is against ToS.
I need to find it, but this was posted by someone higher up in here that they are working on giving paying members the 50 free credits each day and until that is implemented you are allowed to use one more account to use these 50 free credits per day
https://www.reddit.com/r/SunoAI/comments/1dduptr/psa_buying_a_suno_sub_halts_your_free_daily/ top comment from the discord mod there
*reddit mod
is there somebody who also has issues with the website? it's not loading for me atm
Jusst loaded it. Seems fine for me here.
I did have to log in again though.
What error are you getting, 503? Cannot connect with cloud flare logo present?
Hello everyone, is there any document showing which commands we can use in [brackets], like [verse], [chorus], [bridge] and so onโฆ?
I explore what other ppl made so far to learn how to command, is this the way one should do?
Be sure to check out the associated entries
Thanks a lot ๐ค๐ผ
Hey what can cause an extension to without any intention of, produce like vulgar content , I asked in General but it's like a frenzy of delusional pre Madonnas feeding the egos of like 8 actually interesting people who are like mods I guess.. well I mean they have purpose I suppose.. but https://suno.com/song/40104316-c6e4-4c3e-b332-302264e798ef
Lord Draumus/Nick Donithan /Make it Real, Female Voice, Nick's Style
song. Listen and make your own with Suno.
Just a quick question, but how do you make Suno do a sea shanty? I'm using custom lyrics, but I put sea shanty as a style, along with upbeat, and using a cajon box drum as a main instrument, but it comes up with kids songs... How do I make it do a sea shanty? Any and all help is appreciated! ๐
Iโve seen people get good results with โEpic Sea Shanty Folk Adventure Classic Chant
Voyageโ in the style of music
could throw in some of these: Folk, Celtic, Maritime, Traditional, Acoustic, Nautical, Chant, Ballad, Pirate, Historical
"Gruff Pirate Voice" didn't work. lol
Folk, Celtic, Maritime, Traditional, Acoustic, Nautical, Chant, Ballad, Pirate, Historical, Gruff Pirate Vocals song. Listen and make your own with Suno.
If you are using 3.5, do not forget a dash for sea-shanty
I have a similar issue distinguishing between Spanish and Portuguese word. Does it help to spell the word phonetically?
If it does I am seriously considering leaarning the phonetic alphabet
Any tips for actually changing the musical style when using the Extend function?
@Mr.Tree ๐ณ(@MrTree) you know what would be nice? "cut" When u upload songs/sounds you can cut the part you want and it make a smoth start / end , If you know what I mean
(Orignial post by: SmOkZz)
anybody know how to consistently tell the AI in suno to make duets?
out of 40 songs generated only 1 was a perfect duet
please help vote on this: #๐ณโmini-polls message
Thanks! ๐
I tried to make some songs public last night for the first time, some that I generated then and there. I kept refreshing for quite a while to see if they were sent to the new section but it didn't seem as if they were, has anyone encountered this before? Thank you for your time!
I just asked Claude for the phonetic spelling of the word I needed, and it worked...
oh of course! x) stupid me x) "Learn Phonetic Alphabet"... I keep forgetting we live in the future ๐ Thank you! I will definitively use this โค๏ธ
yo
still credit card only... sad
So what has everyone been doing outside of suno for improving audio quality?
I rip to stems with RipX DAW Pro, get rid of the noise in the vocal track, as best I can then export and do what I can in my DAW to improve the tracks with mixing. In reality, I'm not doing any new songs until hopefully the quality improves with V4. Some genres are worse than others and unfortunately it's the genres that are the ones I need that are most affected!
What genres are affected you think?
Most of the instrumentals that I make are meditational/ambient and these are quite often bad. Country music seems affected, too.
Does extending a track to 15..20 minutes always end up incoherent?
This started pretty unlistenable (deliberate) but got worse as it went on. To the point I think it's a bug:
https://suno.com/song/3c532ae3-f9de-4f4e-a4d0-326ced85e74a
8-bit, acid, gabber song. Listen and make your own with Suno.
I've had that problem too. I figured it was just because I asked for the sort of music which might reasonably start with a big loud f-bomb. But that doesn't seem to be the case for you, there.
Also, this:
https://youtu.be/xm6tPMjp9WU
A phonetic explanation of how Ken seems to be dropping the F-bomb. Note: this is an update including the original low-quality audio from TikTok.
0:00 Intro
0:26 'B' and 'F' are pretty similar
2:18 'CK' or sound effect??
I have question how do we write in music that you want to play some part in specific octave and how to we differentiate between notes or accords and the duration of them and the pauzes between the notes or accords? For example if I would like to pass some melody of the Chorus to Suno, Could I write something like this? E#1/8,-1/4,G#1/8,-1/4,C#1/8 when playing straccato? 1/8 here means the duration of the note or 1/4 of the pauze (-). But from this suno cannot know how to play if accords or single notes (Arpeggio). And also, how to tell which octave to play in? Finally, is there also a shortcut for "Time signature"? Instead of typing Time signature 4/4, tempo 90 bpm, could I type something shorter like: 4/4, 90 bpm?
Hey guys, I created a bunch of songs but now every genre I try for some reason creates pop songs >( even when I write (gregorian chant) as the genre. What can I do?
<@&1108435550475653161> has problem when it tries do generate czech characters like: "Uฤรญme se ลกpanฤlลกtinu" there are these "ห" above c and s, pronouced as English [ch] and [sh]. But when it draws the diacritics into the video frames, it does not have correct position (horizontally and maybe vertically too). You can check it in new song (see: Stลeda, ฤtvrtek, dลevo): https://suno.com/song/08dffef9-6a45-4c7a-b7b0-9529468ef2d8
japanese traditional, koto, bossa nova, Clean voice, no reverb, distinct teacher's voice song. Listen and make your own with Suno.
Hey guys, I created a bunch of songs but now every genre I try for some reason once the singing start it switches to pop or rock >( even when I write gregorian chant or medieval as the genre. What can I do?
Did Suno's server crash?
i tried to make about 50 songs of phonk without sucess , could someone help me ? i can't find all the instrument needed to create phonk musics ?
also does suno knows how to make phonk and dark trap ?
"Hello, I have added credit to the wrong account. Is there a way to fix it?"
please write a mail to support@suno.com . they can fix that
does it also happen to you that you write a female voice and a male voice sings it? Or it goes from feminine to masculine and vice versa, in the course of a sentence... Any idea how to avoid this, please?
I created many songs for one lyric but i couldn't be successful at best vocal, any idea ?
thank you so much
Yes this is normal.
Why does it show I have 5 credits left when in reality I have to have 10 ๐
napsala jsem pรญsniฤku jak si chlap namlouvรก buchtu a vรฝsledek je jak z genderovรฉho ลกkolenรญ, dฤs ๐ฆ
Suno isn't loading up.... what's going on
Need more information. Is create doing nothing? Error 503? Cloudflqre showing hist failure? Etc.
It is likely this issue: https://discord.com/channels/1069381916492562582/1252384524319326310
Best solution: Use VPN if you have one
Otherwise try again in a few minutes
Aye it was a host error. Its returned to normal now lol
- Suno v3.5 = expert composer
What is the cleanest most accurate tool or service you have found to extract a vocal stem from a song? Thank you.
Personally, I've found the best results with Audacity + the OpenVino AI plugin. UVR has a lot of great options and vocal-specific models to choose from, but I hadn't found the specific model (or settings therefor) that cleanly separated the two, and occasionally the vocal track would blend with either the 'bass' or 'other' track depending on the tone. YMMV ๐ป
Could you rephrase your question or elaborate? I don't quite understand what you're asking.
Sounds like one of your generations had it's credits refunded, so you got the 5 credits attributed to that song back.
Generations cost 10 credits, and there's no way to generate a single song at the moment, so unless you upgrade to a paid membership or wait until the credits refresh, it's functionally the same as having 0.
Sometimes less is more when it comes to metatags / style prompts-- it's very easy for one to overpower the others and have too strong of an influence. I don't know if Suno supports accentuating a prompt with brackets, (eg. (prompt = 1.1), ((prompt = 1.21)) ), so one method would be to only use the required style prompts to get the initial sound/style of the song you're looking for, possibly with v3 as it's more versatile for different genres, then extending the rest of the song with all of the metatags & style prompts that you want to add onto that initial existing sound.
At 21m 22s, you're asking a lot. I'm not sure how you're selecting your extension points, but you have the ability to create the exact song you want. Are you just hitting "extend" at the very end of the generations and hoping for the best? Are you extending from "Get Whole Song" or just the previous extension? There's a lot to interpret, a lot of different things that can influence how a song will turn out, and it's hard to accurately advise without knowing more context.
important takeaway-- I think previously a moderator shared the fact that Suno has a 2m "context memory", so when it's making new sounds, it can only use the previous 2 minutes to understand what it should sound like.
For dark trap, try it as dark-trap (the dash treats it as one genre, without it, it will likely treat it as dark (emotion) + trap (genre))
IIRC, 3.5 forgot the genre phonk so the best thing to do for that is make the first part in 3.0 then Extend into 3.5 (1 min longer than 3.0 extension) like harber mentioned
Friiiiiiiidaaaaaaaaay
Toker, Paimon, Hamster; hope all is well โค๏ธ (and others, I just saw their names recently)
I got annoyed trying to make that bass fadeout sound nice last night, and just copied a single bass wave and overwrote any wonky sections, using the repair tool to smooth over any places where they didn't match up perfectly.
I would still fiddle with it, but my wife swears she's not able to hear the weirdness that I can.
I think she's low-key telling me to stop effing with it and just work on something new ๐คฃ
It's worse when they come in and tell you that unprompted... 
WHY DO YOU HATE MY GUITARS!
/glances at @flint haven
'was half expecting an allcaps lecture. lol
Yeah, I got stress sweats after sending a message with "effing" in it.
Could be worse. At least it's not a bridge.
i can't unread it!!
Someone report hamster for abusive or harmful targetted language.
@flint haven get them
I find that using UVR on extremely high-quality settings does a significant amount to improve audio quality. By high-quality, I'm talking about using the additional splitter settings to enable a pre-process model and a secondary model. This makes stem splitting take A LOT longer and sound A BIT better (enough to notice, but not miracle)
Plus you get to put the song back together in the DAW, selecting saturation, EQ, and compression settings on each track to taste.
Not sure if this would be the right channel, but I cant seem to find the info on the website, does anyone know how to get in contact with Suno's legal team I would like to investigate the possibility of using songs created with Suno in a game I am making
Hello.
I often find that while the generated music is interesting, there are sections in the middle that I would like to refine.
I would love to have a feature that allows users to select and regenerate specific parts of the generated song, especially the middle section.
For example, the current Extend feature allows users to specify only the start time (MM:SS), but it would be great if it could be expanded to also specify the end time (MM:SS).
Could you please let me know where I can find information about such requests or the status of new feature development?
This sounds like "in-painting" (where a section of a song gets re-generated.) This is coming in the next version (v4) but there's no info on when this will be available yet.
If you made the songs on a paid plan, you own them and can do whatever you'd like with them.
Subject to your compliance with the terms of this Agreement, if you are a user who has subscribed to the paid tier of the Service, Suno hereby assigns to you all of its right, title and interest in and to any Output owned by Suno and generated from Submissions made by you through the Service during the term of your paid-tier subscription.
This is covered in the FAQs: https://suno-ai.notion.site/FAQs-b72601b96de44e5cacd2cd6baa985448#f0f1441180484d6094206e84e334ba36
Basically if you are creating songs during a paid subscription, you're free to do whatever you want with those songs.
Other than that, I don't know of a legal team contact - I'd go with asking support@suno.com for further information if you need it.
Hello! Please email billing@suno.com if you are unable to find the answer to your question on this page.
They also say in a different part of the TOS, that the songs might contain copyrighted content and it's up to you not to infringe any IP rights with those songs
Thank yall!
Whoa
This is OpenAI's tokenizer, so it might be handling it differently, but it shows that there's clear differences in how it's interpreted.
I keep forgetting about the tokenizers x.x
https://platform.openai.com/tokenizer
Yeah, differences with the hyphenated version even between gpt3 and gpt3.5, so no telling how Suno would interpret it, but it's helpful to keep in mind even just to look at prompts from a different angle, especially when you're not getting the expected results.
Which reminds me, I think Toker mentioned the other day that negatives aren't really respected in style prompts, which explains why "no-bass" makes the track THUMP. ๐
Have there been any changes to the model in the last few days? I've noticed that I can't generate instrumental music now. Even when I toggle the instrumental option or add "[instrumental]" in the lyrics area, it still adds lyrics at random. Here's one example:
https://suno.com/song/29c9e7c6-1559-428d-a4b4-a65790b0ff0a
Musica Andina, Peru, Bolivia, tambores, zampoรฑa song. Listen and make your own with Suno.
Has anyone noticed that generations seem to carry some parts of the previous prompts? I like all types of music, and sometimes I shift my creations from one genre to a totally different genre. For example, I was creating some heavy metal songs with low, growly male vocals, and then, when I switched to creating electric pop, with female vocals, it still created the songs with low, growly male vocals.
It could be something in the styles that's driving it to belt out the lyrics. I'd probably try hyphenating Musica-Andina first and then maybe change or shuffle around some of the others
Yeah, I'm going to continue testing different styles and tags. I'm just wondering if others have had the same experience
Extensions use the previous clip for context as to how it should generate the proceeding content. 3.5 is a LOT better at this, so if you're trying to specifically distance yourself from the previous section it can be helpful to drop any style prompts or metatags that could have an overlap. Alternatively and more aggresively, you could extend Clip 1 (2 minutes long), use the first couple seconds of Clip 2, and then extending from that tiny clip to create Clip 3.
Without "Get Whole Song" and just giving a small clip to extend from, you have a better chance that the new extension will be a new "thematic" piece instead of being closely associated with the original segment.
Swap from bluegrass to chiptune ~1m
Is there a reason that the Video exports always use the original first graphic assigned to a song rather than one you added later?
it also appears that the song title in the video clip stays as the original title as well.
after i created a song, I changed the title & thumbnail image, but the video export shows both the original name as well as original graphic...
does anyone know how to only use the "get whole song" to just get two parts ?
Video is created with the song, doesn't regenerate it when you add a new picture. Perhaps you could mention it in #๐กโsuggestions-web
The video generations starts when the audio is ready. So your changes are not seen, a re-gen is needed for the video. Not available yet. Maybe in the future.
Not possible. Download the part wavs and mix in a daw or an audio software like audacity.
ugh that's what I was worried about.. @weary wharf thank you forletting me know
If you click "Get Whole Song" for part 4, it will grab all previous sections. Like the_builder suggested, download the separate parts, or I'd suggest getting the whole song from part 4 so it has a consistent audio quality (or as best as possible),and trimming the first part out rather than stitching multiple together
@charred merlin Thank you doing so now. Iw as going to submit this for the contest.. but I don't know i can submit something via youtube or whatever since i won't be able to put it back up on suno
where to vote ?
Hi,
I made a song and wanted to upload it on streaming platform. (I have a paid plan)
I tried with Imusician, but they refuse to work with AI content and specifically with Suno.
Do you have experience with uploading your work ?
I'm looking for a distributor who work on commission and can deal with AI generated content.
Thanks
A lot of people seem to be using DistroKid
Yes but DistroKid have a subscription fee. I'm looking for a distributor with commission.
Dude, don't ping the devs. There's no word as of yet, keep your eyes on #๐โupdates and #๐ขโannouncements
At the bottom left of the first post (will appear in the top left if you scroll down a bit)
@charred merlin Yes but I need to finish my indie game and 3.5 has major issues
...
I need to look into different distributors to understand the differences.
I love Distrokid just for how easy they've made everything-- they handle all of the hard work, they've helped consolidate all of my youtube channels into my personal one (no more "theharber. - topic" for my releases), they take 0% royalty, I can upload as frequently as I want, and it's like $30/year.
Now I know there's different considerations like the content only being active while I'm subscribed, and there's extra fees for paying to "leave a legacy" so the content stays covered even if the subscription cancels. In terms of that, there's CDBaby which seems to handle the legacy stuff differenly.
All in all, there's a lot to consider regarding how you plan to release you music, but in terms of just getting started & hitting the ground running, I can't recommend Distrokid enough.
'seems pretty straight-forward with them
Before you sign up for a distributor, investigate and read the terms of service and check the chat here for the history if they will accept your music.
Generally, what I had seen was:
โ
DistroKid and Ditto are known to accept and distribute AI songs and
โ TuneCore and CD Baby are known to reject #๐กโsuggestions-web message
I recommend taking a look at Top Music Attorney coverage regarding different distributors ( #๐ฃ๏ธโchill-chat message )
It will give you a good understanding of what to look for in their Terms of Service.
You especially want to find out how they distribute royalty and for any clauses that will make them retain it instead of giving it to you if they terminate your account (see TMA video about TuneCore and Symphonic as they had those very shady clauses).
In general, I mostly see people using DistroKid here and a few mentions of Ditto and LANDR on occasion. Check out all services, see what they provide and how much it will cost, etc.
People learn the hard way about TuneCore rejecting AI music (typically they cover it up by stating it sounds like another artist music but does not provide specifics.)
does anyone know if there will be available the choiche to save the single tracks of the generate song?
You can Extend the track you liked to make a longer version. See https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
(I assume that is what you meant?)
I just opened Distrokid, can I proceed easily with the loading with the tracks generated with the Suno pro account?
I haven't looked into Landr very much, but I saw their automastering service being one of the reasons people went with them.
TBF, I didn't really like the quality any better than I was able to get with a MINIMAL amount of mixing & mastering knowledge + Audacity.
And I've fallen in love with BandLab for importing tracks to add pedal effects, etc
If you mean being able to specifically download the drums/vocals/guitar/etc from a song you make, there hasn't been any specific mention of it. In the meantime, search for "stem separation" in this conversations history for more info โค๏ธ
Yeah, same. I saw some people reason solely being the remastering and I was like "I am sure there are better ways available".
I heard of BandLab but didn't try it out yet
Hi Paimon, no is not what i mean. what it will be a nice to have is to be able to save in different files, the vocal track, the drum track, the bass track, the synth track and so on...
Oh, stems. See Harber's response above #๐โsuno-school message
Look at my post to you from a couple minutes ago, ended with the heart
Best way I can describe BandLab's create section is it's a comfortable midground between FL Studio and a one-shot automaster.
I'm using it to touch up songs after doing the majority of work in Audacity. Pretty sweet that the prebuilt effects suites give you access to all of the different filters that compose them. Adding a simple "loud room" echo to a piece really helps set it apart
Ah, sweet
thank you guys, i check the history
It's not something that I'm looking into myself, tho I did test it out of curiosity. I've seen that a lot of other people are using it, which is why I mentioned it.
well very interesting to read that i'm not alone in this request! The fact is that some melody line, some vocal are really interesting and it should be wonderful take it into the daw to build the song whit the exact arrangement that is in mind. I'll give a try to audacity + openvino
not sure, if its as bad as distrokid mixea, i really rather do this myself! ๐
Yeah, Mixea is the same as Landr in that sense-- I didn't like the quality of a one-shot automaster as it "paints the entire song with the same brush" after attempting to figure out what genre it should be (so it knows how to apply a generalized filter), which is why I use Audacity for the majority of my work, then BandLab for applying filters and a nice middle ground between one-shot and overmastering. #๐โsuno-school message
just uploaded with distrokid today and it show me "dont you wanna master your stuff with mixea" and had a comparison for each song and maaaan that was just loud and the mixes sounded primarily annoying wich was ugh ๐
maybe it was a bit too lofi for the ai taste
but sometimes this stuff is intentional
server down ?
tried the softube flow suite in the last one and i am really incredibly happy with it. just dial in your stuff and then fun with finetuning some knobs while listening. it doesnt get any easier than this and you can still go very deep
thats quite awesome, no ai needed
Not for me
ok :0
ahhhhh thats why
why do my tracks speed up after 00:30 secs ๐
Not a common issue from what I'm aware, perhaps report it in #๐ชฒโbugs-web
Another option I've been familiar with but I haven't seen mentioned before is Amuse.io / so that might be one for people to look into as well for another option
That looks like a pretty interesting choice. I've never heard of that site. Thanks for sharing. It seems fairly inexpensive, too.
I tried various approaches. Initially I could extend from a point that made sense lyrically, but then I gave to and extended from the end of the segment.
I couldn't see that there should be any difference between extending from whole track or the last extension, because the two minute window would ignite pretty much everything beyond the last segment anyway.
But I did both, randomly.
how long does it take for an account to be deleted? i requested deletion well over a week ago and my account is still up
One thing I guess might help is if I get something standalone the sounds the way I want it to sound, and I use the at the very end I fade (beat-match, etc.) it to the end of the music I wasn't to extend, then I can hopefully extend it in a way that joins up but is mostly inspired by something separate.
Which I think maybe Suno is best placed to implement internally.
Feature-request: that^^^
Hi, are there any unsupported instrument names? For example, when I tried to use 'hurdy-gurdy' and 'psaltery' for medieval songs, the software didn't recognize these instruments.
Is it possible to rewrite the lyrics on the lyrics page? I loved how the first verse turned out, but had queued it with the chorus and hated it. So I rewrote the chorus and extended from the end of the verse, but the old chorus is still visible and I'd love to drop it from the lyric page cause it never gets sang. Lol.
https://suno.com/song/7c955a31-d10a-48a2-831b-ebc00d333a35
Soft pop, male vocals, dreamy, intimate song. Listen and make your own with Suno.
If I want a song in a certian language do I put the language name in the "style of music" section?
Writing the lyrics in the language you want is typically sufficient.
The issue will come to the accents/pronunciation though. Would be best to try a local + language for that in the style prompt like
- Hong Kong Cantonese
- Boston English (I wonder if that would differ from a New York English?)
- British English (or England English?)
kk ty
Old copy-paste made for v3.0, I haven't tried for 3.5 but the metatag reading is much better:
Getting duets... that is more luck than anything else. Try to use male and female sounding names in tags, that might help, but it is still hard to achieve.
A few people were able to do it but at an expense to their credits:
3Daizy's duet
MrTree's duet
Suno.Wiki entry
Again, getting it working on purpose is hard.
Alternative methods:
-
It is easier to make a duet by doing something like generating instrumental part 1 then Extend it twice (one extension where you specify for male vocals in the styles field, the other for females), or something similar for each alternating part, then edit them together in a DAW.
-
Even easier (and cheaper) is to make a full track then swap the voices gender using other AI tools like kits.ai (see #๐ฃ๏ธโchill-chat message ) or your own voice with a RVC model (see #๐กโsuggestions-web message ).
You can use a DAW to edit which voice track plays when.
You raise a valid point about the difference between getting the whole song or just extending if there were a really short context memory-- I'd only seen that comment once from a moderator, I'm not sure if it was in regards to v3, v3.5 or both, and from what I'd seen a member of the staff hadn't confirmed. I would be curious to know specifically if Suno is incapable of recalling a particular motif if it hadn't been used within the last 2 minutes of a generation...
That being said, if the song intentionally starts off with a strange structure and changes styles quite often, it's going to have a hard time trying to reasonably predict what you would want to come next. If you feel like some of the generations (outside of the final result) have poor audio quality, report them through ... > Report > Bugs > Poor Audio Quality to help improve the model & help train it against making stuff that you consider bad.
Lastly, I know it's totally subjective, but I think the fact that we've taught a robot to make music for us is mindblowing. It's changed my life to be able to create 2-3 minute songs that I wished my band and I were talented enough to play, let alone asking it to write "a 10 minute long guitar solo that stays interesting the whole time".
I use RipX DAW Pro, which does a great job of ripping - separating to tracks that you can then edit and export for mastering. RThey have a trial version to test it out, but I highly recommend it.
In my experimentation it is indeed the case that Suno cannot reproduce a motif if it hasn't been played in the past 2 minutes.
I second your final sentiment there about how it's mindblowing and life-changing.
Hi all dear, What's about tags to make Nursery Rhymes kid song? Please tell me.
Hello, is there a problem with mixing metal with acoustic guitar? Every time I extend using those options I get my credits back, it's been done more than 20 times now.
Tinda Lee for the win again-- okay, extrapolating from that, just because the outright motifs might not be present, I wonder how much "musical inference" it's able to discern from sections outside of that context window in a reasonably structured song?
Maybe getting a little too out there, but if the context window only contained "the response", would Suno be able to re-construct a likely "call" from which to base future extensions from?
Admittedly my songs rarely go beyond 3-4 minutes, so I don't have a lot of experience where this would become a problem, but in the instances where I've extended from an extension (on v3 before 3.5 was released), it always seemed to produce either a strange bridge based on the last segment, or a coda, but it rarely incorporated the song's overall motif in the same way that extending from the whole song seems to.
And I second your seconding-- Celtic Moonshine friggen rocks
๐ค
testing sorry
aw omg thanks ๐
Did you try nursery rhymes kid song? (I'm not being sarcastic, that's my first suggestion but I don't want to assume you haven't tried it.)
Children's Nursery Rhymes seems to do the job: https://suno.com/song/1d4fb4e7-b656-4dcb-a748-8e8036f9ad53
Nursery-Rhymes should also work, e.g. https://suno.com/song/9a7c16b3-9efd-4b8f-bdd3-5ce0f2e3769a
So, generally if I need to make something longer (such as when I'm constructing a "new" version of an existing song), I do it in parts, extending from different spots within the first 2 minutes to get all the verses / choruses, etc... and then reconstructing it in Audition. That's been the only way to keep it cohesive.
FYI whoever: v3.5 is very good at reproducing what v3 creates, but because it's better at structure and following the lyrics (usually) it can make for a better final product than relying on v3 or v3.5 exclusively.
Thank you dear, I tried a lot of Nursery Rhymes kid with this tags(Nursery Rhymes Children, kids song, happy kid, female vocal, short intro, piano, xylophone, drum, bass) and other tags, It 's not bad for generated song, but I need metatags for good then this.
Hi, not sure if this is the right place but can we ask questions about suno bark or is this only for suno.com?
my 200 credit goes like that!
Refund time should be 20 sec I think.
Thank you so much dear, I will try more.
Jรก to pรญลกu pลesnฤ tak, jak ลรญkรกte, v placenรฉ verzi, a pลesto se to dฤje. Kaลพdopรกdnฤ fajn potkat tu ฤecha. Aลฅ se vรกm daลรญ samรฉ krรกsnรฉ pecky ๐
While Bark was developed by Suno.ai, it's been made open-source and I don't believe its actively being worked on by the team. As such, I believe this server is primarily focused on Suno, though there might be some users who are knowledgeable on Bark.
2 minutes back in time, + 2 minutes forward in time, so in the final song, the motif may be as far apart as 4 minutes: #๐โsuno-school message
You can DM me Bark questions, but no promises I respond very quickly. But I did just get back home after a week and am at my Desktop tonight.
Finally, someone doesn't think I'm crazy! Now I just have to tell all the doctors that...
But in all seriousness, it seems to me that songs using the "extend" feature have lower audio quality that ones that don't.
so i used (succint guitar outro) and the singer said it over and over
instead of actually having a succint guitar outro
Iโd like you to do a test for me, if youโd be so kind, and have enough spare credits to do so?
I notice that when listening to the extensions separately (listed as โpart 2โ or whatever), the clip sounds quieter.
However, when I choose โget whole songโ from part 2, the audio seems to have averaged out and the โdipโ is either gone or alleviated.
Do you notice anything similar?
@narrow stirrup I had that happen too today, it kept repeating "clear angelic vocals" . Must be a bug on the server. I did report it as a bug.
๐ i joined the server just to ask that question but im glad its not something im doing wrong. Thanks!
Oh, Iโm at the point where I kind of find it adorable when the robot makes a misunderstanding like that.
โSilly robot, thatโs a metatag in the lyric box, not a lyric!โ
Sometimes that will happen if the AI misunderstands the tag, or just because it felt like singing it.
Perhaps try placing that in the style prompt instead of the lyric box, making just an instrumental, and add an extension for lyrics after getting the style right.
great advice. thanks a lot!
I find it helpful to think that itโs not explicit instructions, youโre โtalkingโ with something that can โmisunderstandโ. โค๏ธ
I tried and I don't notice a difference, but maybe I just dont have sophisticated enough music senses. I thought all the "get whole song" did was cut and paste the pieces together
im really good at midjourney and other ai's but this one is new to me so im still learning ๐
Great place to come from in terms of understanding how it all works.
Extensions are sort of like taking a bit of the song to img2img (or audio2audio) into a new continuation
yeah i wouldnt say im lost at all. I understand what you say when you reference ai behaviors because ive had em all
3.5 ROCKS at โunderstanding the homeworkโ when given a previous song structure, v3 is willing to get a bit more experimental and make stuff that youโd find 6 hours deep into a jam session ๐
@charred merlin Very grateful for all the good you add to this server! Thank you! ๐ซต ๐ชจ
Glad to hang out and try to make myself useful โค๏ธ
Thanks for your kind words, David ๐
Oh also, is "get whole song" supposed to use credits, becuase it doesn't for me.
Zothmire โ I thought it did? I guess Iโve never actually checked ๐ Well huh. Gonna check that in the morning.
Bankshot โ The Suno.wiki has stuff on voice tags, but I think thereโs been more discussion on that previously, searching this channel for โholdingโ โextendingโโ Paimon, Furrybeans, SixStringToker, any of the mods; just a few helpful names to look out for who might have discussed it previously
doing heyyyy actually worked perfectly
Keep on rocking in the free world
๐
How do you get out off dark modeI? Can't see anything on the Suno site.
It looks forward on an extend?
No, it doesn't. I think they know that would be super uncool, like paying twice.
While generating, the memory is fixed. So if the previous song had a motif two minutes in the past, and then you start generating, as long as it's generating it's not going to forget that motif. So if you get a maximum length generation of 2:00, then the total distance between the motifs could be nearly 4 minutes
Oh I see, I misunderstood what you were saying.
This is the way it works now, practically, Suno could decide to do something different one day (even without changing the model).
I have some compositions that I created about 10/15 years ago in Cubase. Original material. Can suno rework tracks using the same melodies and remix them or keep same structure but filter them through the Ai system?
You might benefit from the Audio Upload feature. You can upload between 6 and 60 seconds of audio and then extend it using Suno v3.5 to continue the song. I did this with one of my soundtracks and it turned out great.
Is there a way to edit the lyrics on a song's page? Every time I extend a song, it keeps the previous lyrics and then adds the new ones. So the final song's lyrics on the song's page are super long with lots of lyrics that are no longer in the final song
Guys when weโre doing audio inputs, do we extend from the first second or the last second of the recorded audio input?
What is the logic of Suno?
If on a free account I click 5 times to get 5 attempts to create a song I lost 50 credit. i.e I got 10 songs, lost 50 credits. To compile the song I need at least two days = 100 credits. i.e as a result I lost 100 credits on one song.
With a pro account I got 2500 credits and I click 10 times do get a result of one complete song. But in fact I will pay 500 credits (because I "waste" 10 songs on one click). This means that I would be able to create only 5 songs per month. But I see it unrealistic. For example if I have long text of 31 verses, Symphonic metal orchestra, this could take like 2-3 weeks to get one song. Because there are the problems with the voices. And I need quartets, not duels. Possibly on sung-dialog two. I could use the OpenVino, but - on linux the Audacity crashes with the plugin. And to use the kittool.ai it would take another extra resources (money)... Yet to pay for distributor. So the music would be still very expensive. I would need to test the lyrics/title/style instructions with generating one song per test. Not songs on one go. Otherwise the credits are clearly wasted.
What is the maximum length of the final song?
Maximum length of the final song is 5 or 6 minutes I believe, one of those two
Actually nvm there is no limit, but the maximum length of the first song is 4 mins, then every new extension is 2 mins max
I asked because recently I got feeling like the song could be about 6 minutes but it was 4:20 whole song duration. It was cca 31 verses. But it could be my incorrect calculation.
Last. Anything after the "extend from" time gets completely ignored when generating the extension.
Not sure I follow the part about wasting 10 songs on one click. Are you referring to this part of the Pro sub? If so: You always get 2 songs per press of the "Create" (or "Extend") button.
What "running jobs at once" means is how many times you can press the button before you have to wait for some songs to finish generating. Pro users can have more songs generating at the same time if they want to. You still just get 2 songs if you press the button once.
There's no maximum length but really long songs can have issues from what I've read.
Check out the length of this one: https://suno.com/song/52e02bc8-59cf-43a9-99f2-cb4ff6c283e7
Thank you for explaining. This part was the confusion for me.
Oh, interesting song. So it looks like it can handle pretty long part without problems. I see some problems with the Symphonic orchestra there.
I just found a bug that lets you generate free songs
How many free songs?
oh wait nevermind, maybe it's a UI bug. Hmm weird anyway
Did the credit counter show the wrong number?
sort of, it's deducting the credits late because it thinks they've been refunded at first
it's because i'm continuing from a 4 second clip that had it's credits refunded, so the extensions also say credits refunded at first
Oh really? That's an interesting one ๐ sounds like a cosmetic glitch. Might be worth mentioning that in #๐ชฒโbugs-web
xD
Are you doing this on a free plan or subscribed account? (Just looking at the number of songs being generated here)
sub ofc
Was just trying to figure out what else might happen with the bug. Was curious if it was letting you get around the concurrent generation limit (e.g. free plan = 2) but as you're subscribed that's why you can make more ๐
can anyone tell me why i have concent restriction, i cant make it public. i have pro paid version
Sounds like you're using an audio upload with vocals or copyrighted material.
they figured it out in general chat
I don't hang around there ๐คทโโ๏ธ Appreciated though
My internet is loading other pages, but suno is failing to load for me:
This has also just been answered in general chat
OK it's working again
Getting frustrated, it doesnยดt want to end a song Iยดm working on. A few hundred credits down the drain. Annoying as hell. I have tried every prompt there is to use. AAAAAARRRRGHH!!!!!
Extend where you want to end with the extension set to v3.0 when using ending tags (e.g. [end] in lyric box and "end, short" as the style tag) as 3.0 is much more reliable for ending than 3.5
I assume it repeated the issue after refreshing?
You would want to post that in #๐ชฒโbugs-web
Thank You so much, that worked great. ๐
Would it be possible to get a list of all the instruments Suno understands? It's annoying adding instruments to a prompt just to have them ignored completely. I waste so many credits trying to get it to use a particular instrument only to realize after many tries that maybe Suno doesn't even know what instrument I'm talking about or perhaps it does, but doesn't know how to replicate it, so it replicates something close.
Better yet, it would be nice to have a page with all the tags that have been proven to work on Suno. They could even be split by sections like "Instruments", "Genres", "Voice modifiers", "Adjectives", etc.. This list could be updated as the model gets trained further.
Hi there! I agree with you!
Only by using a strict, technical, and proven notation, you can then produce expected and professional outputs!
I wish there were strict rules to follow, like in every programming language.
Unfortunately, I see a whole bunch of different notations are used to generate the same styles and songs.
It has to be consistent if you want to be a professional. Hopefully, Suno will come up with a final notation guide someday.
should not useing [end] for it to stop, seemingly my songs forces it self at times to be 4min
Consistency in tags would be great, though I really don't expect to see that very soon. The devs are still probably trying to figure out what works and what doesn't, too. 
I'd argue that a true professional can learn to adapt in the absence of such things...
Do tags like high quality, best vocals, do anything lol?
I know that's a common part of image generation prompts, I'm even "guilty" of it as well, though I haven't tried it specifically with Suno. I'd be curious if anyone has experience trying similar generations with/without those types of prompts?
tbh I'm switching off from it and it seems like it does not make a difference. It seems to be very hit and miss.
Also I noticed vocal quality just dives off the deep end the longer the song goes on.
Also I critique this program but it's amazing lol
I don't have hard evidence but since I started putting "crisp vocals" into my prompts I've had much less of an issue with that. It's probably just in my head though.
The drop in quality can be avoided by getting creative with your extensions, pulling off the same spot a few times rather than extending the extension, and then assembling the track in Audition or whatever. It's not a perfect solution, though, because the randomness can sometimes mean you have a rough time joining the various verses together.
So, I couldn't sleep last night and I decided to try something new. I made an instrumental track with "female vocals" in the prompt, and then after I got one I liked, I wrote lyrics based on the nonsense words being sung. It was kinda fun; broke through a bit of a writer's block I was having on this particular song. I didn't end up using that initial instrumental, but it still was super useful just for finding something interesting musically.
I don't think it's in your head. I've been experimenting with various vocal metatags and I find they all have a dramatic effect. Adding "Emotive Electronic Voice" to my tags when I want a pop song seems to reeeally bring it to the front of the mix like here https://suno.com/song/001f844e-65d7-43a5-a54b-2ee48b28a792
ultrapop, hyperpop, synth driven, emotive electronic vocals, viking metal, epic battles, cinematic, chanting, song. Listen and make your own with Suno.
anyone know what stem separation algorithm to use for UVR for isolating saxophone successfully?
id try sonauto first then to to other more traditional means as it can often fluke and give you super clean stuff
I have 60 song playlist,
when loading second page
and when editted name second page song name not changing name until refresh page
I've noticed that when I generate a prompt that is very different from a previous prompt, parts of the previous prompt are used for the current prompt. Does suno have some kind of memory? It makes it hard to switch genres from creation to creation. Sometimes it makes for interesting creations, but it's frustrating. Especially because sometimes it takes many generations for that aspect of a previous prompt to disappear.
Is possible to edit a already generated music, change lyrics, cut and add things?
how do you prompt for electrical guitar solos? they always come out sounding like a guitar sound on a keyboard, when it comes to distorted electrical guitar. saxophone, mandolins, etc, don't seem to have the same issue
the attack is all wrong
@vital plaza it's not a programming language that's the beauty of it. Instead of thinking it's broken think of it like this, somehow it's taking God knows how many languages with different grammar and structures and its interpreting all of it into music. It understands Tibetan. It even understands ancient Akadian. Old Norse. You can write an entire song in a dead language, including the prompt commands. Think about that.
You cannot fit logic and art in the same container, the more you have of one the less you get of the other.
They have no idea the power I have found in my hands..
Anyways yeah, I don't think re-use prompt on extended songs is supposed to let you go over the limit-
Question, Is there a specific reason why you require that?
yes, i write all my own lyrics and don't need weaker verses repeated three times before the song concludes
Iโm not sure if you have access but you could try fading out after it ends on its own for the verse, maybe when the melody plays etc. are you using any type of daw or strictly trying to have the generation end
It sounds like they are trying to use (end song) as a prompt.
Usually [end] and/or [outro] helps with this. v3.5 is stubborn. Switch to v3 for the last part as it obeys these prompts a bit better.
Great point. I personally have been using outro with 3.5 and havenโt had a problem yet
The restriction isn't designed to limit you, it's designed to give you an understanding of how much Suno can understand.
https://platform.openai.com/tokenizer
AI needs to break words into tokens to be able to understand them, and each different type of AI can only "hold/understand" a certain number of tokens at once.
Apologies, I already know, I was just pointing out an issue on the web design not actual ai 
Roughly assuming that 1 word is the same as 1 token (slightly more, usually), Suno.ai is telling you that it's only able to hold 3000 words before it starts dropping all of the earlier words you gave it.
Tagging is usually what helps a model understand what's important about a particular sample, and therefore reproduce that aspect when a user queries for it. For example, when training a LoRa, you might tag a set of photos with "old brick" and "old brick wall" and "old brick pillar" to teach it what "old brick" things look like. I would be interested in knowing how samples are tagged in Suno, assuming they're working under the same theory (they do seem to be).
The things I would expect/wish to work seem not to. For example, suggesting a key, key change, timing, tempo, with something like [Key D Major] [Key change A Minor] [Timing 4/4] [Timing Change 7/4] [Tempo 1.5x] or similar, would be a delightful feature. Also perhaps [Male Vocals] [Female Chorus] etc. But tags like these seem to be ignored. If there are any fundamental music tags that the model really does understand I would love to know. I'd rather specify [Timing 11/17] [Key A Minor] [Tempo 90] or something than try to find a genre tag that occasionally outputs a signature close to what I want.
How exactly to report bugs on suno? not using suno support email. I don't see the ability to describe my problem. A lot of problems is with drawing lyrics or diacritics. It seems like because of expanding the text, the old lyrics (which should be deleted) says on the frames of the playing video. Which is very bad because listeners don't see what is to be song and what is the meaning. Like in my last video which I inserted in playlist of Japanese trainging lessons, the singers sing something and I dont see the meaning. Could be check in the middle of the song and later. But I see a channel of suno bugs here on discord. How to get the lyrics bugs there? We need some feature to remove the lyrics which should not be there, but it could be too complicated, the program should do it self.
โฆ > Report > Bugs
Try with [BREAK] [OUTRO] [END SONG]
usually works for me when Suno doesnt want to end a song
Can i make another account to buy another payment with same card? I need to make a song and my subscription refresh on 29 this mounth. I have just 5 credits now. ๐ฆ
i think u can buy extra credits
Yea but, its 400redits its 4 dollars, to much.
The terms of service specifically mentions that you canโt make use of multiple free accounts, but it doesnโt seem to mention any restrictions on purchasing another subscription โค๏ธ
It seems like you know already, but just for the sake of educating othersโ any unused credits from a subscription donโt roll over between months like Extra Credits.
I have no idea if using the same card to make the subscription would be an issue
Yes I know, I've done this before and got blocked. I have been paying the subscription for 4.5 months without any problem. The thing is, this month I finished the 2500 credits way too early. I don't even get the 50 free credits daily as a free account when I reach 0 credits or 5 credits like I have now. And now I need to make a song by the 27th of this month, having only 5 credits, it's not possible, buying additional credits is over the real price, 4 dollars 400 credits, and I use 200 credits to make a song, until it comes out the way I like it. And I wanted to ask if I was making another account, with the same registration data, with the same card to pay for another subscription. There is no problem.
Aaa ok. Thanks ๐
Thx, but I tried that already and then I tried the "change to v3.0" and in "Style of Music" typed in "Short End", nothing else. This have worked everytime so far. ๐
Try on end of song (Instrumental) (end) or (fade out) (end) or put on last text (Outro) and after text (end)
Or go to anothers songs who are better end for you and check the final text.
i have realised that when you put a prompt in the song description, the system understand my prompts well. the problem i cant add lyrics. when i use the custom section, it gets the lyrics but the style of music is so random to what i describe not getting what i described. i hope you allow us to add lyrics to the describe song prompt. or make the style description in the custom section to understand complex prompts better
Hi guys, any idea how I can ensure I don't throw out 70% of all songs because they don't have an end? I'm talking about instrumental songs, so no lyrics to throw an [END] into
You can use [end] in instrumental songs that have lyrics (not using the instrumental toggle), just add "instrumental" to your style prompt. (Or if you are extending a song that doesn't have any singing, it probably remains a pure instrumental even without that.)
Alright, thanks
how can I change lyrics after I create a song? there is no remix button anymore.
anymore
๐
I found it very odd and suspicious that it will let you say "f***" in the lyrics. But it won't let you say "bastard". Wtf
what genre do I add to make the voice clearer, its very unclear by default sometime
- tolerant restrictions in โtermsโ and โwordsโ (example: Nig***zzz Black Rap-Jazzz) - most likely "won't miss".
The Remix button has been renamed to Extend.
"Remix" was renamed to "Reuse Prompt" and does exactly the same thing it used to do (copy the prompt and lyrics.)
"Extend" has always extended the song with a new part and is not related to this.
Oops, I got them mixed up
I created part 2, there is a bad beginning, but after that it is exactly what I need, can I get rid of the beginning in part 2?
Just shared my workflow on how I create a song in Suno and what I do to master and clean the audio...
#1254413503687692388 message
If anybody wants to take a look, there is a PDF file, as the content was too big for a single message.
And also has audio proof that how good mastering can clean a Suno song. Leaving screenshots here as well.
Go back and forward to see the difference...
If you have questions ask in the workflow thread, I'll try to answer them.
Continue with that. Create a new part 2 and mix in a multitrack audio editor. Think it like inpainting.
how to make the ai sing another language?
Just write it in that language. Is it not working?
Gonna move this here.
so
Um.. How do I delete account? I accidentally made new account ๐
Or is there just no way to delete accounts?
(speaking about Suno account, Not Discord to just clarify that )
You'd need to e-mail support@suno.com to ask them to delete the account that you don't want.
But, if you're not concerned about Suno holding information about that account (e.g. username, profile picture) and simply don't want to use the account with Suno, just don't log in to it.
Not too concern, but just that I can't even disconnect the Google off it. ๐ Since I accidentally used my main-gmail to make the otheraccount. And don't want it there. But thanks for the info ๐
it would take to long to do but ok
If you are using generation for lyrics, tell it that you want it in "x language".
Like "a song about air conditioning in hot weather, lyrics in German"
Do you ask for help here
ok
Yes, we ask mostly about how to do things that we are stuck.
Would it be
[Verse]
[Verse1]
Or
[Verse 1]
?
Trying to figure out better formatting techniques
If the lyrics are not that different "length" wise, to keep the same melody verse -> verse
If the verses are too different, verse -> verse 1
I'm considering upgrading to a paid subscription, but one thing I wanted to know before doing this is whether there is a way (some sort of label, etc. in the UI) to see which songs and clips I created when on each plan, so I know which ones I am free to use commercially, or is it something I would just need to keep track of myself?
There isn't. It would be great if there were a label or some indication of this though. I just go by the date and time I subscribed (which I have an e-mail receipt of.)
This is awesome i was looking for a guide on ai audio clean up like this because i am not a sound professional, but my takeaways is, there is no quick way to clean up a song that has distortion in SUNO (??) You have to look at the spectogram and spend time (Days, weeks) fixing things... I have a few songs that i really like but have some frustrating distortions.
Pff... Trying to write an answer ๐ Bot want's me to keep it short.
With LANDR I get a 90% success with the tracks. 10% is the metallic sounds I get. I need to go in and clean it.
If it's for personal use / pleasure, stick with Audition, or Bandlab. No need to deep clean.
For streaming platforms, you need to put some money and time to get the best -- if this is the case check LANDR, try a month with the Standart price.
If the distortion is too bad, it might not be possible to recover the song at all as well -- I have some old songs like this.
But Suno now generates pretty decent outputs. Even the muddy noise can be cleaned 99%.
In Bandland, how can it be used to clean audio? I tried some effects but I'm a newbie and it got worse. Which ones are recommended?
Go to the mastering, not editor.
Ah jeez, how is BandLab so good?
4 different mastering styles for free. I genuinely expected it to be a one-time free service, then included in the membership. I just..
thanks to both of you. I dont know it. I'll try ๐
@weary wharf, good on you for making a workflow guide! Thanks for your contributions to the community โค๏ธ
I haven't had a chance to check it out in depth, but I'd been planning on making one as well-- the only problem is my workflow keeps growing and improving with each song! ๐คฃ ๐
3.5 gave me 20 seconds songs twice today, out of 4 generations. Anyone seen that? Normally, it beats a dead horse to make a song 4 minutes. Both songs were tagged [chip tunes] and [glitch] ... not sure if that would do it, or why it would.
Isn't the randomness part of why we choose "glitch"? Sometimes it's not a full length song, but sometimes it makes some WILD stuff to extend from.
Same for me actually ๐ I wrote it while I was mastering the song I shared. Win win ๐คฃ
Maybe it could be, for someone. I wasn't expecting it. They did sound interesting. I don't know if I extend it that I'll get a very long extension either though. What if the extension is 10 seconds? That's a lot of extensions to get even a 1 minute song.
What would be really nice is to get 4min of glitchy music, then chop it up in a DAW and use it as samples.
IMPORTANT QUESTIONS FOR TEAM: When we upload audio to Suno, or custom Lyrics, does Suno use that music or lyrics to train its AI?
This is really cool place to learn. What does this manage to do? Can this clear vocals or remove unwanted basses where the song generated by suno is too overpowered? And should I use it before I would do the vocals change, as you suggested few days ago for duets?
I don't think that makes sense. What Suno generates is not that high quality when it comes to sound. Training with AI generated data tends to create worse models, so no one does that.
Even OpenAI dumps most of the data they are gathering, because ChatGPT gives pretty useless and ๐ฉ answers on most of the chats, bad for teaching what is "true" too. Not good for training data.
i make alotof 8bit and video game music, music box too all gen short gens. i believe it is because the music is faster, this often happens when suno is going really fast in my experience
Sometimes Suno doesnt listen to my promps, like if I were to put "Rock Chinese music", or something along those lines, it wont listen to the rock or genre I put. Also whenever I upload a video I feel like no one ever watches it, as Ive uploaded about 15 on my alt and made a playlist but it never gets seeen
*Seen
Hi Im trying to generate music using your sdk for Bark but i really dont know the difference between Bark or chirp. which one can I use to generate music or train the models to create vocals and music as individual stems
I got some longer ones by putting in various [instrumental], [build], [solo] etc tags in the custom lyrics box. Also, if you haven't already tried it, short vocals like ya! (Yaaa) are sublime with 16 bit or probably other bits I haven't tried yet ๐
Yeah i tried some short vocals but they repeat every 10 secconds
But i just stuff them with [] commands
the way i make suno's i usually only take 20-40 secconds then i shift the style prompt, so it's not a big issue for me
I believe suno only recognizes 1 genre as the tracks genre - but you can get elements from another genre using , "genre" Cross-genre, "genre" "genre" "genre" Fusion, "Genre" Influenced,
those are just the ones that I know
be careful where you put commas if you want to blend them
Fusion is probably my favorite although Influenced is really cool too and they all sound different
Hello all, is it possible to take out a Chorus from one of my suno song ? There is a lyric mistake in the verse before and after the chorus, but the chorus is like.. perfect!! thank you very much
I know there is this song input since last update, maybe i can use this ? Thanks
Gonna have to hold out for v4's inpainting for that. For now, all we can do is fix that via extensions, which means backtracking ๐ฆ
Fusions are blowing all my credits. Ive been trying to play with instrumental fusions with latin beats, and just came up with a bunch of steps i just try to see which one yields what i want. This is dependent on the genre, but I think some genres fuse better than others, and what point you extend also matters. These are the different thigns i try: 1) First enter both genres on the music style either separated by commas or do "fusion of "genre1 and genre2" (most of the time doesn't work, but sometimes you get lucky), 2) On the music style prompt Enter the first genre, then the instruments for the second genre. 3) Create the first track with the first genre, then extend and rewrite the music style with the second genre only (for me this doesnt' often yield a true fusion, but a base of one genre with solo instruments of the second genre, it is also important where you extend at) 4) Sometimes i put the base genre on the music style, then use instrumental tags on the Music Lyrics for the second genre.
I get your point.. but then, Im uploaded high quality professionally produced WAVS, and if its buried in the T&Cs that they are allowed to train on uploaded data, then they are essentially getting additional access to high quality data to train on?
I love doing fusions as well. One trick I learned relatively recently is doing "Fusion of [genre 1] rhythm and [genre 2] melody" if there was a particular type of fusion style I wanted.
Here is an example with rap rhythm and Hip-hop melody:
Have you looked through the terms & conditions to see how they handle your uploaded content?
It states that they release all interest in and rights to the content you produce, but there's no word on the training data in the Terms of Service. If you want a more detailed answer, you'll likely need to contact them directly at support@suno.com
From the Terms of Service, it's not explicitly stated that Suno trains their models on uploaded data. However, the broad license granted by users to Suno allows them to use, reproduce, store, modify, distribute, create derivative works based on, and otherwise make available the content uploaded to their service. This implies they could potentially use uploaded data, including high-quality professionally produced WAVs, for various purposes, which might include training their models. You might check with an attorney representing your interests to be sure. And contacting Suno directly may yield meaningful answers as well.
It's clearly stated in the terms that when you use Suno, you give them right to do with your music what they want. You also claim that you are the author of the stuff you upload, as far as I remember. They want to make sure, they can do with it what they want in a legal manner.
argh, please implement an [END SONG] feature so I don't get two minutes of the AI daydreaming about my lyrics and melody.
anyone else experiencing a timeout?
Wow ill try that one thanks for the tip!
Extend where you want to end with the extension set to v3.0 when using ending tags (e.g. [end] in lyric box and "end, short" as the style tag) as 3.0 is much more reliable for ending than 3.5
thanks for the tip, but my recommendation still stands,
It happens time to time on certain regions. If you are getting error 524 (aka cloudflare host error, likely due to DNS issue), you just need to wait a few minutes to try again or access with VPN in another region https://discord.com/channels/1069381916492562582/1252384524319326310
halloween trap, jungle, uk drill trap, breakbeat, factory, euphoric, drum-and-bass song. Listen and make your own with Suno.
One of the downsides of OpenVino (and other stem separators from what I can tell) is that it groups instruments by drums, bass, vocals, then all other instruments into a single track.
Does anyone know of any UVR models or similar stem separation methods that can further separate the other instrument section?
Is there a way to control the final mix levels. Like increase vocal level balance if they are not very audible? I'm using the website dashboard if that makes a difference. Thanks ๐
What I did was wait a while and then open the page again
here same;/
Some options are a bit expensive;
For instance, Spectral Layers 11, the new version from Steinberg can separate up to 7 tracks including saxophones.
It can also further separate the Drums into Kick, Hihat, Snare, etc
Can some other people test this for me? I may have found a way to fix 3.5s issues with ending songs.
After experimenting with multiple genres and lyric lengths I found the following. Keep in mind this only works if you write or generate lyrics beforehand 3.5 still doesn't like prompts.
More than 1400 characters and the ai never wants to stop singing.
If you include the following exactly as written at the end it ends more frequently than without.
[Instrumental Interlude]
[Instrumental Interlude]
[Instrumental Interlude]
[Instrumental Outro]
[Fade Out]
Nothing else really helped me at all, I tried so many suggestions from reddit, youtube, and here; but, after burning 2000 credits these two things are the only things that Influenced whether the song actually ended reliably. And keep in mind it's experimental AI it's still gonna mess up no matter what you do so be patient.
With that in mind, after following those two rules, I haven't had a song end incorrectly in the last 20 generations
Is there a way to have it generate sound effects only? Like bones breaking or something?
Yes, sometime v3.5 has issue with song ending, sometime Suno create song ending themself and ignore my [End] order, so i always end my song with only [Instrumental Outro] in the lyrics menu so the next extend usually end the song, if still not then i add only [End] with the next extend, most of my songs don't have problem anymore, but it's stiil an issue in v3.5, ie: https://suno.com/song/99ca7f3c-3dd8-45d7-8f51-b216ab608313 i don't expect to be this short.
Nu Metal, Nu Metal Vocalist song. Listen and make your own with Suno.
While I donโt want to say no, Iโd say itโs highly unlikely, since by default the AI tries to interpret things as music.
That being said, there might be specific prompts that can produce sound effects with a higher rate of consistency, but Iโm not aware of many experiments into the matter, beyond a song that featured a gunshot.
It seems like people donโt want to spend their credits trying to experiment in that kind of way, though I would be curious to know if anything turns up interesting results.
A related side note as pointed out to me by SixStringToker, โnegativesโ donโt seem to be respected in style prompts or meta tags, so โno drumsโ is likely to produce drums , โno instrumentsโ is likely to produce any sort of instrument
What I find does work though (atl with the examples you gave) is like vocals only, guitar only
Or [insert every instrument you want in the song but dont list drums] only
omg that worked like a charm, and didn't even cost credits. lol
How do I use the Suno Chirp Bot? When I try to direct message /chirp says it couldn't accept my message as I am not a friend?
Chirp bot was retired in April 2024. To create music now, you need to use the website https://suno.com
Some helpful information can be found at #1151542418991239168 like the community wiki https://suno.wiki
No problems, didn't say Chirp bot was retired in April 2024.
Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com
Is there any word on an Official API?
IIRC, the devs said they will considerate it in the future but other items have priority at the moment
Does it matter the order of the phrases? Since it seems to pick out the strongest terms on it's own.... or can you minimize a certain term by putting it last?
Ok..OK... this is weird, but is it legit for someone to use what I say in discord as their lyrics without my permission?
Because I really feel like that's not cool
Especially if you keep my name in there
Like is that legit?
That's not your own writing
violin, anime, synth, powerful, electric guitar,piano,piano aesthetic, pop song. Listen and make your own with Suno.
Anyone here must have experienced the same thing, like when composing a song, for some reason after the song ended, Suno was still producing it so there were strange sounds. Or even the vocals talking about the lyrics. ๐
I experience this very often. ๐
someone said try to [End] the song with v3.
I did. But it's still the same.
it happens to me too, maybe a new bugs or something, so i always extend in different sections to avoid this issue, it works but need some credits, hope Suno fix this issue.
I am trying to get the AI to do verses and Chorus in speccific ways, but it wont. am i doing something wrong there/
is there any word on an API yet?
Chirp/
/chirp
Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com
suno chirp bird
Hi there!
If youโre ready to dive into creating music, head over to Suno.com.
For more detailed information, check out docs-guides-resources.
If you have any more questions, just ask in #general-suno where you can always find our community, moderators and guides to help you with most questions. ๐
no matter what i try, suno site always tells me "An error occurred."
You donโt have implicitly copyright or ownership over everything you post in a public forum.
Read the error messages โค๏ธ
No official API has been announced
To be fair I think that is a particular Suno error message in its entirety ๐. If it keeps happening @tame blade write what you are trying to do step by step so someone else can check to see what could be causing it
Since it was less than 3 minutes after a couple failed attempts at using the discontinued /chirp command, followed by public, descriptive error messages, I figured that was what they were referring to instead of something on the site itself.
๐คทโโ๏ธ weโll help either way
Tiime for experimenting with Suno.
It sounds like you might be, but this is incredibly vague.
Anything else to share to help us help you? What are you using as a prompt? What are you expecting to hear VS what are you getting as a result? What different prompts have you tried?
I am trying to get the Chorus shouted in an angry voice, like an explosion of rage and frustration, with the verses in a soft voice.
So far i never get the chorus right.
I use promts that define the spicifics for chorus and verses,
Like "angry, shouty chorus, Clear female voiced soft sung Verses
#1252286293992407140 message this might help
thanks, imma try that out
This is meant to be reserved for technical discussions, please save self-advertisements for #๐ฃโsuno-showcase or #1154790975240155247
Hi,
When distributing your suno creation, do you also distribute it with Youtube content ID ? (also on socials like Meta, Tiktok, ...)
It seem giving issue with AI creations.
It require to proof that I'm the right holder.
How did you fix it ?
Thanks
Kind regards
I use Landr to distribute (and have successfully done YouTube ContentID). You own the lyrics and composition when you pay for Suno, so you do not need to check boxes like "Uses samples I have a license for" or anything like that. Which means you are the lyricist and composer, legally. If you need proof of this, a screenshot of your Suno subscription and receipt, that also shows your email, should suffice.
That said, I highly recommend being transparent about how the music was created even though you aren't legally required to do so. First, because it's the right thing to do. But also because YouTube and some of the other services have their own requirements about this. That said, if you upload your music in a video, YouTube now requires you to indicate whether it is "synthetic and seems real" in the details. This has changed very recently. You can be transparent elsewhere by including info about the music's creation in the fields of the song itself (such as in the comment or "other artists" or whatever) or just by talking about it in the description and Bio of places where the music appears.
hint how to write a promt for chord refinement, is there such a possibility?
Try just including the chords, like tablature. [A#] or whatnot. I've had some success doing this.
Thnks
A helpful tip is to experiment with covers using available tablature, just to see it working. Or, even better, use an existing chord structure and write your own lyrics to it.
Currently experimenting with expanding an existing song using an upload of the first 60 seconds, with the intention of "upscaling" the quality of a v3 song to v3.5 while keeping the structure and melody intact. First thing I noticed is when you extend from this clip, you can get a 3 minute generation from Suno v3.5 rather than just 2 minutes. Second, it will map lyrics you include onto the uploaded clip. So the extended bit will start where that cilp leaves off. Pretty interesting!
I don't notice an appreciable increase in overall sound quality, though, any more than I get extending the original generation of the song from the 0:59 mark. But, when I upload the final mastered version of the song, the quality of the extended bits does reflect that change.
Next, I did the same thing, but included the first 60 seconds of lyrics twice, adding
[End]
[Start]
between them. That led to a successful "restart" of the song from the extension point. Again, I don't notice much of an increase in quality over just extending from the original Suno-generated clip. But I haven't compared them too thoroughly yet.
Ahhh, and a few generations later, I am hearing some quality improvement with structure / melody intact, so the randomness is definitely a factor in how closely it adheres to the original. There might even be a use case where I can swap improved vocals into the old track. Very fascinating.
So, the first clip is my original Suno v3 Generation of these lyrics after minimal compression added in Audition. The second one is an extension of it based on the method detailed above, just as an example:
thanks for the update :), I wonder if they have any plans for it at all - at least they should be able to give us word on that
Quite a few people want an API. I added it to the known suggestions recently: https://discord.com/channels/1069381916492562582/1253766525458841610
So they know people want it at least
how to add bot to custom server?
following the link https://suno.ai/discord/add-the-bot goes to this server
/chirp
Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com
"But how can you upload "your songs" when they are registered by Suno and having their content id?"
This part is confusing. Why do you think Suno has registered your songs with content id?
Save your subscription info, email receipt, and payment info if you're concerned about it. But the only way your Suno songs will end up with ContentID or IRSC codes is if you do that yourself.
I suppose Suno could in theory do that, but that would directly contradict their TOS, so if you are worried about that you probably shouldn't be trusting them with your money.
The first paragraph is quite simple: the company is ensuring that they cover themselves from potential legal issues in case someone does use somebody else's work.
I mean, it is pretty straight forward. It says you need to have all rights/consent for the audio being uploaded and, in turn, grant Suno the right to use said audio (otherwise the company would not be legally able to upload, let alone use, the audio uploaded on their server)
Like, if you upload Billy Joel lyrics and get sued by Billy Joel based on your Suno song generated with them, Suno is reducing their legal liability for that.
That is what distributers can do, yes. But Suno is not a distributor. They don't send your songs to anyone except you.
^^ Suno already does implement an inaudible audiomark on their product.
They do not need to submit what they had generated to Content ID (which also sounds like a costly process to do). How did you even think of that was happening? It would be impossible for me to upload to YouTube if that was the case.
If your content was registered on the Content ID system, it is very likely that a bot snatched up your song and distributed it
Yes, distribute as in making it available to anyone using the service (e.g. to anyone who has the URL to the song, make it playable in trending, etc. People would mot be able to listen to the songs if thr company was not legally able to distribute the song to them on thr first place)
Not distribute as in providing to third party services
And, I mean, if you do not agree to the terms, the solution is simple: do not use the service ๐คทโโ๏ธ
It sounds like you're confusing the legal term "distribution" with the word "Distributor" in regard to music services and ContentID. This is understandable, but incorrect.
It sounds like you should cancel your subscription.
Do you mean distribute as to "provide access" or to copy the files to third parties services like spotify?
Look for the definition of "Output" in the terms.
Suno is a private company. If they decide to ban or otherwise discard your account, that is their prerogative and the process if that such a scenario should happen was mentioned in the ToS.
(Why you are particularly worried about that is another question...)
Can you tell me what exactly is the point you are trying to make?
Their point is based on an incorrect interpretation of the TOS and they don't seem much interested in being corrected. ๐คทโโ๏ธ
The first term obviously. If Suno was providing to third party services, that is the term they would use and clarify within the ToS: third-party services.
Thanks for explanation.
You explained, Tinda's explanation was more clear. Thank you. No need to continue this conversation. Wrong assumptions on my side.
how to find suno chirp bot?
so no way to broadcast suno music to the audio channel in custom server, right?
@unique stone is v4 coming soon
โ Warning logged for jungleman3333. I couldn't DM them. || Do NOT randomly ping the team. You have done this several times recently.
Dude's made 9 posts, all blatantly disregarding community etiquette and pinging people directly.
Multiple times asking explicitly about v4, and when spoken to directly, the response is "I need to finish my project, so I'm allowed to do this"
@charred merlin Yeah random pings that improve the ai for sure so random
Pinging the developers does not make anything happen faster.
@unique stone you constantly send messages in here how does not pinging the devs make any difference from you or me
/chrip
Chirp is retired, and no longer on discord. You can make music now on our website: https://suno.com
If you want to suggest anything to the team, go to #๐กโsuggestions-web. Like the bot said, the team DOESN'T want you messaging them.
I see
Since you've asked multiple times, no information about v4 has been made available, follow #๐ขโannouncements and #๐โupdates to get information as soon as everyone else gets it.
If you're not satisfied with the quality of Suno's current output, you're welcome to stop using the service and request a refund from billing@suno.com, but you're not welcome to ignore all of the community rules that everyone else has little-to-no trouble following.
As you read already, Chirp bot has been retired (in April 2024). To create music now, you need to use the website https://suno.com
Some helpful information can be found at #1151542418991239168 like the community wiki https://suno.wiki
If I creat a music and it cut the last world, how to solve it?
You can Extend a song and use the last word (though better to go a bit before that and extend from a slight break, you will have to see what flows best tor your song)
https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
Is there a way to get higher res audio quality from the final output?
I think post-processing is your only way after itโs been generated
You can use iZotope RX to clean it up a little bit, but thereโs only so much you can do if the quality is bad enough
today there is a lot of static in all of the songs. Can this be fixed? It's making great songs sound awful.
try [Genre Switch to Glitch-core]
hahaha
[Shaped Noise Percussion]
when suno give you static, be like bob ross - tur the mistakes into art
Will try that. lol great idea. Hopefully they will fix it though.
Um but on a serious note, using "Spectral Manipulation" in the style prompt of an instrumetal can balance some of it out
You can also prompt Dynamic Multi-Layering in style, and use the [] Command [Ballance EQ] if nessecary - these are my tricks
but spectral manip is the most effective
guys help me pls
basically when im trying to tape from what segment i want to continue it doesnt let me
Thanks! I'm going to try that now.
Wow, it works pretty well! Thank you so much!!
got any other tricks? I am not familiar with music production terms yet but I am taking a course in EDM music production so hopefully I will improve.
I made some songs, but I can't seem to make more for some reason
yeah you can look at my page, i have tons of examples ranging from absurd to hyper detailed prompting styles
https://suno.com/@joshuasca
follow and like too to stay updated, im always coming up with new stuff
Do you still have credits? Daily credit refreshes at 0300 UTC+0 (<t:1713236400:t> in your local timezone)
Each time I try to distribute with different distributors, the quality control come to me to proof that I'm the right holder. They mention that if it's made with AI help it's refused.
So I wonder how to be able to have it done with all these social networks.
I'll try another distributor again.
Imusician refuse to distribute on any platform if it's made with Suno.
RecordJet refuse to distribute on content ID if it's made with AI help.
oh, I don't, I didn't realize that how it worked, ty ๐ซถ
Check out other resources in the mean time at #1151542418991239168 like suno.wiki
Thank you!
Hey, I take the wrong subscription how to cancel and get refund ?
You are soooo creative. How did you even come up with all of that? It's amazing. Brilliant! I like your music a lot.
I have a permanent brain injury, i lost alot of functionality to half my brain when my skull was fractured - but the other side compensates for it
so I'm right hemisphere dominant
Blessing and a Curse
Amazing...so sorry you got so badly hurt. I'm glad that you are enjoying your music creation. The right hemosphere is full of secrets to explore. People spend whole lives trying to operate from that place because they say it can connect you to the divine and divine inspiration. I hope you continue to heal.
'I actually wrote a song about that sort of meditation, called '3 hours 'til morning'
Hello. Me new here, So, what is "RawVocalSamples734"? Are there more RawVocalSamples, or is it just an autogenerated name for my channel?
As I said above, I use Landr and they know I'm using Suno and it hasn't been a problem. You can find my songs on Spotify, Tidal, etc. I'm sure there are others but that's the one I use.
What do you folks use to clean up the muddy and staticy sounding songs that you like other than the quality? I want to clean up a few songs so I can hear them clearly without all the fuzz and annoying static.
That's the profile name that was randomly generated for you. You can change it with "Edit Profile" from the menu on your profile pic.
thanx a lots
It's not super easy at the moment. Or even fully possible if your source is distorted enough. But you can try separating the stems using Ultimate Vocal Remover and then try some remastering in Audacity or Audition. But so far all my experiments with this have resulted in something that, at best, is about equal to the wav originally generated when you do some mastering on it. For now, try extending the song you like from various spots and redo the verses to see if you can get a better version. Also see my above posts about uploading the original, from this morning.
hello, I was curious. I had to extend a song because it hit the 4 minute mark before ending. Is there a way to merge the 4 minute song and the 2 minute extension so that it publishes/downloads as 1 song?
On the part 2, click ... > Get Whole Songs.
It will combine the two parts for no credits! https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
perfect, thank you!
Thank you! I will have a look.
I don't know why it does this sometimes when extending, but it likes to skip down to the next tag. It kept doing it no matter what I tried, so I put some garbage lyrics in the front and then it skipped to where I wanted ๐
i use custom and instrumental mode in my music, style and title also inputed, but i still can't create the music.
Usually that means you're extending from a weird spot, so it's like trying to add a junction to the tracks in a tunnel.
Several ways to overcome that and you've found a reasonably good one. Alternately, putting the last line of lyrics just before the extend can help, too, depending on how recent they were.
The way I think of it is, you might want to branch off mid-tunnel, but unless you give the model context it's gonna try branching after the tunnel. Whereas if you supply the context it now knows to branch before the tunnel. That's probably a little simplistic (definitely), but it works for my brain.
Yea, it probably was because I was cutting right in the middle of a musical phrase. I wanted to keep the beginning of the phrase but not the whole phrase.
how do i get a song to not have drums?? I have tried ethereal, tranquil, ambient, slow melody, quiet in the style. can I add anything in the instructions in the song to make it stay calm?? like:
[chorus]
[calm]
This is a battle I have fought with the ENTIRE time I've been playing around with Suno. No matter the style prompts I give or tags in the lyric area, I still get some sort of drum beat and a music style that sounds like "synthwave."
It might require "rolling" a few generations for experimentation, but you could try placing an extension point in the middle of the beginning of the phrase? Depending on how "strong" the pattern is, and if it takes context from anything else in the song, the beginning might naturally resolve similarly to your preferred generation, except it wouldn't be so tightly restricted to how it can extrapolate from that melody.
If that makes sense?
The issue there is that "calm" is wholly subjective, and apparently Suno doesn't recognize "negative" prompts (so "no saxophone" is interpreted the same as "saxophone"). "Calm" means no drums to you, but some people find drum & bass techno to be calming, some people find the simple drumlines in Beatles songs calming, etc.
However, if you only want vocals, we've discovered that using "a cappella" works as a style prompt (but not the anglicized "acapella", interestingly enough), because Suno is less likely to misinterpret "a capella" because there's a shared understanding of what that term means.
Similarly, I wonder if there's a musical term that indicates "no drums", or if that's too specific of an exclusion to have it's own shorthand term (like "without guitar" might be specific enough that it's just noted as such instead of expecting musicians to know every italian or latin phrase ๐ )
Anyone have any suggestions to make [spoken] work at different sections of a song? I can't get [spoken] to work at the beginning of a song, and rarely somewhere in the middle, but it always seems to work near the end.
so how do i get it to have no drums?
What's your prompt?
ethereal, tranquil, ambient, slow melody, quiet, piano, harp
Out of curiosity, was my previous post unclear / confusing, or are you simply looking for exact instructions to get a song without drums?
from a week ago, i just can't make SUNO to sing like 80s way of sing, now all my songs even if they have an 80s style intro, its start singing with that pathetic and synthetic way of sing from 2000s to nowaday
is there a way I can get it to stop making my voice robotic? I want to do a recording and it sounds awesome however the program puts it, minus the robot overlay on my voice lol
(hope this is the right place, not trying to derail any current technical issues/discussions!) Does anyone know how to copy the voice or timing of a song I have previously created or is it random every time? Even if you copy the prompt and lyrics? I wanted to change a couple of lyrics and see how it would sound instead. If this is in an FAQ I would appreciate any direction! thank you!
i'm not sure if there is a better way to do it but when I want that, I do an extention of the original song just a bit past the first couple timings and first few seconds of voice. it should pick up the same stuff
I will give this a try tomorrow! thank you!
Send an email to billing@suno.com but expect that with recent events there may be a backlog on emails so it may take a bit to get back to you.
I'll try flagging it for attention but understand that it may be a bit for a response.
i would love to have negative prompts, if i can do wishlist suggestions here
More appropriate at #๐กโsuggestions-web (better yet, upvote for the suggestion at #1252298754787377225 > https://discord.com/channels/1069381916492562582/1253341197917884530 )
done and done
Is the site down again?
how do i get suno todo instrumentals with 3.5? its not working the way it uised to for me
try using [instrumental intro], [instrumental verse], [instrumental], [guitar/synth/whatever solo] and various versions of those in your lyrics box
im already doing [instrumental intro/outro] and [instrumental] and it does nothing
that's frustrating. what genre(s)?
Nu Metal, Melodic, Harmonic, Aggressive, Rock, Rap, Angry, Punk,
try taking the rap out of your prompt. rap is super lyrical so you're more likely to get hallucinations out of that and nu metal is already rap-rock a lot of the time so its a little redundant i think... try that a little and see if that helps
yeah its still skipping the instumental prompts
took out rap and nu metal
aswell to see if thatd help
that's frustrating, i guess sometimes it's just roll of the dice?
been fairlly consistently ignoring the prompts for a few days now
maybe you made every single instrumental nu metal song that can ever exist already
you went too far!
i can only download the song in mp3 format but not in WAV, i don't know why.
Hello question, is there a way to upload more than 60s? or is there a way to combine uploads? @final hazel
Where does it state that they can do what they want with your uploaded music?
Why should Suno use your voice and for what? Once you have paid, you also have all rights to your sounds, beats etc.
Is it better to use 3.0 or 3.5 for instrumental music ... I didn't get the best results with 3.5
Hello! Please email billing@suno.com if you are unable to find the answer to your question on this page.
but they wouldnt be training on the ai generated output , I am uploading professionally produced WAVS, so am I giving Suno rights to train their data on that? its an important question that needs answering by the team. Why would I upload my hard work compositions and productions only for them to the train on that. cheers
I dont believe my query is quite covered in the terms.
any ideas?
There is not
okay then i have to wait for upload till 200 secs
I guess, they deleted my post
i got a problem, i can't download Wav file anymore, only mp3. each time i tried, it failed.
I heard this happening before and that clear cache "could" help
okay thank you, i'm gonna try it ^^
It works, thank you so much !!!! I was dealing with it for 4 hours now
Sorry to hear that.. but still: VICTORY ๐
Thereโs a gap between data experts and non, so itโs hard for regular people to detect the level of data collection going on online
But a lot of it is even you can use ai to find peoples vocal range that you like and apply it to your song generated
Cat singing with appropriate music
Yes try [end song] or [end song instrumental]
Wow I happy someone has gone down that rabbit hole and come back to share. TY
I'd assume them to be rests
Would it be possible to upload a song of mine and suno add the voice without retouching the music?
While Suno is capable of reproducing the music, it's difficult to gain complete control over what the generated extension will be.
In my experience, I've uploaded a ~60 second clip, and the same motifs and musical passages were repeated in the extensions, but they each had their own levels of differentiation-- some were MUCH closer to the original uploaded audio and almost acted as natural extensions, though often they included stylistic variations.
That being said, you absolutely can upload your own instrumentals and ask suno to add lyrics onto the extension!
Hello together;
Do you have experience with controlling male and female voices? I wanted to create a song where the first verse is sung by a woman, the second verse by a man, and both voices sing the pre-chorus and chorus together.
So far, I have only managed to have the verses sung as described above occasionally, but the parts that were supposed to be sung together were just sung twice; once by the man and once by the woman.
Here is what I have tried so far:
Lyrics:
(Verse 1: Female)
(Verse 1: Male)
(Pre-Chorus: Both)
(Chorus: Both)
(Verse 1: Female vocals/voice)
(Verse 1: Male vocals/voice)
(Pre-Chorus: Duet)
(Chorus: Duet)
(Verse 1: Female vocals/voice)
(Verse 1: Male vocals/voice)
(Pre-Chorus: Duet - Male and Female vocals/voice)
(Chorus: Duet - Male and Female vocals/voice)
I have also tried different variations for the "Style of Music".
The best result I have gotten so far with V3.5 + Full Lyrics is this:
https://suno.com/song/73eef78b-fa5c-442a-972e-152c1dd3577a
Perhaps you know some users who use duets more "controlled" in their songs, and one could read up on that a bit.
Greetz
Epic Rock song. Listen and make your own with Suno.
anyone else having issues with it giving male vocals when you ask for it?? it was doing it an hour ago and now it refuses
Is it possible to get credits back? I wasted almost 500 credits because suno isn't following my style instructions anymore. The last one that worked was "rap" and every song after that was also rap even though i prompted "Opera", "Country", "Blues" and so on... the instrumental beginning is always ok but when the verse starts its just rap!
Any tips on making a chorus hit hard?
Not directly, no. You can email billing@suno.com, but there's no guarantee you're entitled to anything back.
Also, you can report those songs through ... > Report > Bugs > Ignores Style Structure to try to prevent further issues in the future.
I reported every single song
It seems like suno involves the lyrics too much within its calculated styles ๐ค
Any way of getting custom sounds to add instruments? I made a acoustic riff and uploaded it, and would love if it had piano and other instruments, but looks like it is really heavily influenced by the sound, so it is only giving me acoustic guitar songs no matter what styles I've tried to put in
Don't try to generate the entire song at once if you feel that something is overpowering the style.
If you want a rap song, take out all lyrics and other style prompts & generate something that sounds like a rap song.
Then extend from that and THEN add lyrics into that extension so it already starts off like a rap song.
Or replace rap with whatever style you're doing. Break the song down into smaller and smaller parts that you can more easily control, instead of throwing your masterpiece into the prompt window and expecting Suno to juggle all of it-- work collaboratively with it!
โค๏ธ
I will try that. Thank you very much.
Good morning ,
I just joined the server
I would like to know how to delete your suno account
(I'm French but I use Google translate)
Not sure off the top of my head, I believe it's through request to billing@suno.com or support@suno.com
Please make sure you take steps to cancel active subscriptions beforehand though, to make sure that you're not charged accidentally while waiting.
mine is also doing this right now, i think the bot's having a bad day
how to do it via discord?
Not an option.
Le service sans frais de Google traduit instantanรฉment des mots, des expressions et des pages Web entre le franรงais et plus de 100ย autres langues.
N'est pas possible via Discord
aaaaa I can't do it any other way, right?
merci
how to earn credits other than inviting people to register
The report?
Seulement#๐ณโmini-polls, abonnement, ou achat supplรฉmentaire (only polls sometimes, subscription, or purchasing extra)
aaa I hadn't seen the message that replied to someone
Does Suno use an API?
No, there's scam websites that will try to steal your money while pretending to be Suno, though. https://www.suno.com is the only official way to access the service
Yes! For example;
-
custom, instrumental, style prompts = "hip hop, 1990s rap, los angeles" make a handful of these, and remember we're not trying to get the whole song finished at once, we want to find THE COOLEST intro
-
after finding a neat hip hop beat, place an extension point a few seconds before you want things to change, like adding lyrics or extending a particular cool sound (there's times I've extended from the first 2-3 seconds of a song because i REALLY liked a particular noise at the beginning)
-
once you feel comfortable with your hip hop sound, and you know "what the song will sound like", try playing around with different extension points for inserting your lyrics to find where it sounds the best.
If you throw everything at "the robot" at first, it will have no idea how it's supposed to juggle everything, it'll drop most of it, and you'll likely end up with a mish-mash that sounds like "generic pop", maybe with a couple choice genre elements thrown in.
But using this method, you're breaking the song down into smaller and smaller parts that you're able to comfortably control, then adding things onto it, like a juggler who starts off with 1 ball, then 2, then 3, and eventually chainsaws ๐
(let me know if I can simplify anything, I tend to speak too strongly "in pictures" and metaphors <3)
@charred merlin Youโre incredible
Thank you for committing so much time and energy helping members! Always a big deal!
okey just realised something. if the structure of the lyric doesn't fit the genre it is more likely to strt hallucinating
so if it starts to hallucinate alot your genre selection is likely wrong
Thanks dude, that means a lot ๐
Suno and the supportive, collaborative community mean a great deal to me, so I'm happy to give back โค๏ธ
It shows consistently! Lucky to have you here
hello i have violin in the beginning of my sound, i would like him not to use it anymore for the rest, any solution?
And you checked to make sure you were working from a clear create tab? Sometimes I forget that I have been extending something so I don't hit "clear" I just change the prompt and lyrics without scrolling down to make sure there's nothing in the extend spot.
try things like [style change] or [switch to keyboards] and change the style prompt too. Can't guarantee it'll work but this is a good starting point.
ok thanks i look
๐๐ ๐ best advice here, don't miss it! ๐ ๐ ๐
(Slow, doom-laden intro with distorted guitars) what's the best way to convert this to a tag?
oh and thanks for any help
too bad it doesn't work. ๐
Hi, I wonder if for instrumentals, the theme is taken into account for anything apart from the title. For instance, if I go into the custom mode, I have only the option of choosing the style and the title, but I doubt the title is used to generate the song. For instance, would a "Snowy Christmas" title create a song with Christmas typical sounds?
Doesn't seem to...
https://suno.com/song/a76a0a7f-049b-4a44-87c4-db6b2e41ae75
Dark Horror, Movie Score, Eerie Slide Guitar, Haunting Lead Guitar, Tuba, Polyrhythmic Trap, Not Dave, Or Kevin song. Listen and make your own with Suno.
I would say it doesn't!
NExt I would try extending the song until you get 2 minutes with no violin but still feel like the melody is right, then extend from there. I've found a lot of success by extending and using
[End]
[Start]
and variations of that to create new versions of songs that stick to the original. And if you've gone long enough without violin it won't even know to use it. v3.5 only can look backward 2 minutes.
ok thanks i look this
so when I put an end to it, instead he improvises on part of my text ๐คฃ
Lyrics didn't help ๐ https://suno.com/song/6336a16f-520e-4cc3-9fa4-d8313e6b0685
Dark Horror, Movie Score, Eerie Slide Guitar, Haunting Lead Guitar, Tuba, Polyrhythmic Trap, Not Dave, Or Kevin song. Listen and make your own with Suno.
Just keep trying. I'm at the stage now where I'm such a perfectionist I spend 1000 credits on pretty much every song I make. But it's worth it because you can stumble on some very awesome results if you give it enough time. Infinite monkeys on infinite typewriters, and whatnot, right?
I'm up to 1100 just for her, and I'm planning to put in a lot more as it's a text that's close to my heart.
Well, the end / start thing is pretty reliable, but as with all these tags, none of them are 100% reliable. Just try again. ๐
I even tried to clear my browser data and cookies โฆ all the same. Itโs the third time suno is doing this and like I said: I burnt around 1500 credits stuck in one genre no matter what prompt
If you've tried following this #๐โsuno-school message, please share some more details so we can help figure out what's happening; provide the lyrics, style prompts & what some of the generations are
Ok I will tryโฆ but Iโm stuck in a โhip hop loopโ so Iโll try a different genre.
I've never heard of Suno actually being stuck on a specific genre, you shouldn't need to clear your browsing history or cache.
It sounds like you're trying similar methods that are all giving you the same results, so if you share what you enter when making a song, we can help try to figure out how best to create what you're looking for
Ok. I took German hiphop lyrics with Arabian words in it. I tried different styles. But every prompt I entered lead to the same result. It always began with an instrumental part of the correct genre I prompted but when it came to the lyrics it was always hip hop.
Easier to help if you share specific examples
Ok give me a second. I tried the extend method with small steps and I worked for a few seconds before it actually turned back to hip hop.
Opera, Chello, Tenor, High Male Voice, Choir song. Listen and make your own with Suno.
Did this just quick and dirty
And just to reclarify-- what roughly are you wanting to create?
An opera
An opera with rap over top of it, or do you just want opera?
Just pure opera
In your lyrics, you're using the metatag [fast rap], which gives Suno the suggestion to make the next section "fast rap"
Sorry that was my fault. I tried it without any meta tags before and it always falls back to hip hop
I will give it another try. Wait a minute
Including metatags that are related to rap are definitely going to push it toward a hip hop feel. I'll try a quick demo to see if I can use rap lyrics with opera, since I've never really tried that
I have a question about hooks. Can I put what it say's in hooks?
Yes I know all that but the behavior is strange since 3.5 for me. It happened three times now and even without any meta tags. I even tried completely new songs with completely different style prompts where, at some point, suno falls back into the previous genre
I might be confused, but if you want only opera, wouldn't you use [opera] to keep it away from rap/ hiphop?
Yes. I even tried without custom lyrics and entered โRussian polka about cookiesโ where suno suddenly starts rapping in the chorus
v3 is better at giving a versatile range of styles, while v3.5 excels at the actual song structure itself, but can lean toward "generalization" of songs (pop, rap, etc) without a strong musical structure for it to follow. For this reason, I often find myself starting songs with v3 and then extending with 3.5 (but 3.5 has given me some awesome introductions to tunes, not to write it off)
Thatโs a great idea! Will try that too. Thank you
good morning guys
guys I have the portal to distribute with Believe and I'm trying to upload some material taken from suno
Aaaand the same behavior as alwaysโฆ after several new tries it suddenly worksโฆ I really donโt know whatโs up.
let's see what happens?
Do you think they'll throw it away or will it be released in stores?
Have you contacted Believe support, or read their Terms & Conditions?
I have a contract with them for distribution
no
You'll have better results asking their support or reading their Terms & Conditions than asking a different company's discord
ah ok
For instance, Soundcloud doesn't allow you to distribute AI generated music, but some people still manage to sneak content past them.
what genre i put in v3.5 to get similar results?
but what if you don't say it was generated by artificial intelligence?
Then you're breaking the terms & conditions that you agreed to, and they'd be within their right to take your music down if they discovered it.
Before signing up for anything, it's best to know what you're getting into โค๏ธ
I have read the entire contract and I don't see that there is a paragraph that talks about artificial intelligence
@charred merlin
how do I use the uploaded audio refference in a new song that I'm about to generate?
In terms of user interface, the uploaded audio is treated like a Clip 1 of a song that you then Extend. There's lots of nuance and extra detail (it makes a big difference whether your upload is very short or 60s long) but that's the core flow.
Is there an effective tag for making a Verse only vocals. Or at least mostly vocals.
โA cappellaโ seems to work better than โacapellaโ
has anyone successfully made phonk ?
Any tips for remastering Suno songs?
IIRC, for phonk, you are better off generating an initial track in 3.0 then Extend the one you like into 3.5
how long does it take for an account to be deleted? i sent a deletion request 2 weeks ago, and my account is still up
why is paste disabled now???
youre expecting people to type everything out again and again?
Is the credits of subscription plan will reset in the next month when it's still available? For example, when I still have 500 credits in this month, will it reset to based on subscription plan? Or will it be added?
Reset, credits from subscriptions do not carry over
Ooh, I see.
So, it means that the subs plan users should spend the credits within that month, right?
Exactly, yes.
You might as well spend them
Note though, that any bonus credits, either bought or gifted to you from the devs do not expire
Just the ones you get through your subscription
Well, thanks for the info. I'll decide to either upgrade to subs user or not.
Not a problem at all.
Hey guys.. You might want to see this! HOW TO MAKE SINGING AVATAR WITH HEDRA.COM
https://youtu.be/KOYoEvvtffE?si=B8aa38DMqpDqjJKu
Dive into my latest video where I showcase my workflow for creating captivating singing avatars for your Suno songs. This tutorial provides a behind-the-scenes look at how I integrate AI tools to bring music to life with dynamic avatars.
The song used to show can be found here:
Suno: https://suno.com/song/07133ad4-782a-4076-9f57-936c0a2752c1
Yo...
That sucks and waste credits
It's a solution though, since if you get the style of music you'd like doing it then it isn't exactly a waste, is it?
I actually don't see how it's a waste at all
Can we use kitten faces? ๐ผ
They say it will come soon
lol. w000!
Speaking of cats... LOL!
I discovered a funny little trick to get a layered (and often robotic) voice effect. (Using the [robotic voice] tag consistently results in something almost incomprehensible to me). So anyway. If you add "cat sounds" or "cougar sounds" in the style prompt. ๐คฃ i've replicated this effect thru many songs that have nothing to do with techno or robotic sounds, but heres a test illustrating my point.
Cougar sounds, detroit techno, cat sounds song. Listen and make your own with Suno.
YO another workflow from 3Daizy? Youtube is doing well tonight ๐
Ideally you would want to grab the .wav and load it into some type of daw. There are options available, Iโm currently building Soniqs DAW (located in my name) that Iโve implemented some amazing feature based on what Iโve learned the community would like to see. Itโs a web based DAW that is cross platform with no installation. Thereโs a demo live on the website you can try. One of the newest features Iโve implemented is stem separation. Once you have separate stems you can add effects such as adding reverb / delay etc, and tweak the EQ of the individual stems. After that is done you can have a more polished track.
Can anyone suggest a way to get a "shortened" extension to a song. I worked up something that turned out pretty well, but it cut short, so I extended from a decent spot; it flows well, but the extension goes on and on and on despite me throwing in metas to [Fade Out] and [End] it just keeps on going and ends up sounding like poo as it ends.
Secondly, it has a spoken point that sounds decent, but always ignores my tags for [Female Voice] [Female Spoken Word] and multiple variations I can think of to tell it to speak with a female voice. Keep reciting the words in a male voice.
If you're trying to change the voice mid lyrics, it's difficult. If you search Duets there's some practical formatting advice but it's not reliable, still needs luck. You can try to switch in an extension with the style prompt and brute force "female, female, female" though.
Today the site is having trouble crashing and crashing all day
I did try to brute it by doing just that, changed up my style prompt to say "spoken female, female narration, female voice, gentle female word" - still get a dude reading the text. The rest sounds melodic still, matching core theme of the original
Also, FYI, the whole song is "sung" with a female voice, but it wants to read text with a male voice, despite all tags calling out [female voice] [spoken word] before all "read" text.
If the song doesn't ever change voices before that time, the longer duration the clip is you are extending from, the more resistant the voice will be to a change. What this means practically is that if the Extend From time happens to be very close to the start of a Clip (5 to 15 seconds, maybe) then the odds a higher you can change it. I just read your message and this is not the situation here, but I'll leave my response up, as it covers that case.
possibly the spoken' or 'narration' tags could be correlated with male voices. but it could just be a quirk of that particular song, sometimes that's just how it is.
The male voice is honestly just a small annoyance, thought it DOES fit okay. My main concern was actually getting the extension to NOT trail off after my tags to stop the song/music.
I'm trying both v3 and v3.5 - keeps rambling on after [Fade] and [End]
You can add 'end' to the style as well
I'll try that.
Does Suno has a developer api for mobile app?
HA HA - It did the opposite now. Silly thing. It SPOKE the final lyrics in a male voice, then SANG the spoken part in a female voice. This AI is goofy.
I've had a lot of success (65%-70%) with extending the song, setting the "extend from" timestamp to where you want it to end, but go into custom mode & in the lyrics box add:
[Outro at 00:15]
[End]
Most of the time it makes a 30-90 sec extension. 20% of the time I get a 15-30 sec extension.
That's a good suggestion. I'll give that a shot too. I've already burned nearly 150 credits trying to get this to work, because I REALLY want this song work out.
Here's the final... I think. I like how it "finally" turned out. https://suno.com/song/cda2e18d-b60f-4cc2-97d3-cd501f35b9db
Dreamy medieval ballad, female vocals and choir sung chorus, no percussion, spoken intro, spoken bridge, spoken outro
song. Listen and make your own with Suno.
Hi guys, I'm having a lot of fun with this fantastic and promising creative AI, thx for the great job!
As an undergraduate in CS, I'm very curious about the techniques behind Suno and planning to devote my future career in AI for audio industry, either Black-box for public or White-box for producer.
Any suggestion on how can I study SOTA audio generative models and something related to them (preknowledge etc.)?
Suno doesn't talk about their music model, but you can look at the source code to Suno's TTS model Bark which is open source. But it's just the source, there isn't a paper or training code, so I think a better place to start is Meta's MusicGen model https://github.com/facebookresearch/audiocraft/blob/main/docs/MUSICGEN.md
Does anyone have tips for generating good spoken word with v3 / v3.5?
Alan Watts type of style
There definitely is trained data in that realm. This i.e. is quite close to the Alan Watts / Motivational Spoken Word .. but it still is hit and miss with the prompting and could be cleaner
fast spoken word, rhythm, studio ghibli, japanese instruments, rice fields, spoken word song. Listen and make your own with Suno.
Thanks for the recommendation!
it will be added ๐
@craggy delta ๐๐ป๐๐ป๐๐ป
Mark Duplicate Playlist Songs
this was driving me crazy so maybe somebody else finds the little script helpful https://github.com/JonathanFly/chirp-studio
maybe I should have tested it more than on one playlist, but going to sleep, probably works fine
I want to make music for exactly one minute, but putting "Duration: generates clips up to 1:20 minutes long" or "just 60 seconds" does not work. Expert tell me what to do.
Not yet possible, but hopefully in the future.
Do you know about the Extend From feature, if you can generate a song that ends in 60s or 1:20, then you Extend From that time (in the middle of the clip) and try to end the song early, even the original clip didn't.
I understand. Thanks.
Is it practical to make a song and then use a DJ tool or something to increase the tempo and keep it to 1 minute?
If you are editing the song yourself, it's not too hard to Clip of a song that is self contained in about a minute. It's just hard to do that in a single official Suno song link.
If you need it 100% exactly 60s, then yeah, probably have to use a DJ tool
Thanks for the clear answer.
Are there any new ways to end your song without wasting too much credits to get a short ending?
I seem to be experiencing a large-o problem with this at the moment
there is a drum called " lali ", but i can't use it because there's an artist called " lali "....
You could post that in bugs channel, not sure if they are fixing collisions like that, but worth a shot
Hi, anyone have suggestions for controlling the length of an instrumental? Like I need one minute and not 4