#📚┃suno-school
1 messages · Page 24 of 1
Ooooooh thank you so much
@pure bay no it doesn't sorry
What do you mean? Who doesn't?
Original:
Part A of the lyrics
Part B of the lyrics
I trimmed at the time when A ends, to expect:
Part A of the lyrics
But in the whole song that I got, the lyric panel still displays:
Part A of the lyrics
Part B of the lyrics
Now you understand?
Try cutting a few secs.
not cut but you know
like minus a few secs and if anything is missing add it on the following section
I meant that I just wanted to erase B cuz it is the wrong lyrics. I cut at 1:09 but B still appeared in the lyric panel when I got the full song.
Oh yeah! Other social media sites can be just as bad. I don't know if the problem is down to the money-mad owners or the growing millions who just want a voice.
What I find strange is why people sign up for a site related to a particular interest, say Suno, and then talk about everything but!
It's just me going off on another old man's rant on a site that is about Suno! 😏
Have a good one!
one silly questions: How do you change the vocals? Is there a document or a guide to learn how this works??? If you only use the free version, you have only a few tries per day to figure it out.
I don't want to reuse prompt I want to do a remix, doing a remixes of my own generations was a way for me to turn a mediocre song into a banger, I don't understand that it's cancelled

I periodically search posts of Discord users who I remember had some very interesting prompting. Not very elegant but can find good stuff that I would have otherwise missed
I understand. This is a current limitation. It would be good to have a feature to customize the final lyric output.
Plus, are such 2 metatags eligible?
123.05BPM,
B minor.
BPM No. B minor usually no (but maybe if you are very lucky.) But other more general music terms do work better: Legato, glissando, etc. My guess is "Minor" probably doesn't work as well as words that mean "sad" but I am not sure.
BPM is a difficult thing to control. Best you can do is use other words that indirectly influence BPM (by genre, words meaning fast or slow, specific music terms).
I used them to try to exactly stipulate the song to be fast at that rate, but failed. Do you know what metatags I should use for fast tempo exactly and quantifiably?
I am not very knowledgable about that specifically. What genre of music are you doing?
123.05BPM, B minor, Male, Very Fast Tempo, Pop, Electronic, Indie, Chill, Mellow, Soulful, Sad, Melancholic, Relaxed.
"adagio" "allegro" "fast" "slow" "energetic" these are words I use sometimes, that can change speed.
For EDM or Electronic music, my guess is the best prompt are EDM specific music genres that are known for having specific speeds.
123.05BPM equivalent to which one?
I don't know EDM music very well, sorry. I bet somebody else might know. There are very many EDM sub genres!
But using the number 123.05BPM, that willl not work. So I would remove that.
Oh, one tip: If the speed is too slow you can try something like "...Sad, Melancholic, Relaxed, fast, fast, fast, fast"
repeat the word
I tried "fast" but still not fast enough. Plus, I tried "Male", "Male Vocals" and "Male Singer", But still generated female voices.
For example: "Male, Fast, EDM, Electronic, Indie, Chill, Mellow, Soulful, Sad, Melancholic, Male, Male, Fast, Fast" I removed "relaxed" because that sort of also means "slow"
"mellow" and "chill" might also sometimes make song a little slower, my guess.
But "relaxed" is the one I think most likely to make it slower
Remember you can browse other people's songs and listen - if their song is the correct speed, you can try the same prompt. That doesn't cost credits.
https://sonoteller.ai/PYRjpF_7VHI
In this webpage, there is a panel "Themes", comprising:
Politics (80), Injustice (70), Oppression (60), Unity (50), Action (40)
Can these words be eligible metatags?
Song analysis, lyrics analysis and music analysis. Understand what your favourite songs are about. Get accurate and in-depth music autotagging with SONOTELLER.AI, a tool powered by artificial intelligence including a comprehensive summary about the song lyrics and music. Analize your song, analyze genres, analyze moods and enhance your musical u...
"Politics (80), Injustice (70), Oppression (60), Unity (50), Action (40)" = My guess is probably these words do not work well. But I am not 100% sure, sometimes strange words do work as music style. But low chance of success.
"Defiance (80), Frustration (70), Urgency (60), Rebellion (50), Solidarity (40)" = These are "mood" words from the lyrics, emotions. But this list is a little strange choice to use to describe music style. However, these have a higher chance of maybe working. Small chance of success, but higher chance than the first list. Maybe worth trying at least.
"Sad (80), Melancholic (70), Relaxed (60), Reflective (50), Intimate (40)" = These are "mood" words from the music analysis. These probably do work. These are commonly used to describe songs.
I would recommend removing "pop" because "pop" is very generic music genre word.
OK. Plus, I tried "Male", "Male Vocals" and "Male Singer", But still generated female voices.
How to solve it?
Two different methods:
- Repetition. "... male, male, male, male, male" repeat many times.
- Start with different lyrics that do not always generate female voice (sometimes the lyrics has an effect on this) and then use "extend from" feature to continue the song with your real lyrics as Clip 2. You can choose a time very early in the song, as long as after a voice voice stated singing.
"dirge" will give a male voice, but also "dirge" is a style of music so it may change too much.
I saw somebody say "German" gave a deep voice for Electronic Music, even if not using German lyrics. But I did not try it.
What about languages and dialects? metatag them in style panel or lyric panel?
Just use the correct language in the lyrics. If you use the name of the language in the style, what usually happens is the music style is changed. Sometimes that is okay but often it is not what you want. For example, if I put "irish" in the style what happens is I also get more "irish music" with fiddles and bagpipes. So that is only helpful if you want that type of music.
For dialect, that is difficult. It might just need luck if you need a specific dialect.
For dialect, maybe you can try using a 'throwaway' clip 1 where the lyrics are very clearly in a specific dialect to make it easier for the AI to guess correctly. So these lyrics are just filled with unique dialect phrases. This is not ideal because the beginning of the song, the 'throwaway' clip, will need to be removed in an audio editor later. But it can work. However it may only work for some dialects.
But not all dialects have specific words. Like "East London English"
Plus, can I repeat "male, male, male, male, male" in the lyric box to spare room for the style box?
Also, can I write:
[Fast]
[Fast]
[Fast]
in lyric box?
And even by titling the lyric directly, making different parts of lyrics sung at different speed?
So it doesn't work all the time, but here I just asked ChatGPT to write the most over-the-top East London English (Cockney) accent:
https://suno.com/song/04ed8df3-c48f-4331-bf79-6eca0c4f32d4
it worked, a little maybe? let me try a few more
It's not Cockney accent, I can discern.
yes it's hard to get a very specific dialect. I was curious how hard 😄
Plus, can I repeat:
[Male]
[Male]
[Male]
in the lyric box to spare room for the style box?
Also, can I write:
[Fast]
[Fast]
[Fast]
in lyric box?
And even by titling the lyric directly, making different parts of lyrics sung at different speed?
Yes, this can help, but the effect is much weaker than the style box. But it can help.
closer?
https://suno.com/song/47274bfa-2a15-4963-b82e-6ea02e60cd2e
https://suno.com/song/5478cbad-9ca4-4fa3-b39c-5a372132cdce
It is hard to get a specific accent. This is "Suno Hard Mode"
Last two, gotta get back to work. I was curious how well the idea works:
https://suno.com/song/8c9d203f-4264-45d8-b87b-b9f5d8844849
https://suno.com/song/27225749-f804-4b35-bfc0-652c5962ef00
How did this song end up so British??!??
https://suno.com/song/5be76253-bd58-4930-8265-4d768ed5069e
"British rap" is all you need?
I was trying to do it with lyrics but maybe "british" in style works better, at least for that accent. Keep it simple.
No, I tried Taiwanese. But it's too complicated for you to understand. Thus I used Cockney as an analogy.
I understand. My examples were also an analogy, to show the concept, and maybe the same method could work for another dialect. The reason I created so many examples was to satisfy my personal curiosity about how well this method works.
I try to achieve something similar - control voices, dialects, etc. Blasted so many credits in Suno and Udio without any luck. 😦
I can do it in Bark pretty well, but it is harder in Suno. In Bark there is no 'style' prompt so the effect of the 'lyrics' (the bark text input) is much stronger than the lyrics in Suno music.
Wait, so Bark is better than Suno V3??
Just for the record, there is also a variant of this method that can be used for foreign accented languages. For example, let's say you want English but sung with a French accent. Then you can prompt Clip 1 in just French, or maybe verse 1 in just French, then swithc to English for rest of song.
Bark is a text to speech model (also made by Suno), so the same concept of accents and dialects applies, but it is not used for music. (well Bark isn't supposed to be used for music <#🐶┃bark-showcase message>)
I tried both:
[Male]
[Male]
[Male]
in the lyric box,
and "Male, Male, Male" in the lyric box, but still generated female voices.
Can you link me (or dm) a Suno song you made where the music itself is close to what you want to make, but the voice is female instead of male?
This is what I want, kinda. Have you tried to achieve something similar with Suno V3?
Fast, Fast, Male, Male, B minor, Pop, Electronic, Indie, Chill, Mellow, Soulful, Sad, Melancholic, Defiance. Frustration song. Listen and make your own with Suno.
I need to step away but I tried a few quick, I think "German" might actually work better than "Male" ... 🤔 But I don't understand the Taiwanese so maybe the singer has a strong german accent or makes no sense. Very small test probably just RNG but interesting.
https://suno.com/song/71264a8c-02ca-40ca-8e4c-7631f7151709
Here is all my tests:
https://suno.com/playlist/bc8d78b2-6493-446d-8b39-bb0b65637708
German, Fast, Male, Male, Male, Electronic, Indie, German, Mellow, Soulful, Sad, Melancholic, Defiance. Frustration song. Listen and make your own with Suno.
I should not have used the whole lyrics in the "Continue From" tests, too long for just 1:00.
Much obliged🙏🏻🙏🏻
Btw, do you know which user is the tech employee of Suno in this server?
But at last in this test the deepest voices had "german" in prompt.
Anyone listed in "Team" Is a Suno employee. I think it is a pretty small company so everyone is "tech" to some degree. But "georg" and "bigtony2651" are the ones who most often talk about the most tech stuff in this Discord chat. If you catch them active in chat can ask questions (usually on EST time, maybe mornings) but please don't @ them directly when they are not actively chatting with folks.
Aren't vaughn and z?
Yes they are Suno employees, I don't know any more details. During workday EST time you will see many more people on Team online.
There are only 2 members in the group "team", as I see.
It is 5 AM Est time, almost everyone is offline and sleep, yeah
I used "German", but still female voice:
https://suno.com/song/9acd7291-feb7-4a4b-b289-4e663fee0283
An examnple of a specific prompt in the ones that worked was "German, Fast, Male, Male, Male, Electronic, Indie, German, Mellow, Soulful, Sad, Melancholic, Defiance. Frustration" so it had German and male. But also you can see on the playlist that even with this prompt there were many still female. But the deepest voice songs did have at least one "German" in them which I thought was interseting.
Probably "Defiance. Frustration" are not having an effect, or the effect is very small. I think other style words that will work better than those. Maybe "defiant, frustrated" (slightly different phrasing as adjectives) are worth 1 or 2 tests, but my guess is they have small effect at best. However sometimes I guess very wrong...
BTW, in the playlist I linked, please free free to "Extend From" and build on any of those songs and make your own song from any that you like. By using the "Extend From" time feature you can start from a generally good voice/music and try expanding out to rest of song to "dial it in" as you want.
I tried "Fast, Fast, Fast, German, Male", but still female voices:
https://suno.com/song/800f45db-b4d6-4dd4-86d0-39059266fae9
Fast, Fast, Fast, German, Male, Pop, Electronic, Indie, Chill, Mellow, Soulful, Sad, Melancholic, Defiance. Frustration song. Listen and make your own with Suno.
In the playlist, you see the songs that are "clip 2" what I did there was extend your song from a time like 00:05 which is before the singer started singing. This way you can keep the music you like, but keep retrying until you get lucky, and the voice is correct. Example using that song: https://suno.com/song/04c5f7da-cdbd-4065-a4df-4b0ac858e1db (this is a "get whole clip" my style prompt was actually "German, Fast, Male, Male, Male, Electronic, Indie, German, Mellow, Soulful, Sad, Melancholid, Emotional")
I don't why this works better when prompted like this, but almost every voice was male (bottom of list: https://suno.com/playlist/bc8d78b2-6493-446d-8b39-bb0b65637708)
When you continue from a clip, I recommend first doing a "get whole clip" on the clip before extending it again especially if it is short.
Fast, Fast, Fast, German, Male, Pop, Electronic, Indie, Chill, Mellow, Soulful, Sad, Melancholic, Defiance. Frustration song. Listen and make your own with Suno.
In this case I think the I chose a time too early, the music is changed a little too much. Maybe 10 or 12 seconds would work better. (That link is 00:06 so the music is not as well established)
Though I like the song...
TLDR ON EDM GENDER PROMPTING FOR ANY FUTURE DISCORD SEARCHES.
For prompting male EDM voices:
- "German, Male, Male, Male" worked better than "Male, Male, Male, Male" but both worked better than "German, German, German, German". Just add a pinch of German?
- Prompting the voice from an existing song (with the wrong singer gender) using an Extend From time like 00:06 or 00:12 seemed to work more reliably than the same prompt as a fresh generation. So basically take a good song with the wrong gender voice, prompt it from before the voice starts singing for the change.
Just a quick test with one prompt so maybe just random chance results but interesting and a bit surprising, worth a deeper test.
Hello there. Just a small question.
Does Suno afford importing my music piece? Maybe a short audio, midi or gtp-tab to show the melody I want
I've seen this idea posted, but worth repeating. switching [chorus] for [Refrain] or other non-chorus tags will lower the chance of choirs/layered voices
I'm wondering if anyone can help me. Sometimes suno is naturally ending a song properly. I can appreciate it would do this if eg the lyrics are too long and it runs out of time, but for other songs, it does it even when extended or even if the lyrics are finished - it's still ending the song really abruptly for no reason. Any suggestions on how to stop this? It doesn't seem to happen with every song.
Remember you can use the "Extend From" feature to put in a specific time that isn't the end of the clip, and by doing this, chop off the part of the song that was cut off. You can instead pick a time at the end of a line of lyrics where it is easier to continue into a smooth ending. But a small change you can try is adding extra space at end of your lyrics:
[verse]
lyrics
...
...
Something like that. Or even add [instrumental] tag at very end.
Ah perfect thanks. Yes I was ending it before the end of the clip but it was extending it and then cutting it off again. But I will try the instrumental tag and see if that works. Thank you so much
\chirp
@untold briar I'm not there yet but I've made some progress on bass
everything related to this prompt is basicslly what I could only call anti-music
and it's awesome
I've had difficulty broadly getting that deep bass
but here it almost always sxores
scores
I'm in the process of analyzing what specifically in it is the source of the solid bass
bc trap does not always produce such deep bass
on suno
you know me, still the local "anti" skeptic. not that anti doesn't change things in useful ways but just that it doesn't stand out from the set of weird little style words that also change things (as opposed to the set which changes almost nothing, which certainly does exist). here i randomly swapped it with aaaa which is not a word I have tested to probably that is doing little to nothing, and this prompt is mostly coming from the other non-anti words
8 samples to add to your database https://suno.com/playlist/25c61cfa-0a6d-4763-a3ed-8e1edea15bb4
possible, but idk, I have definitely done sets of up to 8 gens on both ends, anti definitely, for example, is more likely toclead to vocal hallucination in instrumental
the "(" may also have a pretty decent effect, haven't really tried it much in v3
this particular one I think anti doesn't do much
it's mostly trap here
I've "simplified"
however, it doesn't have as much general weirdness
hang on, it needs accidental
also the bass is not as deep
it's weird, like for some reason that other combo just gets waaaay deeper
i haven't tried accidental but I bet that might be real good, it's both a description but also doubles as a literal music term
https://suno.com/song/76c01b36-16ea-4f25-86fb-4bcd3401aa08
sister gen. You can see these are just not as interesting (imo) as the weird combo of random junk
but actually nah, it's probably the goth
bc it's going to add repetition
hang on
is goth music actually specifically repetitive in real life, or is that just the way the style word works out? I just checked it is for sure repetitive in Suno at least.
Sorry for my question, but what that message mean and how can I resolve it after I put the command /chirp and lyrics? Diese Interaktion ist fehlgeschlagen.
Diese Nachricht kannst nur du sehen • Nachricht verwerfen
Sorry, it means the Discord chirp bot /chirp is not working. Currently only the website is working to make music. However the website is version 3.0 and /chirp bot was version 1.0 so you probably would prefer the website version anyway.
Thanks for your information! I'll try the webapp!
it's very repetitive
defining feature
Hello, some times when i do prompts with unknowns words to makes it used by a song, it generates 'Lost in translation' song, do you know if there will be something to not use control about the syntax used in prompt ?
it's good though, but I like everything basicslly except that kind of death metal where you can't hear anything
well, actually no
I like ramstein
Bella lugpsis dead
is like, peak goth
it's def interesting bc genres like that have a lot to gain from large context windows
since buildup and composition is really really extended
HI guys, can someone link relevant discussions or document about prompt usage?
like 1: How to write the best prompt, 2: How to make song similar to something
prompt people may like this
https://suno.com/song/d42d379e-ed9b-402f-ba1d-c2e26f9a6f71
it's managed to isolate pure repetitive bass
ultra-instrumentalized trapized trapccidental ultra-hyper-math-trapstyle neurofunk song. Listen and make your own with Suno.
Hello! HELP! I'm getting errors constantly when I try to create something new or when I want to extend some songs
What's happening?
I'm from Belgrade Serbia btw
me2 ; Ping me if you get a response
New musical style I've been working on
ultra math-antibass anti-hyper-gothllo trapelloized jewish nigun40404040404040, accidental song. Listen and make your own with Suno.
it's horrific
but I love it
Are you aware about this
https://www.suno.wiki/
Slow-paced Chinese folk song,Passionate emotions,String ensemble,Brass instruments, Percussion instruments,Female chorus song. Listen and make your own with Suno.
epic, Roman Military Anthem, cornu, hydraulis, lyre, Verses with storytelling, gritty vocals song. Listen and make your own with Suno.
Are there any plans to remove text limits for the lyrics boxes and the song description. also is it possible to extend past 2 min
I am aware of the wiki, but there's nothing relevant. Different generative models have different syntax, i am sure also Suno model has one. But how can we improve our job here?
Btw, how do you say "a song like Frank Ocean"? Is there a way to prompt this?
what's up with the thumbs down emoji
idk... 😒
I just use something like "Frank Ocean-style"
using the artist instead of the song works similarly
Better yet, research descriptions for Frank Ocean using musical terms.
If I give the bot or website lyrics that has (chorus) (Verse 2) etc will it understand that?
If so how do I make it sing the song but have a gap of music in between each sonits not just one continuous thing if singingness?
https://suno.com/song/d4f2ac4a-2f49-4414-915b-c56cf0f38518
this is an interesting technical progression
cronchy-cranch-gronky shrank-bonkums, flarfy doodlebop song. Listen and make your own with Suno.
since it carries the lyrical repitition into the instrumenal portion
I can't test if prompting "mama" in the style carried over
but I suspect
as it also sings "cronchy cranch" which is a style tag
I got a question. What would be the proper way to extend the songs? Do you guys do Prompt > Extend > Full song > Extend and repeat or do y'all keep extending each part and then merge the last one to make the full song instead?
*Also, extending seems to mess with the lyrics for me. Like, it duplicates whatever I add and it's missing from the first part. Will send an example in a moment.
Posted it on general but it was lost between all the messages. Thought it'd be better to ask here instead.
Here's an example
See how it duplicates because I specified the [Outro] and [End] parameters but it couldn't get to that part because the extension is only 30 seconds long
You know sometimes I look into my library and read my prompts thinking i must have been quite deranged while cobbling them together. And then I read yours and feel sane again
Your first question depends on if you want it to reference only the last part or as much as possible of the song. If you merge after each extend it will be more consistent, but keep in mind the model has limited memory so it can not reference a chorus that happened 3min ago for example.
If you want to change things up it's easier to do without merging after extend.
Your second question is caused by extending yes. Each extend throws all the lyrics you put in during the generation of said extend irrespective of how much of the lyrics where actually sung or from which point you extended.
Quick question, are you able to download a WAV or AIFF instead of mp3?
Hey
Does anyone know of a way to make ai parodies by changing up the lyrics a little bit? Keep the same vocal/voice tone/voice but just change up some words.
Maybe there's a way to make an ai sing based on the instrumental track?
You need to use brackets and parathesis.
@ashen surge
Otherwise it will hallucinate and include those words for lyics.
Also, I noticed when extending a song, the section " Style of Music " you can add more spizaz.
dramatic end for the outro and towards the end of the track.. the ai started to whisper..
I hope this helps.
I like it. Some limitations but you will get what you pay for in my opinion and this is only v3.
Is there a list of terms like [Instrumental Bridge] or something I could use to force it more towards a structure?
You force it by repeating the words that need more weight
repeating the words? Won't that just make it sing the word over and over?
I'm trying to make it create more space in a song
maybe add a vocal breakdown
Can I edit the start of a song that's already generated or use one for a basis? Like use the same voice?
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
What if I only want to make one copy of the song? It always makes two and I dont want to pay for two at a time?
And why is it charging 5 credits? Why not just sell 5 times less and use only 1 credit for each? kinda deceptive isnt it? Tryna make it look like youre getting more for your money?
If there is nothing that costs 1 credit or even less than 5 there is no point to using 5 credits at a time.
hello, my last 50 or so songs ive tried to create had an absurdly large amount of very robotic voices, I tried many different genres for hours now and changed up lyrics, generated lyrics and it's all of a sudden so much worse than yesterday. what happened?
please let me know how you did that, i am having trouble making robotic voices
hey all i have a question how does one extend a song past 1 extension as It seems like I can only extend a song once.
dont leave lyrics window empty, add more tags
I just made a 8 minute song
prematurely ending a song is a problem with this i notice. wasted a lot of credits with that happening
@ornate belfry i have one issue im seeing. if i add lyrics it seems like the lyrics box and song description have a limited number of characters is there a plan to fix this and remove the character limits
They need to refund more, its BS
?? what do you mean
Udio doesnt do this
you mean they need to refine it more
Premature song endings
ya agreed there
they should have done that before they started accepting money
I just subscribed to the pro version as I would love to use the tracks in some games i and my business partners are looking to develop
won't be long until theres better open source models anyway. cant blame them for trying to grab money while they can
the model isnt the issue
they need to tweak fields on the UI
that is my only gripe
and i need to figure out how to extend it more than 1 extension to the song
they need to post what tag words they used in training
is there a place to post such feature requests
what happened to the vocal quality? It seems to have gone down hill from a couple weeks ago. Cant get a single song to not have constant reverb and voice doubling.
the chirp command is suddenly not working for me in my channel on my private server. did i miss something?
i'm getting an "application did not respond" error
seems like the longer the song the less it pays attention to your prompts
you cant change the music type in the middle of the song either
ive only seen this happen in other ais when the prompt gets really long, I am assuming they keep adding to the prompt every time you extend
or its paying more attention to the last segment than your prompt
I hopped on here to see if anything caught my attention.
It's interesting to see Udio being mentioned. I suppose it was expected. I've taken a look, and I'm sure many of you have too. My first impression? It's in beta and it's free for everyone, so I'm curious about the future. What changes will be made, and what will the cost be?
Udio is fun to explore and seems to have some really innovative ideas. However, since it's new, you'd expect them to try to outperform Suno. They must have developed their model by using Suno and other AI music generation tools. This is just how competition in the industry works.
So, I do expect Suno to respond with some updates to stay competitive in the world of music generators. I certainly hope so!
They all need users, so this competition should ultimately benefit us, right?
Moreover, I don’t want a generic app that produces generic sounds and songs—the kind of parody music that's prevalent today. I want to use these AI generators as tools to foster creativity and produce unique sounds and lyrics that are truly our own, not just a random throw of the dice.
I’m looking forward to the ongoing battle in the development of AI music, which has only just begun.
in an extension i want to add a verse and a chorus. in the first part the ai made the chorus start quitly which is nice but i dont want to have in the second part. but its alsways starting quitely now.. Any ideas how i can prevent that and have a powerful chorus?
Are they removing Chirp from Discord all together?
Hey everyone 🙂
I was wondering if there's a command in Chirp that allows for a sentence to be spoken? Or maybe even sound like one of those trainstation announcements? I tried [announcer] but it only seems to work in Bark.
I tested Udio, it's only better than suno in terms of sound. Everything else is light years behind. Suno can do granular specification in the prompts, it builds synthesizers as specified, follows the chords, follows German female voice etc. Udio does something, hallucinates, can't cut to the second, changes style 3 times in a 3 minute song. The sound quality is good but everything else is clearly behind Suno V3, rather Suno V2 or worse
Thats not true. you have actual control over the music with Udio, beyond the first section you have almost no control with suno
also its free, and they arent deceiving anybody by pretending like 5 credits are 1 to make it look like youre getting more for your money
just wish it werent so slow
Ok, let me explain: in both cases I start instrumentally with chords. In Suno I cut after 10 seconds. With udio I have to take the 33 seconds. The fact that I can jump back and forth with udio is nice, but with Suno I cut sequentially and granulaterally. With the same prompts I get better music from Suno but I have to work with it. Don't just throw text in. I throw in 10 dollars and get the rights. What about the rights at udio? Competition is good for us but the buzz around udio right now is a lot of smoke for nothing to me. I'll keep paying the 10 dollars to Suno for now, I've generated music with Openai Jukebox on my own gpu infrastructure when it wasn't popular and I still think Suno is the best generative music model.
post-(the seminal work of doodoopoopoo.com, their full evil soul poured out into the blood of piano) gothline song. Listen and make your own with Suno.
I accidentally recreated moonlight sonata, but variated it
Can anyone answer to this question please?
That is Not possible. You can try Meta Open Source Models for instrumental
What model do you mean?
I mean just show Suno the melody of intro, to get it and use in the chorus...
There is No Way for that. You can try to describe it using chords and instrumental. If any attempt Sounds like what you had in mind you can continue from that. But you cannot Import afaik
do titles help with the song? or are they just.. normal titles
They are just titles
That's something. How can I describe chords? [A-Cm-F-G-E-Bm...] inside the textbox?
In the prompt Click instrumental and put them in the prompt where you describe the Style of your Music
divided by minus, comma or space?
Unfortunately there isn't a known reliable way to prompt for chord progressions. I see users try different ways of formatting chords and some of them do report better-than-random-chance success, but typically in pretty simple cases. Off the top of my head maybe just one single chord like [Cm] in the a verse in a lyrics prompt for an instrumental. I haven't check this lately though and my info may be out of date, could be some users found better method
I tested it in V2 one time using external software because I'm not musical enough to match the chords, and even when it worked, it seemed to ignore major/minor entirely. But V2 is quite different than V3.
Yes, that's the my experience in V3 too, Jonathan!
I see it gets the main chord (like C#m phrygian minor), but that's all I could make reliable
If you feel like doing experiments, this format for melodies/notes did sort of work a little in V2. Example:
Hap[C4_1]py[C4_1] birth[D4_2]day[C4_2] to[F4_2] you[E4_4]
Hap[C4_1]py[C4_1] birth[D4_2]day[C4_2] to[G4_2] you[F4_4]
Hap[C4_1]py[C4_1] birth[C5_2]day[A4_2] dear[F4_2] Su[E4_2]no[D4_2]
Hap[Bb4_1]py[Bb4_1] birth[A4_2]day[F4_2] to[G4_2] you[F4_6]
Even when the notes didn't work it often made very unique singing styles:
https://suno.com/song/e1d753e4-9357-477e-a745-deaa19e82151
Why do you use V2 when you have V3?
That's from January
So one thing I've noticed is that after continuing a song a few times (e.g for a slower song), in V3 it tends to forget the melody/verses and I can't seem to generate anything more. Have others experiences this / know how to get around it?
Try "Get Whole Song" on each clip before continuing it
👏
For reference, I tried doing get whole song on the furthest point that I had gotten, and it was busted (just generates random melodies after). I'll give it a shot in individual ones - thanks
Yes, Get whole song - is a clue. You can generate 1 more second - to introduce voice or new instrument you like, and then generate the melody of it this way.
I blew through a months worth of credits in a few days 🥲. This is very fun to use
Even with that 500 bonus hah
I did my first song for 8000 credits yeah 😅
But I needed exactly_that_kind_of_pause in that shitty place 🤪
The heroic epic of The Song of Roland https://suno.com/song/d73f99ac-2998-4cd2-8129-62641df39d66
Medieval,gritty vocal ,gospel, Gothic Soundscape, Dark, melodic,haunting violin, Dynamic Swell,Inspirational,epic battle song. Listen and make your own with Suno.
🥹 no violin 😭😭😭
🥲
hi, any idea why i get this message? "The application did not respond"
/chrip
Chirp is retired. Please use https://suno.com for the WebApp
Thanks
/chrip
post-anti-instrumentalized anti-hyper-gothbit song. Listen and make your own with Suno.
/chrip
neo-(ultra math-antibass) anti-hyper-gothline anti-trap song. Listen and make your own with Suno.
I'm happy to see someone with more knowledge than me has had a look! I so easy to get wound up in this... Do I need that!
Many thanks for your time and reassurance!
POV... I noticed this abbreviation in the Wiki Notes: Torchy-lounge song, from the POV of a piano player.
Abbreviations save space and time, so I was wondering if Suno likes any other abbreviations. Then I got to thinking: What others would I use!?
My gut instinct is for anything with "more than one word" if you can find a single word that means a similar thing it has a higher chance of working. or try "using-dash" to combine two words. or abbreviation, same idea. Even maybe using a foreign language word that happens to say in one word what takes multiple words to prompt in English, reportedly some german words were good for this
neo-(ultra math-antibass) anti-trap-gothline neo-neurofunk song. Listen and make your own with Suno.
end switch for that eerie fade
using that weird but extremely effective prompt
alternately (but incomplete)
https://suno.com/song/330ee5e2-340f-4514-8c73-f6d7c181b7a4
neo-(ultra math-antibass) anti-hyper-gothline anti-trap song. Listen and make your own with Suno.
Style of Music. In Custom Mode, I created prompts to go into Style of Music. For example, Welsh choir front, gentle koto accents, and tranquil backdrop. Male, I realised that Suno seems to have a preference for what is placed at the beginning of the prompt.
With the Welsh Choir at the beginning, a choir, 'well an attempt at a choir,' dominated. Then I tried rewriting the prompt with gentle koto accents at the beginning, which were prominent in the song.
So I thought I'd share this intriguing discovery. It's a small tweak, but it could potentially be a game-changer in your music creation process. Perhaps, it might be of use to someone, empowering them to experiment and create unique musical styles.
I'd like to know if anyone else came across this or other strange finds.
i was going make a joke about prompting the windows start sound and you already literally prompted it jasldkfjas;d fjlwa
(or something with enough resemblence anyway)
You're not telling me I've got to learn German now 🤣 Only joking! I get where you're coming from.
You do have to think carefully about your prompts. With only 200 characters, including spaces in Simple Mode and Song Description and 150 characters, including spaces in Custom, Style of Music, it's a balancing act between speaking Suno's language and being creative!
Prompt ordering is not supposed to make a difference, however I am not certain at all order doesn't matter. But it not simple to udnerstand: changing the order will make things different so it has hard to figure out if things being front or back matters, or if changing anything just changes your output randomly, and since you probably decide to change up a prompt when it's not working, if you 'regress to the mean' and get a more average less-bad prompt, it feels like things are better. another thing to check is maybe adding a trailing comma like "blah, blah, blah," might have some effect, but I did not test that, just something i wrote down to maybe test some day
Does Suno support Gregorian Chant?
my mental model is that a space is like 1/3 of a comma. it's like a weaker grouping effect but still there. for example if you remove all commas, suno does not explode because you wrote one long sentence that makes no sense interpreted as one single phrase or genre. just a hunch this is wildly speculative
These are the two audio tracks changing the placement of Choir and Koto. Can you tell which on starts with Choir and which with Koto?
to wildly speculate a bit more. maybe tags found in songs that have 2 words, compared to tags found songs that have 10 words. imagine the training data is randomized. does a randomize 2 word song look like 'front, back' or does it look like 'first word, second word' ? If you compare with songs tagged as "first, second, third, fourth, ..., tenth". the tldr would be maybe some specific set of words really benefits from being in front, but not others.
i have no idea how anything actually works, kind of silly to speculate at mechanisms I think, instead of just trying prompts 😄
are these style prompts the same length or is prompt 2 shorter?
I started out with the intention of trying to find certain things that worked, and you could rely on.
My idea was that once I had a consistent working framework of tags, brackets, etc., and I'd learned the basics, I could then start being creative. Just like you do when learning any new instrument! Consistency is something I think Suno need to build in and I'm sure it will come.
The AI music generation of yesterday is a million miles away from what it is today, and tomorrow, it will be so different again!
What a great time to be learning this new instrument!
Oh I got to step out, didn't notice the time. agree fun time to experiment
I used the same prompt; I just moved the words around.
I would have to test a loooot more to be certain, but it appears to matter. That being said, that may simply be due to the hyphens producing tokens which end in either a space or a "-" which in turns primitively creates an "ordering"
ie I have tested gothline-hyper-anti vs anti-hyper-gothline
and there's a dif
however, I didn't spend 200 tokens on it
so like, it's tough to know until you do that
those should be considered single words, so it absolutely matters there. but specifcally where i think it isn't supposed to matter is comma seperated list ordering
Anyone know if there is a trick to edit lyrics so it alters the phrasing when it's vocalised?
The tags work well like: (verse) (chorus) etc.
Do anyone knows, how i become lot of views?
Where is the best place to share this link ?
Facebook, tiktok , instagram etc?
uptempo electronic pop song. Listen and make your own with Suno.
Building up your own followers. There is no magic way tbh. 🌼
Ok thx for the fast answer 👍
Hello, is there a prompting guide for v3? Has "quality prompting" been tested? Anecdotally adding "clear" and "crystal clear" seems to help somewhat to alleviate the "distorted voice" problem.
gg
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
#📚┃suno-school In a whispering wind, the echoes of time,
Whisper tales of the days left behind.
In the depth of the night, where shadows dance,
Weaving dreams of a forgotten romance.
The years they pass, like grains of sand,
Slipping through fingers, slipping from hand.
Memories fade, like colors in the light,
Leaving traces of a silent fight.
But in the heart, a flame still burns,
A yearning for the days to return.
For every moment lost in the stream,
A melody lingers, a silent scream.
So let the music flow, like a river wide,
Carrying us on its gentle tide.
For in these notes, we find our release,
A solace for the soul, a moment of peace.
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
Got this welcome message https://discordapp.com/channels/1069381916492562582/1096209586245083248/1229185946809794670 but none of the links work. How do I get access to create /chirps
chirp needs new jazz!
[Instrumental Intro].
I can't create an intro even if I fill out the form like this.
Does anyone know what I should do to increase the probability of making an intro?
@celest summit , your recent activity of posting the same YouTube video in multiple (8) channels has led to a 1-hour timeout. This action is considered spam under our community guidelines. Please use this time to review our rules about posting content. We value your contributions and look forward to seeing you participate constructively in the community once the timeout is over.
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
Chirp is retired. Please use https://suno.com/ for the WebApp
nope you got to use the site now.
Read #📢┃announcements
帮我写一首歌表达不想上班的心情。要求中文的,旋律欢快的
catchy, cheerful, nihilistic, pop song. Listen and make your own with Suno.
is this...sino-ska ?
ska is...?
Ska (; Jamaican: [skjæ]) is a music genre that originated in Jamaica in the late 1950s and was the precursor to rocksteady and reggae. It combined elements of Caribbean mento and calypso with American jazz and rhythm and blues. Ska is characterized by a walking bass line accented with rhythms on the off beat. It was developed in Jamaica in the 1...
I used catchy, cheerful, nihilistic, pop
Ok figured I would ask in here. Is there a way to get suno to use particular chords?
how do i make music here guys?
Idk
Any way to make the AI sing slower? Sometimes it feels like it's on a hurry trying to finish each verse
No, in short. I have tried may times, many ways, and struck it lucky just a couple of times.
Hi,
Understanding the functionality of Suno is crucial as it empowers you to make the most of the platform.
In essence, Suno creates lyrics and accompaniments from prompts of 200, including spaces or less in Simple Mode and only 150 in Custom Mode.
There are so many variables involved that the odds of achieving what you hope for can feel similar to rolling dice and hoping for the best, which can be expensive.
To use the dice-rolling analogy, if you aim for a six, the chance of achieving it with:
one die is about 17%.
Two dice, the chance decreases to just under 14%.
Three dice, it drops to less than 5%,
Six dice (or six variables), the chance of rolling a six is only 0.00214%.
Thus, if you consider the variables in Suno like dice, you can see why it's hard to get what you are expecting!
We need to know how to control the odds, which means we need to learn how to use Suno efficiently, consistently, and with assurance of the outcome.
Understanding what tags are and how to use them. How do you construct prompts in Simple Mode and Custom Mode, and know the difference? How to build lyrics using set parameters, such as brackets and tags. We need to know how the lyrics are used in the creation process so we are not left wondering, 'what!?' We need to understand why something has not done what you expected! Without such knowledge, the odds of us being creative and achieving our goals are slim.
Imagine putting someone who can't drive in a car with six pedals and sending them on the motorway. I suppose Suno isn't so dangerous, but I wanted to make a point.
If I were a cynical person, I might think the company has a vested interest in not showing us how to play effectively!
Getting back to your question, the best documents I have found are here:
https://www.suno.wiki/
These are good, but I think they open up more questions than giving answers.
I wish there was a dedicated teaching/learning forum, monthly tutorials and live interactive events. I think this is the only way to move forward.
Enough of me going on.... time for a coffee! 👍
/chirp
Go to the #🐣┃chirp-beta-1 channel to use the /chirp command, and check out the #🐣┃how-to-chirp channel for step-by-step instructions for making music using Chirp on Discord.
Neo-thrash, Groove metal, post-thrash song. Listen and make your own with Suno.
I'm trying to get a consistent male singer. I've attempted Male, Voice Male, and Male Voice in the prompt, both at the beginning and the end. However, I still do not always get a male voice.
Any suggestions, please? I
Is it not possible to define gender?
after your first section theres almost zero control over what happens
good for them i guess since it takes half your credits to make one song
I just got a segment with 40 seconds of silence and no refund
maybe too early to be accepting money or at least give more refunds
Can someone please explain that when I Click Create for the first time, the Clips produced may be of varying lengths, but when I Extend the Clip, the Clips are always 58 seconds long?
Sorry, wrong channel. Carry on lol
I always use "Male voice" and it works for me
Is it just me or is it really difficult to get a chorus sound exactly the same for 3 or 4 times in a song?
wie kann ich das jetzt nutzen?
/chirp
Why do i have no access to the generating channel?
Wondering the same thing
Ok so not just me
in Basic Plan, does "50 credits renew daily (10 songs)" mean the user will not earn extra 50 credits if he has already had more than 50 credits?
Because you have to use the website now. see #📢┃announcements message
/chip
In Style of Music, do you add 'Male voice' to the beginning or the end of your prompt?
Hi, I think you can now only generate songs online at Suno.ai.
@ivory bramble @copper temple @mellow pond @abstract silo @quartz canopy @elder oasis @grizzled cradle @raw talon
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
Shouldnt be charging for <1 min segments if you didn't use [Outro] tag. Pisses me off.
I am not a developer. I am a volunteer user/moderator. For feedback use #💡┃suggestions-web
Short simple styles, blank styles, long diverse styles, randomly typed letters in style 🤷 every day I think a different method is better for clear voices
i mostly do "female-fronted" or "Male-Fronted" and that seems to work
Sometimes I do (no joke) "German, male, male, ..." for male voices in EDM
Well that's different and I'll certainly give it a go! Many thanks!
😯
LEt me know if it works for you like it does for me 🙂
Hey there! I hope my request is in the right spot here because i'm pretty new into discord and using AI in general. Is there any beginners guide to prompting in suno?
I have no idea how to use discord 😦 What chirp and bark mean?
A good way to start with suno is via our community website https://suno.wiki
You can go to your library. Note there is pagination, you can change your current page listing at the top right
thank you! ❤️
my bad, i had Liked button on
It happens, didn't even think about that scenario
The wiki is good, but its also a lot of fun to just try things and see what it makes
Think what you want your prompt to convey, and when the AI mucks it up, just tweak it till it does what you want 🙂
Absolutely agree. Not only can you find new and interesting sounds to add to your personal styles, it can make you have your own sound. It can make your style to test things out and do things on your own.
Just be nice and share some of what you discover along the way with the rest of us too!
Is chirp website only or can I run it like bark?
i will certainly share some if i make some good stuff i am now also looking in some odd fusions, like Jazz and metal
Exactly this! Fusions make for new sounds and styles. They can even help get closer to a style or sound that you're looking for by blending them the way a painter will use red and yellow to make orange. But it's the whole gesalt that makes the style or song work. And even common things will work as styles sometimes. I once used hamburgers and bratwurst as a style...
🙂 bradwurst ist sehr gut
i like it from time to time 🙂
I just looking for a fushion between the strength that jazz exhumes, and the anger behind protest metal songs. so if i can combine those two like they are both working together but also music batteling, that would be great (and when i get the prompts etcx just right imma even sub so i can use the thing in games i am making with rpg maker)
Hi everyone, is Suno having a bug that it can't extend the song? I have been trying to extend the song for 2 days, but after each time, it's refunded and the extend version has no sounds. Not sure if this is just me or everyone else is having the same issue?
is it something you would share? i can try extending it from your link
Thank you. Here's the link. https://suno.com/song/eec926a6-eb18-491f-8da6-92eec70eb1d1
piano ballad melancholic dreamy song. Listen and make your own with Suno.
any specific timestamp, or just the end?
Is this happening to a particular song? If so.. make sure you are not continuing from a part that is fading out or has silence because it will result in failing clips. To avoid that, you can move your timestamp a few seconds earlier to continue from. This often helps.
4:50 total.
Cool. let me try it out
Thank you so much for all the helps
I think it happened when I enable (instrument) option and closed the lyrics part, it error out. I retried with leaving the lyric part empty and generate. seem to be working now. but got about 58 seconds and 24 second part 2
i usually get about :58-1 on extended generations unless i try to prompt for shorter, for like outros or breaks
Cool. I am gonna check youtube for tutorials, thank you so much for your help! I really appreciate it!
is there a way to download my songs in uncompressed audio quality e.g. wav ?
Why am i getting 'Message Failed to Load '
hello, I was wondering, is it possible to create a duet on Suno. For example, I would like to extend the song but instead of the woman's voice singing I want to make it a rap by a man. Managed to get the rap part but not the male voice, is there a tip for it to write in the prompt or are duets not possible for now ?
ooh, that is a good question, I would also like to know
i tried putting in the style description, also in the prompt like this [Chorus Male] but didn't work for now
uhh where do I make music now?
its possible, but not easily done currently:
https://www.suno.wiki/faq/style-and-lyrics/duets/
is there any info about the architecture of the song rendering pipeline?
or is that closely guarded? I'm struggling to imagine how this is made but i'm guessing it starts with midi, then individual instruments? idk
how does it achieve consistency of instrumentation for a whole song then? thats wild
whats wild is the flagship models from google and meta over the last few years have only produced noisy garbage
that almost resembles music but is clearly computer generated mush
had me thinking that genai music was not going to happen for a long time, then this appears out of nowhere
Hi hi. Anyone know why “continue clip” option isn’t showing up? Trying to be to make a complete song for the first time
Thank you!! 🙏
I can't write all of my lyrics, how to add more in the custom lyric ?
If I use the pro plan can I ?
okay thx
thanks
did the singing get better recently or is it genre specific? Prompting a "Catholic prayer" gets great results but little backing music /shrug
how about Catholic Gospel?
Thanks for the answer! I'm just struggling with finding the wiki 😆
Yes! Found it by just googling suno.com wiki 😂😂😂
Great, i just like the faff about with it, sure i can make only 5 songs a day, but that limitation does help me focus on doing things more streamlined
hola alguien me ayuda
I have a question
Is there a way to make a long build up with vocals?
Because it once happened to me when it generated the result
I will be very thankful for a reply
i heard you can use / to extend the hold on a sound, like Hi///// is supposed to enlongate the i sound
Also Hiiiiiiiii can work
hello... i'm trying to extend a song... but it keeps failing and refunding my coins.. what could be going wrong?
Try restarting from a different timestamp. It keeps getting refunded because the new clip is under 10s because the AI thinks the song should end.
oh is that a thing? ok... I'll give that a go, thanks 🙂
so, I've found a spot then, in the song that could be extended... can i now add text to the lyric box and they will be inserted in the 'extend' .. and then the rest of the song remain the same?
How can the ai just cut in the middle of an incredible sung verse? how it usually can't it reproduce the exact same melody, even if the lyrics are the same? Or it finishes a song brilliantly doing an outro, but then adds some weird instrumentation like it's a totally different song?
So it needs a cut at the end option
Can i do some song with more than 2 min? i need like 5 more secs! 🥲
how can i do it?
also I am experiencing some bugs apart from weird nonsensical generations, that don't stick to the lyrics, or laguage, or melody...
i just need time for a 3 more words hahahah
like I'm in my library and suddenly I can't get to any other page
you see, I have 40+ pages of mostly useless generations.
you see I can't get to page 3, it's grayed out.
it's easy
just click on extend, and then the extension you create has an extra option get the whole song and the ai it will merge them for free
and i need to put 2:05 min for example?
you need to put the time you wish the ai will continue the 2nd part from
by default it gives you the ending time of the song you are extending
ohh, i'll try, thx! 🙂
there it is your option to basically cut out the last part if you don't like it
or half of the song or more,... well you get the point 🙂
I assume there's no way to change art for a track short of doing it yourself after downloading, yeah?
I actually like the cover for once, lmao.
But some of the finish art thumbnails suck balls.
yes the cover ... changes depending of the cover of your song that's on the top of the playlist
I tried this. I couldn't ever get it to continue the same verse, but I could do the bridge/chorus so ended up just ending it. Half way success. But I appreciate the help!
I'm struggling with it like crazy too...
I at least found a good way to end songs. If I put [End] along with [Fade Out Quickly] at least songs don't go forever.
the ai makes a very catchy verse and there is no way it can repeat the same way, except from the previous parts that are in the memory and are sung at random
For reference, the issue I was having was when you generate a ton of continues. In the normal case I have no issue continuing the song when it's only the 2nd or 3rd time
the best way to end a song is typing [Outro]
It just forgets after a few continues 🥲
but an [Intro] is rarely effective as a command on top
chiptines?
Chiptunes*
what are chiptunes?
oh don't worry I'm probably older than you 😄
or maybe not
I was a kid too when NES came out
I just really like the style and not enough artists make it in the style I like, so I figured I'd do it myself. 😄
Yeah of course 😄
i lost my song, F xD
i dont have any chance for upload my .mp3 right? (cause i have it download)
/chirp
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
I noticed, usually the first 6 generations are the best, even if you do 50 more there is a high chance they will not top the first 6
sometimes it just comes with the first generation that gets everything perfect, verse, chorus, bridge, pre chorus, everything. But then it's really hard to make it continue on the right track.
maybe is there a secret prompt? the ai does not understand commends like keep the song the same
yeah I got my first song great and it strays away from it with the same genre text.
or continue from previous part
That's the biggest nightmare, I am struggling with this for days
what I found out is that if it has lyrics in the first part there is a chance it will repeat them at random, but sung the right way.
if you type the exact same verse there is a very little chance it will sing it as it did previously
works with short choruses of 2 sentences best
I'm thinking of doing a "Extend song from 1 second and put in new lyrics to see if I can keep the song style.
I did a song with a chachy part, that has become the chorus
it is structured like
blablabla - blabla blala bla
blablabla - blabla blala bla
but the last 2 words are different
anyway it repeats them every time the right way with the same melody, so this is surprising.
the ai will often skip letters, sometime whole words and whole verses... like it's very random, especially the first line of lyrics, when doing an extension you write in is often not even considered by the ai
but still I'm falling in love with the songs I create. Even if I'd say, almost always there are glitches in a whole song.
but often I am just blown away
so it's worth it
I would also LOVE if I could reuse the same voice
save those voice parameters as a preset
also another thing I forgot to mention before when I posted the screenshot of the disabled pages, this happens as soon as I delete a song in the library view.
also beside showing track length in the library view, i would love to see date of generation.
also it seems as son as I extend the track looses some audio quality depth, there is a noticeable difference if I play the extended and non extended piece. I think the more extensions I make the more auto-tuned the vocal sounds. not to mention how the vocals can completely change, like it's a whole different person singing the new verse in the extension.
/chrip
What is "prompt flagged for moderation"?
https://suno.com/song/2d179693-0e09-49ba-968d-9d84151bd87e Hi friends - please any help would be greatly appreciated- This shit is pissing me off - I have worked so hard to get it to this point and I want it to repeat the line “I need you to know” again and then do “sure as marrow in the bone” sung exactly like “I need you to know” any ideas on how to make this amazing but sometimes stupid tool achieve this?
Did you try extending it?
[Outro]
I need you to know
Sure as marrow in the bone
[End]
And add "finale" or "end" to the style tags.
Suno tends to lock on to the chorus from the first generation and has issues if you try to add something with Chorus tags but different content.
/chirp infop
one time I had Suno whistle the melody of the song, but it's never done it again. I've tried putting [whistling] in brackets and in the song description, but it won't do it. Does anyone know how to make Suno whistle the melody, please?
I really want to just remove the sure as marrow in the bone part all together but suno is not getting the picture even after I tried removing it and regenerated I don’t even know how many times it keeps putting it in there
How to avoid suno to end my song before I want? When I try to extend a melody the generation usually tends to fade it to end, so I try to extend it again from seconds before it seems to begin that fade but the new generation also tends to only go for around 10 seconds to end then, leaving most of my songs for less than 3 minutes.
Do users earn credits by reporting bugs in songs?
nice
So is there a way to access the general licence we are given?
Chirp
Hello! I'm new to the channel. Can anyone please help me? Under server guide I don't see the how to make music with chrip messages, it shows nothing and I can't get started. I've paid for a month to test the app
See here: #📚┃suno-school message
Is it possible to prompt the vocal?
Seems like the audio quality has gone down lately. levels are completely out of whack and lots of noise
Half the time cant even hear the lyrics over the oversaturated music
Chirp's gone, please make use of the site for generating stuff
/chirp
Just to let you know it didn't work, but I wasn't using EDM. This got me to thinking, does Suno have song structures for each catergory, such as EDM. This would include how the song is created, sung and what options, if any, work. Hope that makes sense! Have a great day!
electronic dance music, thats the acronym
Ah Ty i am bad at acronyms 🙂
it's a pretty common word used here in the past year
Sorry about that, I'm getting lost on Discord again... Someone said try German Male Male.... I know sounds mad!
Dont worry about it 🙂
I've been trying to get an introduction to stop the clip from playing straight into the verse. I've been attempting [Intro] Intro: {Musical Introduction] and lots of variables of the same.
I've been reading the Suno Notes, which, although ok, do raise more questions, and I don't think they have been updated in a while.
One interesting snippet... in Bold... it states The tags are not necessary, the AI will build a song pattern from the lyrics with or without them.
So quesstion 1: Is this the reason that the tag prompts I used didn't work?
Question 2: Does Suno create the lyrics first and then build the song around them?
Question 3: So how do I get an intro, is it down to using a Style of Music that allows Introductions?
the tags kinda work fored for me, buyt not sure if it was the tags or if it was the ai doing its thing
what prompt should I write to finish the song when extend?
My feeling is that Suno Style of Music parameters impact on what tags you can and can't use. I would love for Suno Staff to help. For now, it seems that if a tag doesn't work, it's because the Style of Music doesn't allow the tag to work. This is just my idea! Any other thoughts are always welcome!
i used (fade out and end)
makes sense i tried a breathy song solo, and it didnt do that , what it did worked however
In which field should I write this? I just have instrumental music
Don't words inside parentheses just get sung?
not in my songs
sometimes it did in other things i tried but not in the songs i like
What is different with your songs? I'm not trying to correct you; I'm just curious to know.
i use this style [Big Finish]
ioops
Single Female-fronted aggressive heavy metal, Experimental funk
and some how it doesnt sing in parenthesese
I think it works, thanks)
Got 2 extended ones at 58 seconds and at 18, both end the song)
Amazing! So it does seem that maybe each Style of Music has an influence not only how it creates the songs, but on what, and how, tags can or can't be used.
Glad to hear 🙂
/ create
/chirp
Hi everyone. I stumbled across Suno (chrip?) today and used it to create a song from lyrics I wrote. It didn't a pretty good job, actually. While it's the easiest of the 10 platforms I've tried, it's limitations are, um, irksome, ha ha.
So, my song, with 2 verses, a bridge and an outtro, ends up longer than 1:58. From what I've heard, one needs to use the "extend" feature to "tack on" that which doesn't fit into 2 mins. I'm picking my best bet, therefore, is to clip my lyrics to only include the 2 verses and attendant chorus, then use the extend feature to tack on the bridge and the outtro. Do I have that right?
I was originally hoping to try to generate a Steve Winwood "Valerie" kind of vibe for my song, but had to settle for an early 80s pop love song prompt - but, to be fair, what I've got so far is pretty good. I'm creating the song for a friend of mine going through a rough patch, so I want it to be good before I present it to her. 😁
/create
i recall that Suno doesnt allow songs to sound a like, so you can look up the style and make your own lyrics.... moight have to confirm this with a mod
Hello everybody, I'm new here. I have a full song that I developed in various parts. I'm quite pleased with it, only the first part of the first part is rather bad. I'd like to cut it away and create a new intro that leads directly to the part(s) that I like. Is there any command to delete a part of a song and replace it with a new one?
Sadly that is not possible yet. The team is working on V4 and hope to offer us then "paint-in". And then you can do the things you describe.
Hey folks 🙂
Been messing around with making a few D&B songs for a while now - one thing I've found is on a couple tracks there's this very weird ambient chorus/choir noise that is generated on a few of them. Any tips to avoid this in my creations?
In the beginning of this slice it's quite noticeable
https://suno.com/song/bb5c5c42-ebc7-4115-846c-c76d607be722
melodic drum & bass outro song. Listen and make your own with Suno.
Hi There, i am new here. Would like to check 1) Can we choose the video/music photo 2) Can remove the SUNO logo on the downloaded video for subscriber ?
i avoid writing [melodic] for that reason
@versed valley @versed valley
Chirp has been retired. You must use the webapp from now on.
@prisma cosmos @kind turtle Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
嗨
Chirp已经停用,在Discord上不能再使用。请使用 https://suno.com/ 来使用WebApp制作音乐。
好谢谢
Hi team. Having trouble. All my chirp interactions fail. Can someone help me?
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
Thank you! Loved the chirp and love to see progress! Proud of this technology! Keep it up!!
/chlrp
Chirp已经停用,在Discord上不能再使用。请使用 https://suno.com/ 来使用WebApp制作音乐。
If am I paying for this why do I have to be censored? Not like I am posting my songs on your site in public. Don't prevent people from doing it just prevent it from being public. How is that not obvious to anyone? Are you running a business or not?
99% of the time its nothing censorable anyway. You guys really need to fix that
Sorry but its really aggravating
I'll bail as soon as something even remotely as good shows up open source, I guess
I hear you—it's frustrating to feel restricted. Even though it can be frustrating to face limitations when you’re a paying customer, Suno’s moderation rules are in place to ensure content suitability across diverse user bases and to comply with regulatory standards. However, you have the option to directly input your lyrics into a chat with ChatGPT and ask why they were flagged. This tool can provide specific pointers or problematic phrases, which might help you tweak your lyrics just enough to pass moderation without diluting your artistic intent.
What regulatory standards? Can you produce reference to it?
I cannot. I am just a volunteer moderator. But I know chatGPT is the gatekeeper. So what ChatGPT accepts, Suno accepts.
Oh no wonder.
CHat GPT is a cripplebot. They are so afraid to be sued it can't really do anything useful.
But why are you using it to moderate our own lyrics?
You mean: Suno. I am a volunteer.
That makes no sense. If we write the lyrics there is no point in passing to chatgpt. In fact I dont want my lyrics to be seen by chatgpt
Youre volunteering for Suno, so yes I mean you
It does make sense. Suno could also be used for wrong purposes. Making songs for terrorists etc? And a human checking all, is just not possible.
Well.. this discussion is futile with me. Again. I have a volunteer role and provide answers. Not debates
You don't have to debate but I think its funny you think it makes sense. Hope I can get answers from the team on this.
You can write them via support@suno.com 🌼
I did, over a week ago and no reply.
That is outside my scope.. sorry cannot help you with that 🤷♀️
Well thanks for trying
I think its a terrible decision to have chat gpt moderate content. It is extremely limiting your product.
Whats weird is I made like 5 segments with the exact same lyrics and it went fine but after that it flagged
It would ne nice to know exactly what was flagged.
whats the prompt
they dont use chatgpt for moderation but something else and when you ask why there is any kind of moderation then you must be new to the internet
@rigid iris , please ensure your statements about moderation tools are accurate to avoid confusion. This type of misinformation can lead to unnecessary misunderstandings. You've been timed out for 10 minutes to review our community guidelines.
there is no modmail to dm? but suno website with custom lyrics for sure does not use openai api or they would be violating the their usage policy. the system also does not seem to work or be enabled for all foreign languages vs simplest openai model that does detect with same lyrics
suno has only report feature for the web site published songs? how songs elsewhere are reported properly
This is about moderation. ChatGPT checks your own lyrics for not being too graphic.
Devs confirmed to me that they use ChatGPT to check the lyrics.
Any question or concerns can be directed to support@suno.com
for the auto generated ones yes but for custom ones in non major language they dont seem to manage even the ones that the lowest chatgpt model, 3.5 does via api
ok thanks
Go into custom mode
Can someone post a link to anywhere with technical details on metadata to add to Suno to guide it. For example I tried [doubletime] for a rap but that didnt work. I copied [VERSE] and [CHORUS] from Suno auto making lyrics but would like to know any way to guide the tone. not only double time, but maybe ascending or descding too, that kind of thing.
here are some basics:
https://www.suno.wiki/faq/metatags/
I've have decent success using "building intensity" "crescendo to allegro" in style tags and [BREAKDOWN BUILD] as inline prompt.
to get the bot to sing faster, you might need to double up on the words-per-line...
Lyrics play a HUGE role in how the music pattern is made
What do styles tags like this do/mean:
(electric guitars, bass, drums, hip hop beat :1.3)
(male vocalist and pianist :1.4)
how would I go about making an instrumental that has a chorus but also empty parts for the verses
it's probaly because it's stored in the first or previous parts, where both phrases are together, so the AI just repeats them at random. And it will rarely sing the same words with the same melody again if you just add the first part of the chorus as lyrics.
hi, i have a quick question, is there a way to remove/reduce the chorus filter that gets placed on vocals generated?
oh, this would be nice too many voices have these effects that get even worse as you are extending the song.
I have found a video on youtube, have yet to watch it but it's long and as I quickly saw that guy shows some really specific commands
channel: Chris Unlock AI
video title: How To Create A Hit Song With Suno ai V3, Plus Tips And Tricks
thx man
chorus is a thing that is in the model. It get triggered with certain genres. It's hard to get rid off atm. I use things as a cappella, minimalistic, unplugged, solo, soloist, and it helps a bit.
maybe you will find something useful 🙂
how do you use those daizy? im kinda new to suno
I write [Chorus] and instead of Suno repeating the chorus, I get it to sing the Chorus word
^
In the "style of music". See my video @silent yacht
https://youtu.be/lj8utBd_IUM
Hello, I'm 3Daizy, a moderator and guide at Suno. Welcome to my channel, where I share my journey and experiences with AI-powered music creation. Whether you're a budding musician curious about AI or an experienced artist looking to innovate, you'll find useful insights and practical advice right here.
Join me as we navigate the fascinating int...
thanks
on the suno website, lower left corner you have the Help section
is very basic but good
it's the Suno wiki
ops my mistake lower left... corner
Suno wiki is made by us moderators and guides... it a community wiki.
You can find it here: https://suno.wiki
hello guys
Hey
/bark
I keep trying them with different genres and styles, but can't seem to figure out what the numeric values contribute:
(electric guitars, bass, drums, hip hop beat :1.3)
(gypsy guitar, double bass, violin :1.6)
Any ideas?
lol I am quite desperate
see my last promp
repeat verse and chorus
make longer heavier outro
instruments play all the time
same melody from part 1
faster
😅
deos it work? wel... kinda
even tho there is no lyrics written, of those being sung, they are all from previous parts
and they are not sung as they were before, so this is ... atrocious
it's funny cause my song sings "We'll create a new reality, no longer off-track"
and sing it totally off-track
... is the AI trolling me to make me lose crredits? Cause last 8 generations gave me exactly what I did not ask for.
and it happened before, like I have 40+ pages of trolling generations 😉
ok, finally something that I can merge!
Is there a reason suno refuses to make a violin song?
i see people using weights in style tags, does that work?
did exacly what I want for the final extension... marvelous!
I never tried, what happens if you write [Violin] in the lyrics? or Violin in the prompt?
[Violin Solo] maybe
so all the iterations I've done either change violin to acoustic or classical and those for some reason only translate to piano
I used acoustic arabic melody once, and immediatelly got something beautiful, but the melody was spanish, but it was acoustic
maybe try chordophones
I just found this word on google 🙂
"Hi, how should I make sure that the song I'm creating has music at the beginning and end without the song starting and ending abruptly?"
Guys, strange thing happened, 'lazy lofi saxophone hip hop beat' didn't create any tracks with sax on it, I tried 3 times w dif variations (instrumental only)
add extra word saxophone and start with it to make the AI understand you REALLY want that saxophone.
Chirp bot down with anyone ?
Ahh ok when i tried to make something it says the application did not respond, maybe they have deleted it
I wish they could make the web site work like the discord bot. The web site sucks.
are there any reliable meta tags to force modulations/key changes? like, ideally not to specify the exact key you want, but rather just tell it to switch things up
No, the AI has a hard time to understand that.
How do I stitch/merge the extended part of the song?
3 dots menu, choose "get whole song"
strange, there is no 3 dots menu, choose "get whole song"
you have to do it with the last part of course 😉
investigating further... thanks
Hello, I'm 3Daizy, a moderator and guide at Suno. Welcome to my channel, where I share my journey and experiences with AI-powered music creation. Whether you're a budding musician curious about AI or an experienced artist looking to innovate, you'll find useful insights and practical advice right here.
Join me as we navigate the fascinating int...
can y’all point me in the right direction to learn better on how to tweak and control genres
is that just the wiki or is there more or examples I can see directly with the wiki in action. Thank you
Anyone else have this problem?
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
Thank you!
oh man... i struggel and generated now... more then 20 versions and i don't get a good result... There are cool ones that are but totally noised 😭 Can someone suggest a way to get a good result in quality after a genre switch?
https://suno.com/song/1244d614-d2f8-458b-b570-957d5aa0718e
(after 40 seconds i want to generate it further with the style it gets)
aggressive experimental polymorphic song. Listen and make your own with Suno.
hai/chirp
#📚┃suno-school song
I don't know@pulsar cradle
cn
@shadow hull 我2
看你名字就知道了
how did you make it do "full song" out of curiosity?
suno chirp bot "该应用程序未响应 " why?
同问
没有/chirp是什么原因啊
When can i use the API?
A quick question: Can I add lyrics to an existing generated instrumental-only song, after it is done?
no
Hey Guys- Loving Suno. I am experiencing distortion. Is there anyway around this? I am mastering in Ableton Live which merely emphasis the distortion. Wonder if you will be introducing stems at some stage?
yeah bark for me
d
今天咋回事未响应
Hello everyone, how to make songs that are longer than 2 minutes? In the library, people have a lot of generated songs of 3 and even 6 minutes, but I can’t get more than 2 minutes, what should I do?
/bark
anyone got any metatags to go over 300 bpm? i cant find combinations to get to speedcore levels
what kind of prompting for genre do you use?
speedcore, terrorcore, extratone
there could be some keywords that could make it go higher, like high-tempo, intense speedcore etc
would need to experiment
super fast? 😄
ive used intense but that jsut changes tone and snare sound
insanely fast did nothing atleast
yeah intense tends to do that
also order of words may matter
so you probably want the most meaningful part first
yea i noticed that, the earlier a tag is the more it matters
ive done alot of dnb, and if i leave that tag further back than say 3rd or 4th, suno just makes 175 bpm happy hardcore
Hey so I'm wondering. If I put down another months worth will I receive it immediately or do I have to wait for the turn over?
it will start when your current month refreshes...
if you UPgrade from Pro to Premium, there is an option to pro-rate the difference, but the refresh date stays the same, so it is not always wanted to have a bunch of credits and no time to spend them...
(there is also the TopUp option to buy some credits (that don't expire) to hold you until the refresh date
So I just paid again for a top up and got 2k credits
Who can help me?
At this time, the following options do not appear:
get whole music -Continue From This Clip
This makes it difficult to generate new music.
ui changed, there's now extend button
ok I realized, this serves to continue the music. and for " get whole music?
@sly goblet
basically
you select the clip and you have these options on the side menu, if you click on the three dots
they were on the previous version too, they just renamed continue to extend here
or
just click on the three docs directly on the clip
to get the whole song
I closed the session and started again and remains the same
@inner steeple can you help me?
is the clip you are using that for actually part2+? or the first one?
I can see the extend option on your clip too
if it's the first clip you are not supposed to have whole song option even
Before suno updated, it had all the features. Now, be it part1 or part2 does not appear
how to get better sound quality?i used to use remix and somehow the sound was better now that remix doesnt exit anymore
try it from here
if you use the side view, it may not always be obvious you picked the part2+
on this view it's super clear
Maybe Voce is right. I’ll test it. Thank you so much for the support. a strong hug
trying to test some Chinese dudes Bark Voice Cloning repo in a colab today, but trying to use the style of one Suno/chirp audio mp3 to transfer the style to another Suno/chirp mp3
i don't think voice cloning is music aware, but maybe i'll get lucky since Bark training data does include music and style transfer is a well documented Transformer Text ability
hello - any help from competent users would be very much appreciated.
Is there a trick to extending a song? my pet died very unexpectedly yesterday evening, and I created a song that provides me with comfort, but extending makes a completely different song. It loses tone, pacing, the same voice, etc.
Hey Moose, sorry to hear about your pet. Try changing the "Continue From" marker one second in either direction.
I appreciate it. I'll give that a try!
Edit: It appears to have worked, but there's going to be some messing around to link it all together. TYTY!
IMO, if you subscribe, there should be an option to not have to link 2/3 clips together for a 330-4 min song.
on the final clip, click the [...] button and choose Get Whole Song (free to use, no tokens)
it will stitch the whole song together seamlessly
https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
thank you! I appreciate the direction
testing is inconclusive because i cannot be bothered to convert from mp3 to wav this afternoon lol
any tips on getting the vocals to switch between men / womans voice between verses?
I hope you all have “fetus shredding a banjo” on your bingo cards.
Pleas help! Yesterday it deleted my generated songs out of nowhere! Today, when I remove the self-generated songs from the build, they reappear when I refresh the page! And they can't be removed!
Newbie question: In paid version, when providing your own lyrics, when the melody and rhythm is perfect, is there any way to fix a few words in the lyrics, and regenerate with the same voice/melody/drums? Like keeping the seed when compareing to other generative AI concepts?
JO! Put that in suggestions! Use Seeds to generate the same output/nearly same... that would make so much things better! 😮
this would be amazing
Devs calling this "in-painting"
it's a potential feature for the next model v4 (under development, but maybe months away)
Hello, Sorry if this has been asked , but are we going to see a way to restrict a generate song to a key like having as a tag : “Dmin” ?
I appreciate it. if you're interested, I think it came out pretty good. https://suno.com/song/416fbef9-5e73-402a-a494-85d8aa7bd622
Indie rock, male singer song. Listen and make your own with Suno.
I was extending like crazy and the video output didn't seem to sync the lyrics.
Also, Udio doesn't create songs like Suno as Suno has songs that sound so good that you makes you keep replaying them and replaying them in your head.
Suno should never change to match Udio in any way. Suno v3 is the best and it should get better not worse please.
Does anyone know of any way of converting a suno piece to sheet music? I have a choral piece I created that my choir love and wants to sing, but of course I need to transcribe it....
Share the song link please, I want to hear it.
You can try to stem it then use Melodyne for MIDI extraction
That is quite lovey. Feels very sincere. Glad it worked!
Is it possible to get only a instrumental version of a generated song?
epic, fantasy, choral, cinematic song. Listen and make your own with Suno.
Thank you, I'll give that a try!
Use UVR GUI (free on Github) to extract the Voice as clean as possible. Then you have your Instrumental Version
can also do vocalremover.org
But not as clean as UVR GUI i must admit
I'll have to check it out
UVR GUI can even extract huge reverb and denoise and so on. You can pick your own AI Model and combine different one to get even better results... this is stem heaven! xD
thanks for the tip
Help? Whenver I send /chirp I get This interaction failed? And then The application did not respond, I even tried direct msging the Bot?
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
Okay ty for the help
That's neat! Did you use the accents and ligatures on purpose so that the Latin is sung properly? What happens when you leave them out?
Does anyone know how to add sound effects?
The create button is blacked out I don't know the reason
Do you still have credits?
no
There is your answer 🙂
How do I get it?
You will get 50 new per day. Or you can buy a plan.
Free daily credits refresh at approximately 10pm EST, 3am UTC.
you get always 2 songs which cost 10 credits in total
thank you very much
huh, it will insert audience laughter if you give it prompts for stand up comedy
How do i generate a song with the same "singer" and melody, but with different text?
is this free?
You get 50 credits per day. If you use them all, you get 50 new tomorrow. Songs you make with free credits are not for commercial use though. For that you need a paid account.
Like, is that even possible?
so we cannot use song here and monetize it to youtube am i correct?
you need a paid account for that
thank you
so how can i generate songs here for testing?
awesome ... thanks
/chirp
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
how much does it cost for subscription here?
Free account is free and nets 50 credits daily. These are use it or lose it and do not roll over. The Standard and Premiere plans may cost differently depending on your location.
/chirp 床前明月光,疑是地上霜
/chirp
So many people trying to use the discontinued bot o_o
You need to go to https://suno.com
If you cannot access the website, your ISP may be blocking you so use VPN
Google translation (en -> zh)
您需要造訪 https://suno.com
如果您無法造訪該網站,您的 ISP 可能會封鎖您,因此請使用 VPN
Hi all, subscribed, made a few tunes and impressed with the results but got a few questions, so hope someone can help me:
1: Audio quality is questionable sometimes. Lot's of effect noise or overdrive on the bass, like the volume was set too loud, which distorts the audio
2: regardless if I set male vocals, there's a 50/50 chance the vocals are female or male
3: When I use Lyrics, the song will almost certainly start with singing immediately. I've tried to set:
[Instrumental]
[Instrumental intro]
[Instrumental build-up]
[Instrumental 40 seconds]
but none of these work consistently. I'd like a bit more control over the timing of verse, pre-chorus, chorus, bridge, drop, outro etc.
Is there any way I can control the vocals a bit? Sometimes the timing is great and sometimes it's terrible
Any suggestions?
\start
/chirp
¿Hay posibilidad de descargar las canciones que genero en formato .wav?
Totalmente de acuerdo en todo.
@errant dew
Hello, I have a paid subscription and I have want to ask you a question. I need to cut the beginning of my obtained (united) track. Previously you wrote that currently I can’t do this via Suno website, but I can download the mp3 file and manually edit it with any mp3 editor software. I have a question – do I have the possibility to upload an obtained (cropped) mp3 file back to your server and publish it in your chart?
No
No, just keep regenerating until they sound similar. Its frustrating but eventually itll happen.
/chirp
Korean pop male vocal fast bass song. Listen and make your own with Suno.
I can not get any introduction! The track jumps straight into the first lyrics.
I've tried:
[Intro]
[Intro - Acoustic guitar]
[Pause]
This is eating up credits!!!
/chirp
@ionic bloom @placid oyster @sinful ingot @steel cargo @lapis fossil @summer basalt @graceful condor @leaden grail
Chirp is retired, it can no longer be used on discord. Please use https://suno.com/ for the WebApp to make music.
I did! For the most part it seemed to work. Without the accents and ligatures there were some odd pronunciations!
Wow... not the same guy at all I'm sure.
🙄
I clicked an awesome song in Suno Explore then I clicked what will you create and the song disappeared. 😢
You should be able to extend off the song that was selected!
and having no volume control in that area is dumb
Hello to all,
I’d like to ask for your help in the following:
How can I generate a song that has man and woman sing (duet)?
https://www.suno.wiki/faq/style-and-lyrics/duets/ has information and you can use mine as example to learn from.. this is a duet I was working on: https://app.suno.ai/song/5045e5f8-d213-4cac-9d87-99b370e89adb/
intro, rockabilly, minimalistic song. Listen and make your own with Suno.
3Daizy thanks for the help.
I will test.
Have a great day
You too! 🌼
Try creating an instrumental first, then continuing from a good point of the insturmental by setting the resume time and adding the lyrics.
I do not believe this to be an option at this time. I do apologize.
Feel free to make a post in #💡┃suggestions-web about this though.
Ok, thanks. Already figured out this workaround. I'll post some suggestions in the correct channel that would greatly improve the service. Really excited about the future of this product though. It's only been a few months where it took 10 minutes to generate 2 minutes of audio with Facebook's engine running on a GPU, so the fact it's evolved so quickly is very promising.
Hi there !
Is there any possibility to change the picture of a song ?
Kind of.. The song will take the image of the last section that is used when you select Get Whole Song from the menu.
Argh. That's what i thought. Thanks anyway !
Hey! Everytime I try to extend a song I get either "credits refunded" or a few seconds of silence. I've tried with adding lyrics and tried doing it instrumental and tried on multiple songs, but no matter what it never works. What am I doing wrong?
How do i make a song ?
Workaround for this is to go back a few seconds when you extend from a song.
https://youtu.be/lj8utBd_IUM might be able to help you
Hello, I'm 3Daizy, a moderator and guide at Suno. Welcome to my channel, where I share my journey and experiences with AI-powered music creation. Whether you're a budding musician curious about AI or an experienced artist looking to innovate, you'll find useful insights and practical advice right here.
Join me as we navigate the fascinating int...
noice
Hi, I'm new to Suno and have a lil' question
is it possible to insert multiple style all at one?
Absolutely!
You can get some good results mixing styles. Think of it like mixing paint.
Did anyone test if getting male or female vocals is more consistent prompting it in another language? Like german for example Frau or weiblich (instead of woman or female) and Mann or männlich (instead of man or male) I have a suspicion but i don't want to spam credits if someone already tested.
oh, nice! how do I need to do it?
do I need to type it with a comma? or anything else?
||mean like: "pop, space themed, hard-rock" or like "pop | space themed | hard-rock" or what?||
That will work. Commas. Those are the only thing you need.
okay thanks
hi again, just found something 'interesting' so wanna ask about it
is it just I'm having bad luck, or Suno is designed to not complete a song in under 2 minutes and is it nearly impossible to get a complete song in just one generation?
because I've used all 50 credits to create 10 songs and got 9 of them ended in 1:58, and 4 out of 9 ended just right before the song was about to finish
even though in the end I got 1 song that ended in 2 minutes, completed
Your Part is Ending so the model predicts it has to End. Just cut before: e.g. 1:56 Instead of 2:00
You can extend it for far longer if you wished, you just need to use the Extend button
https://www.suno.wiki/faq/making-music/how-do-i-make-the-song-longer/
Bot is gone, use the web app instead https://suno.com
Hi, is there a way to reuse the same background while changing the lyrics of the song? I regenerated too many different tracks but I am unable to go back to the same song with edited lyrics (Sometimes the AI drags 1 sentence out for 2 bars)
okay what are the rules on whats aloud cuz somethings go thru and others dont like some songs swears dont get flagged and others they do
hi so on the website version, there is no more continue from button, is it just extend?
Continue was renamed to Extend and made into a button like a week ago.
Similarly, Remix was renamed to the more appropriate Reuse Prompt.
The method of extending stayed the same
https://www.suno.wiki/faq/making-music/my-words-are-wrong-after-continue/
oh I see...thank you
Most likely got censored from ChatGPT not liking something. Ask chatGPT itself about what was not appropriate about the [following content] and it should tell you
Unfortunately, the way that the process works, it is random BUT you should be able to reclaim some of the beat by Extending the music at the earliest part where you like it before the vocals.
You could try reusing the prompt but you already know how that goes
I tried but it changed the song entirely, it's all right then I guess I'll just generate until I find something else I like. Thanks!
what is the trick to prompt an outro end to a song? im struggling its coming out as a jumbled mess
manual lyrics
[end]
if you think you might still extend it
[break] might be better
subsequent generations tend to stumble if a previous "Extend" ended with the other 2 (just my experience)
Hi, is there any sure way to generate childrens voices but also maintain genre range?
oh i see
For references:
https://www.suno.wiki/faq/making-music/how-do-i-end-the-song/
Hello guys, does anybody know if it's at all possible not to make the vocals sound all chorus-y and autotuned? I'm not talking about inherent, let's make up a term, ai-fi, it really seems the vocals are autotuned on purpose
I am going to remove your double post and fairly warn you @fierce harbor
Hi all, why i cant see the chirp-beta channeld to create songs?
Chirp is dead.. long live the webapp: https://suno.com (where you can make songs now)
I found using 'clear ___ vocals' tends to help (until the chorus usually) though I know genre also plays an influence
/chip
Chirp is dead.. long live the webapp: https://suno.com/ (where you can make songs now)
@hollow hare how ?
Is there a way to make Suno do a duet between a guy and a girl?
Any and all help is appreciated! 🙂
See my previous post about this today: #📚┃suno-school message
chirp
Chirp is dead.. long live the webapp: https://suno.com/ (where you can make songs now)
Found a way of getting better generative lyrics. Basically put next to the bracket a description what the lyrics are about 3 minimum lines of Lyrics/Chorus
Example [Verse] About Ice cream
[Chorus] Talk about green flavor
Hit generate lyrics
Hey what does it mean when your prompt is flagged for moderation? I was making a song about playing fallout 76 with my friend tonight
is there a way to search throuh Suno public music? the main website only has a few displayed tracks i want to hear what other people have been making but search for specific terms
well, that wasn't the point
what I wanted was having a song, with just one generation. Because, you know, I'm using the free version with limited credit.
also, I kinda found out that Suno is quite unreliable if you wanna have a similar song extension (like having the same chorus part)
Hi, the 'how to make music with chirp' channel says failed to load messages
Chirp is retired.
try the web app at https://suno.com
And here are some starter docs:
https://www.suno.wiki/faq/getting-started/
the clip you are extending from has ended the song, and your continuation clips are dead air... that's why they are being refunded.
adjust your Extend From Timestamp a few seconds before the ending, to bypass the problem
hi /chirp not working in my server
/chirp
@sick parrot @pallid kelp
Chirp (which ran ver. 1) has been retired for a while now. Use the web app at https://suno.com (and use the superior ver. 3)
Also read through the wiki: https://suno.wiki
Is it chirp now or still bark? #💬┃general-chat message
Chirp no longer working in discord ?
/chrip
Hey, guys. I'm new here. In the sense that I wasn't even familiar with this application before. Can someone tell me what the member avatars in my "posts" menu mean? They have just been there for a long time and what do I need to do to remove them? I'm sending a screenshot for clarity.
How do I make the second verse after a chorus sound like the first one? The parts are labelled for me so if I create a second verse (part 2) right after it will create a new "part 3" but my current "part 3" is the chorus... lol. Is there a way to rearrange?
Yo mother
There's no way to mass select songs to mass delete. There is also no way to empty the trash section.
I want to generate like nuts and worried I will not be able to find what I want.
I don't think there is a search section too:
discord mobile is weird lately, but i think these are your friends, or the people you were interacting with in a server, i dont really know, sometimes i get unknown person in there too
The songs have the same title and description. There is absolutely no way to search for what you want.
There needs to be a way to leave custom notes and descriptions to see for searching.
Songs needs to be moved into a folder that can go down a few hierarchies please.
It could increase server usage stuff.
I'm unsure how Suno is setup but if it makes Suno worse, then keep it how it is.
Oh, thanks for the tip.🙏
You were right after all. I get it now. It shows the members I've been chatting with in different chat rooms. I just went to the menu and the avatars changed. Lol
Rename songs, use playlists
@dim glacier @broken lotus
nice ban hammer
I agree, there should be a way to separate projects, there are some things we can do as @royal ibex says however it doesn't help when you want to look at something a page over when you're creating
Although thankfully it's not too much hassle to go to library and back to create
yea you can highlight multiple songs using Shift and or Ctrl and using the three dots option move to trash, same in the trash to permanently delete
I recommend clearing as you go, keeping what you think you might use and renaming outputs if they are significantly different but with same/similar lyrics
It's not a perfect system, I'm sure it will be improved with the help of yours and everyone elses feedback
make full song, cut it short where you want to or when times up and make the continuation, the AI will re-use the same verses if you repeat them later, which makes things sound more uniform like it's meant to be one track, not 2-4 stitched together
does anyone know how to get less noise and impose a single singer?
you could try the intimate prompt - in the song type
ok thanks, and impose a voice timbre / tone of voice (there's a level to specify, perhaps?) ?
hmmm, spoken word or resonant prompt maybe
maybe in the lyrics you could put [sing softly]
etc
the AI doesn't always do as asked and you may need to generate a few times before you get to a point where you think you're on track
Coaxing the single singer is like convincing a squirrel to eat out of the hand...
it feels like we need to do LESS, and try to get the genre to do it... (rock seems to be some of the hardest, soul/rnb voices are usually fine)
it can be done, but one or two words seem to break the prompt (it effects other sound qualities too)
slow sinistercore, melodic bass, sultry percussion, heat, male
(and I don't know why this prompt worked but others don't)
I'm looking to calibrate a voice that could appear in several songs :
Voice Types: soprano, mezzo-soprano, contralto, tenor, baritone, bass, countertenor.
Timbres: bright, dark, rich, clear, warm, metallic, soft, powerful.
Pitch Levels: high, medium, low.
I've also found the more you stitch on it can gain large amounts of static/white noise
I'm experimenting with Rap and it's very complicated to have a voice without noise and without other voices around.
all these tips sound like things I hit too...
I try to prompt some kind of intimacy... (bedroom, close, intimate) which sometimes works but often makes him sing off-key...
'big' words like 'anthemic' seem to make the chorus voices more
– but again it's like standing still hoping a squirrel will come... I don't know a sure trick...
@knotty skiff "more you stitch" ?
I've heard some 80s/90s rap/hip-hop that was very clear, it is doable, you might want to look in the genre and see what others have done and try to replicate
when you extend, then make into a full song
you can do that basically forever
ah ok, then i extend one time and get full song and extend anymore the full song ?
tbh i haven't tried limit testing
yup that keeps it more simple for yourself
and is the 5-minute wait for clear sound a legend?
i'm not sure what you are referring to
the audio looses quality each time.., reverb and lots of echo can carry over into the extended clip as a ringing noise that doesn't go away, and keeps getting worse...
essentially if you hear a compromised 1st clip it will be very compromised by the 4th part... We can do a little surgery with Extend from Time, try to trim around quiet breaks..., but if you have metal or shoegaze genre with lots of distortion guitars, it's gonna sound pretty tinny by the end...
I studied the question and they just put a "90s rap" in the style... maybe they did a lot of testing.
so should we or shouldn't we "get full song" to improve quality or limit extends as much as possible?
ah i see what you mean
it helps but for what I'm pretty sure you're referring to it won't help with, some generations are just better than others
if you are extending from time on a Part 2, and it is less than 1 minute, it's better to Get Whole Song and Extend from the Full Song, so the bot can 'hear' more of the song...
if you want to do a sudden style change, we can extend from time (where the change should happen) then extend on that clip and set the timestamp super short (00:01) – this second extend only has a 1sec memory of the song so it will be easy to change with style prompt...
basically use Full Songs when you want the bot to hear the most..., and force it to forget using 2 extends in a row...
someone mentioned rock being some of the worst for static, I've noticed that it puts in the most when you add [Chorus] prompts in the lyrics, there's not much you can do about this other than re-gen or extend between verses
@broken lotus I understand the logic, thank you. Does this influence the quality?
@knotty skiff in Rap the chorus gives very bad results with a lot of noise, unfortunately, so I don't use it.
what sort of rap
it shouldn't... I personally just generate a few starters and extend on ones thaat already have a good sound quality...
A lot of times the songs are doing something musically great but the audio is tinny or weird from the start, so I consider those loses...
@sand crypt French, 90s
can you link a demo?
100% agree, this is also what I do
Good result but many credits behind ^^
you mean the echoed voices?
behind chorus