#🗣️┃chill-chat
1 messages · Page 13 of 1
I like trying classic terms like 'fugue, baroque, martial' with a few electronic words:
https://suno.com/song/0f162dae-c040-4585-b23c-dbedb7485c8d
https://suno.com/song/cc5602a9-1fc1-4476-8048-93438ee2fbcf
vocals, fugue, baroque, surround-glitchcore, weird trumpet, , martial, surround ",",",",," acoustic haunting
",, song. Listen and make your own with Suno.
vocal fugue baroque, female weird-trumpet vocals, binaural-vocals, virtuoso acoustic-folk, spare vocals song. Listen and make your own with Suno.
@runic depot Oh check 1:50+ on that second clip for some cool stereo vibes (I think that particular segment is V2 actually) https://suno.com/song/cc5602a9-1fc1-4476-8048-93438ee2fbcf
Interesting how you use the morse code to induce nonsense lyrics
I think you normally don't really need it, as long as you have "vocals" or related words in the style prompt. But I happened to be trying it here.
I can't test very much because udio is so slow, but I think I found something that is in suno's training and not udio's
hey guys i need some help i want to know that does songs generatd by suno.ai get monetised on youtube or not ?
Yes on any Suno plan that is not the free plan.
@grim idol i have a paid plan. Will my channel be eligible for monetisation ?
I can't speak for YouTube, but you are good as far as Suno is concerned at that point.
@grim idol do you have any channel or songs from which you are earning?
Sorry, I have a YouTube but I don't monetize it. Or remember that I have it for years at a time lol.
She threatened to time him out because he was confused? That's unprofessional if you ask me
That's really annoying
Clearly you did! You had no idea what he was saying in Spanish. Vito did though. Unprofessional message. Can't believe this server came to this 😓
I think the user had gotten this from Suno
https://suno.com/song/e7a4780a-ccb4-41e8-a591-306717db827f
But they are aiming for this:
https://www.youtube.com/watch?v=l0kc50OdmqA
Dark psytrance, hitech trance, 170 bpm tempo, russian folk, hitech psytranc,
futuristic sound effects, Insane Cosmonaut song. Listen and make your own with Suno.
Shiva by Furious
Album - Beyond Atom by Quantum Frog Productions
Genre - Psychedelic
Quantum Frog invites you to go BEYOND ATOM with special 2 CD compilation...Engineering the mechanics of harmony, facilitating the synchronization of man and environment through music. The arrangements you will hear are reflections of music as geometry, with inc...
I couldn't get a bassline in particular to be that fast
And they left because of that. Great
hmm, hmm
I have some ideas, but that's a toughie
I think I did match overall speed but only as basically a pure chiptune, not that close to the youtube song genre
off-topic, ironically, seems to end up having the most technical Suno prompting discussions lol
so you can see the pacing is coming
it's transferring it from cello to bass that will take a sec
but that's the general idea
I'm grocery shopping tho so I'll come back if it isn't resolved
At least I still have respect for Jonathan Fly and The Hepcat. At least according to what happened right there. I've been in this server for months and never had problems like that
you are supposed to only prompt with FOOD WORDS while grocery shopping. i don't make the rules, just how it is.
Yeah I've been here a month before you. It's a lovely helpful community
Shots fired
Idk man
the user looking for hitech trance @covert oxide so I'll ping them just in case i forget to come back and you figure out the most epic way to prompt it
Hitech me? You won't find that ☹️
a little closer but also further. more like horizontal
post-anti-instrumentalized anti-hyper-gothline song. Listen and make your own with Suno.
it has moments, and alternates
it's sister is also fast
I feel like if someone had the credits, they might be able ot get there
what I did was find somewhere with a fast baseline and continue towards it's end with a prompt that has fast pace. You could likely chain this a few times to get it into a cycle
the issue then is trying to get it to maintain that while introducing more beats. I suspect you will have trouble
this one has the closest in pacing though
it's just not the right soind
i might try something like spamming 'hyper' 10x, later when I am messing aroudn
what are we trying fast bassline?
it incidentally increases pace
but ultr
ultra
I remember from Ultra-bach
it made some unlistenable stuff
It was a user question here: #🗣️┃chill-chat message
because the pacing was brutal
haha, I just like doing weird things users are trying to make
Alright i'll try as well later 😄
and I had like 250 credits
plz do
we should have little contests
like prompting contests
that would be kinda 🔥
it would help the community as it would force people to share prompts
In general, I would watch watch generative AI competition lol
my latest "thing" i like doing is transfer from instrument to instrument but keep melody
I just like figuring out how Suno interprets the tokens. Like, it is surprisingly context aware.
People are mistaken about certain prompts I use, they absolutely change contextually, it has much better understanding of adjectives than Udio for example
as in, in udio if I prompt, it works best with full comma separation
it 'understands' a lot yes sometimes even from the context of the song.. which can both be helpful and annoying if it's very stubborn about it's 'understanding'
whereas with Suno, putting an anti- in front of two different things results in two different effects. More interestingly, it even can shift its understanding of one instrument to another. So if I have a prompt which causes a fact trembling cello, I can manipulate that into, for example, a bass with "basified anti-hyper-cello" for example (I haven't tested with particular one, but similar things have worked for me).
not right but I liiiike weirdly
https://suno.com/song/fe016507-747f-4db3-909a-2d85c7590b12
oh, and this one from earlier was accidentally a dope prompt I'll be using
https://suno.com/song/6abf8db0-a4af-40d4-b4f5-89641e3fa1af
it's basic but I enjoy
post-anti-instrumentalized anti-hyper-gothline song. Listen and make your own with Suno.
always interesting to see this kinda stuff
post-anti-instrumentalized anti-hyper-gothline song. Listen and make your own with Suno.
full song
could improve the outro but it is a complete end
Am I hearing things or are there birds on the transition to this?
Birds are one of the most common non-music sounds all versions of Suno music have generated. At first I thought it was something they did because the Suno Discord command was literally /chirp but probably just lots of songs use nature/bird sounds and it really is a common sound.
yaaaay you can generate cover art now with custom prompts
so in the end I'd say like 1/4 of my tracks will probably make sense on Udio
Suno is better in most cases
Is that a new feature? Damn that is some fast work
Oh is it finally working? Yesterday I had to work around that with doing an intro and stuffing the image "prompts" as tags.. 😄
ya, they didn't even let you heart tracks for a hwile
now you can
here's how you can change covers
Oh thats really cool Yesterday you could not save the cover after editing it just errored out everytime.
So I just made an intro with these Tags:
Which gave me this
it errors out still
just refresh
I wish I could just add my own art
Ah yeah that did not work yesterday was always stuck with these colorful blocks as cover art
Also the auto prompt "improver" really caught me offguard.. tried to do robot standup comedy ended up with a horror movie or something
https://www.udio.com/songs/kqNjLuJuQwRiNrAuX9K8tD
It is good at taking a short and boring list of tags and bulking out the list and spicing it up, but it can go very wrong. The prompt here is "an extremely fast Hitech psytrance speedcore song"
Make your music
Yea great if it works.. if it doesn't not so much 😄
I think the software is perfect for dialogue, only the voice interchange all day
it's not bad for music either, it's just a lot more annoying
I will say their prompt fixer is inferior
to creative prompting
which is obvious, but still, not always the case
depending on how they structured/cleaned their training data
lmao started with psychedelic psytrance... it got the psychedelic alright i guess 🤷♂️
yea, ultra does def get that speed
"drugs drugs drugs"
you could try ultra-meth, as it's also semantically close to math, which also lends itself well to speed.
i feel like I get usable/high quality results more consistently from udio, and vocals tend to sounds clearer
https://www.udio.com/songs/c8oM7ZpAV9qWdP17TUaSjr
I'd love a sneak peek into the future to see how all of this plays out. Children prompting hologrammed cartoons on their way to school
i'm in awe of the way this works as a joke and genuine suno prompting advice
the best style prompts are the ones that have a pinch of meme in it and still lead to great results
like post-nasal drip
if only I had suno credits; this is actually one I made in v2 a long time ago
and also in v3
alternative rock, lo-fi, space-rock revival, chillwave, slowcore song. Listen and make your own with Suno.
I'll gen it for you dm me any prompts
they should give you a sponsorship already
credit gifting would be cool. i'd start the Yolkhead prompting fellowship. i think i'd save credits in the long run because if I mention something he often just goes and puzzles out a solution before i even try anything.
they get mad input in exchange ^^
Interesting... hyper binaural violin asmrcore
https://suno.com/song/8eb3679f-3ee5-4040-9fca-b5930c9d8620
i love binaural, but I wish I could find a similar prompt that didn't push the music into ambient or atmospheric genres so much. Probably not realistic but maybe songs could come out like https://www.youtube.com/watch?v=IUDTlvagjJA
You'll need headphones for this to work. Make sure there is no noise around, close your eyes, turn the volume up a little bit, hold and press headphone on your ears for better hearing, play it and enjoy this impressive audio work ( illusion ). Not a screamer not a prank!
Audio credit: The original audio is created by QSound Lab. I put it into t...
Yea that was kind of my goal but did not work out like I wanted.. still interesting effects it gives
udio just sound like a little competition so far. suno has the first mover advantage and probably a slightly more active discord... but i guess i will check there too
hyper binaural chipped metal ultra asmr, polyphonic polyrythm song. Listen and make your own with Suno.
OK, so, this has left me very conflicted. I'm a big fan of Suno and been enjoying it for months now. But Udio has a lot of pluses going for it right now. I wish the vocals were as good in Suno as they are in Udio.
yeah Udio has some awesome quality
https://www.udio.com/songs/teBRaxvZVUKpR44hgo4UWA
I've tried about 10 gens in Suno with this song and it keeps making choruses and layered singers. Muddled sound.
Udio, pretty much out the gate.
I just hate the workflow.
Reminds me of trying to learn ComfyUI when I got used to Automatic1111.
yeah udio quality is great but the workflow is a pain
and it always forgets the chorus once it gets to the second one
OK, thank you, there's a big problem that no one is addressing, and it's that Udio changes the "feel" up a lot and doesn't repeat the same rhythm or style for each stanza block.
You can mask it a bit though. Gotta start in the smack dab middle and burn through gens.
id assume its because we only get 30 seconds each time but idk
luckily i rewrite my songs so i could just copy the chorus i like best
but it doesnt help that the verse could also just be totally different melodies
it just doesnt feel like were really continuing our songs as much as rewriting parts of songs that are very similar in style and just mashing them all together
hi guys! i've just arrived, can anyone help me? i'd like to know if i can make a api request for SUNO and what their documentation is, does anyone know?
Yeah, especially with how it handles words. It just speeds up talking to fit in the 30 second block. Makes rhythm of the words weird.
You can not.
Behold the devolves into noise machine (Volume Warning):
https://suno.com/song/9901583e-72b7-4e07-a8f7-1defe7d177f2
(I should stop wasting credits now)
hyper binaural recording chipped orchestra ultra asmr, polyphonic polyrythm opera song. Listen and make your own with Suno.
so it's not possible to make requests to the application in any way? just using the site?
yes
Far as I know that is currently the case, yes.
got it! thanks a lot guys
just to give some perspective, I started working on this song when we discussed this and JUST finished it. https://www.udio.com/songs/mc8PRrbLurU8WH7uYUB1sP
Make your music
it just takes SO LONG between choosing parts and long generation times
Ok im actually getting something decent with Udio by just random generation, the only lyrics im manually typing are to make it repeat the chorus it generated... stay tuned
Udio does better at lyric generating ive noticed. still nothing spectacular but better than Suno
i wonder what prompt they use for the lyrics
cuz i imagine they use GPT as well
Gemini I am guessing
i had TERRIBLE luck with gemini but it was probably my fault
Its actually not my style of music entirely
Like its not what i would choose in my playlist for gaming
but it is what i would consider a good rock song
one of the angel investors is Oriol Vinyals (head of Gemini at Google)
ohhhh yeah then i bet they use Gemini lol
one of my biggest gripes is when i click the delete button the little "are you sure box" goes away before i can click it
welp now its stuck generating
Really? That's interesting to me. I, personally, find it's lyrics garbage. I've been using Claude a lot lately.
ive only done a couple auto-generated lyrics to be fair lol
I've heard the auto lyrics build(?) on each other from your prompt?
I don't like LLM style lyrics to that degree though, so that doesn't appeal to me much.
yea, this is the real problem
it's inferior because of this
btw have you tried suno lyrics recently...
i actually havent
have they improved?
and again this means it also cannot properly cohere longer threads of the song together
at least suno partially does, sometimes quite logn threads
something changed?
I tend to focus more on how the lyrics sound rather than what they actually are saying
irl I agree with you considering nirvana and Pearl jam
oh yeah Suno looks like they improved lyric gen too
but for the model, it matters
Like i love Starset, but some of their lyrics are kinda meh if you actually pay attention to what the words are saying
seems like its not just doing ABAB, 4 lines, over and over
also they still had good lyrics
i photographed Starset
which again, means for you to get their sounds your lyrics matter, for example
are we not basically training udio atm?
and the outputs will be way longer after the beta than the issues will be gone anyway right?
one of my favorite photos ive ever taken was of them lol
I doubt that
they will need a new model
for longer context
so a month or more
they'll get runway tho
there's money to be made
ok but that's fine anyway. at least to me.. I don't collect songs anyway, only some
every 2nd prompt is either real stupid or a joke to some degree
rn some serious money probably yes
but wait if there's more competition the stocks crash or plummet
Yeah Udio loses the chorus kinda
@late crown I tested the remixing and the thing with the seed.
For me I can slightly change the text and it stays sameish.. but it still changes a bit depending how much text is changed Here are some examples:
https://www.udio.com/playlists/3k9v7RS45jkTKgMZitqRUt
Make your music
OH I see whats happening with the delete confirmation disappearing, if its generating the page kinda "refreshes" in stages of generation it seems
suno lyric gen seems to be a lot better but it still does that thing sometimes where it drops the genre into the lyrics:
Chorus]
By the way
By the way
Won't you stay?
In this world of shadows and shades of gray
By the way
By the way
Let's find our place
In this indie folk song
A story we embrace
lmao
Are you using the simple or custom mode?
In custom mode i just type what i want the song to be about only, i dont specify anything about genere except in the style
i told it to write an indie folk song called "by the way" in the custom lyrics box then clicked generate lyrics
So you only used remix, and changed a few words?
nah we're still bubblin
well, it's not a bubble
it's real
Yeah if im in custom mode i keep the genre out of the lyrics box, i only put that in style
yeah i usually do that too
Yes, I tried the thing with the seed. It only works if you change one word or so if more then totally different so remix and the seed workaround seem to be different things
but like, there is going to be too much competition imo very soon. I think suno has a good position atm as they much be sitting on a ton of user preference data
wanted to include it because genre can impact lyrics. didnt want hiphop style lyrics lol
I've been trying to get Dio's voice for the past hour 🙂 like that one? I somehow - agree 😄
and then have that msft 💵 I believe
If udio would speed the heck up I could try somthing with the sample you posted yesterday
Interesting! Thank for letting me know 💗 🌼
Tried Udio first time today, i LOVE Suno but i feel that Udio have some nice features and overall quality feels better for now, was able to do that https://www.udio.com/songs/72yin5FbvRep9cFpESTe7r
Was the exact text in the lyrics field indie folk song called "by the way" or write an indie folk song called "by the way" (I think first is best.)
interesting
the first
Sadly remix only works for the first part. Thats why I wanted to check the seed thing.. but I can't make it work properly yet
Well at least you still have access.. New people have to wait for that
IDK, my experience with Udio is that Suno is not just better, but way ahead. Every 2nd and third generation (extension) is terrible, or at least not great. Suno is ahead, currently. The future, we'll see.
udio stopped letting new people sign up?
Udio has a small waitlist, though it was only a few hours for me. Though that was early in the AM (EST) yesterday.
So may be longer now.
They did yesterday but its fine now
I don't know but this is my guess for how a seed would work in Suno. Basically maybe only if you change a tiny bit at the end of your prompt does it say sort of the same. So it doesn't give you 'the same voice' with a new prompt. But... just that little bit would not be useless, even if it doesn't do what most people expect. Just pretty low impact.
so it progresses which is desirable, but yea, you can see it can't look back more than 30 secs
In about [however long it takes to generate on udio] I think i'll have something decent
try a string of numbers
or a single emoji
I found the udio one
the size kinda
i’ll listen when i’m on my PC
gives that away
Yeah i mean i dont know, the track i just posted i am unsure if i would have been able to do it in Suno
Length of audio too.
So thats my problem, you can tell the Suno ones sound very similar, but Udio took it a whole different direction, i prefer both suno's better as far as it being closer to a sound of music i like
Udio generation with extending is really quite pleasing, the lyrics sound a lot less like AI for one. But I put in a whole D&D character's journal and backstory from one of my players, and each extension of the song, covered the next event in the journal, and understood events that happened at the end of the journal, and foreshadowed it (not so subtlety, but still). Quite impressive honestly, for something done on a whim and as a stress test. And we're talking 21 pages of Google Docs.
You can tell easily which one is Udio.
is udio free commersial use?
It's obvious right?
cannot create a song that's built with longer verses until you get the chorus. Very short output and does not stay the course throughout a song. It can make enjoyable songs but not traditional structured songs.
if i had to guess id say you uploaded them in the order you labeled. a is 2, b is 3, and c is udio?
on 2nd guess I wouldn't put money
I don't understand, what do you mean? You made the verses or it auto created them and you just dumped the journal entries?
but I call B suno
my reasoning for my guesses, last one had a 30ish second intro, so Udio, and B had the choir vocals lol
I dumped the journal entries and it made verses in order of occurrence, with a chorus to separate them.
Really? What's the text length? Did you have to break it up into ton of little parts?
If you play C, you can see that it stops/starts every 33 sec. Suno is way better for long songs, without interruptions. And Suno may one day allow double input and output for full length songs, Udio is trying to be a credit eater. Suno has a better business model, geared towards creating songs. Udio has a model geared at collecting credits. I know they're currently "free", but each generation is tiny/small.
ignoring the 30 second intro, i found the background music to be more unique in 3 as well which leads me to udio
I didn't break it up no, it was 21 full pages of text, but I can't say exact numbers of characters without pulling it up, and I'm a bit lazy. I did not break up the journal from the backstory either (which was 6 of the 21 pages).
One thing worth trying is using the 'optmized' udio prompts in Suno. It's often not bad at all. So not the original prompt but this list here that udio adds extra words to.
I copied those udio song into Suno for example:
https://suno.com/song/b4c1c8f0-1e54-46b1-8e4f-72facf019bfa
the other gens: https://suno.com/playlist/325d59fc-5906-4c89-9555-c2daa20041fd
Female, Indie-pop, dream, Synthpop, Melancholic, Bittersweet, Melodic, Introspective, Ethereal, Dream pop, Bittersweet song. Listen and make your own with Suno.
ive been doing that a bunch tbh
basically use Udio to generate a song based on an artist, then copy the prompt over to suno to see what it does
Yeah cool, it sound good too you are right, worth it
i havent gotten anything great yet but its been fun to mess with
I do tweak the prompts a little for some words I know are not what I want in Suno, but sometimes not at all. Though usually you have to remove some words just to fit in length leimit.
I would say, above all else Udio did a great marketing advertisement. Suno needs to up its game or Udio can win this race. It's more than just the tech, it's about creators and songs. There needs to be a rush of Suno creations on social media.
Doing that helped me realize that vocalist being the first prompt usually got better singer quality.
Wow... gonna have to test that later tonight.
Yes, however you do have to attribute it
Which kinda makes it tacky
like, I am open about using ai, but I don't like specifically having to name what tools I used in song credits
my issue with Udio is I actually it's less creative, and people are confusing it's lack of context for creativity by virtue of rhe richer sound
it's overtrained
imo
I cam get crazy unique outputs from suno
they're just obvi using a much much more refined model
that's awesome, that it tells you what it did and clears the smoke a little
so it doesnt get shashmaqam eg
I have a feeling even just using a long chunky prompt (lots of diverse words) is better for voices over a shorter one using only common words. Not specifically tested or anything just a vibe
Maybe it's me but i just split stems and i find it easier to work with Udio
It's not about who is better. Suno is way ahead. But, Udio is what people are talking about. Suno needs a splash 3.5 or whatever within the week to get on top of conversation again.
Also, Social Media blasts
maybe "better" tagging of the datset which would explain why it doesn't diverge in creative ways because it has a very rigid definition of the tags.. but idk I have not ran enough tests yet
There was a noice ,like background noice, was able to isolate it and then remove it
may be single parameter set to 'play it a little extra safe and standard' for launch day craziness
no idea on my end. I mean, the main thing I find is, if I slam together a weird musical descriptor, Udio tends to do basically what SDXL overtrainig does, namely you get a bunch of abberations.
it can't "transfer" the knowledge to build something and just pulls out bits and pieces?
Re-release v3a as v3.5. 😛 Ok, just my dream, lol.
yea, I mean, this is care about so little. I think there are definitely a few people who can actually master a track well.
But there are even people in the industry imo who suck as mastering and just absolutely remove all the natural sound.
Like a lot of music has awful background noise, fuzziness, people literally willingly choose to listen to record players
Haiz, the song generated by Udio so far is the best to compare with Suno. I use the same lyrics in the same genre, but Udio is better. I just hope Suno does something; maybe Udio, in the beta stage, gave their biggest commitment, and that might change for their product once they hit the market.
Suno just came out with a new model last month
I mean, it's machine learning so the pace is insane
but like, they're def always working on their models, like everyone atm
I am sure that Suno will improve way faster
It's been fun to work with. I've got some pieces that are close to the sound quality as this, but the vocals tend to go askew in Suno.
https://www.udio.com/songs/44LRjXAxwWwpmRVYYjK14g
just hope suno can solve the background noise issue because "udio" gives a different experience in their audio quality
could be idk. I think it's subjective personally. Like, the music on fallout 3 could be re-recorded now and soind so crisp, you'd realize how awful the sound quality is. But it sounds good.
so like, I think all the stem splitting is like, eh. It's like I writerss groups, people become convinced certain rules will make their writing better and it's like, eh, at best it's a horizontal change
bc it really depends on your audience
my earliest memories are listening to music that sounds bad and being like "what the hell is that?"
and wanting to listen to all of it lol
Of course I want better sound though
like, who doesn't, udio beats on quality
of the individual sound
FYI : I went over to the Udio Headquarters and [Took a Shit] on the Udio office desk
SUNO FOREVER BABY !!!!
i agree, the reverb/noise/echo in the slow songs is disturbing
Once you hear, you cannot un-hear.
You are absolutely right! I hope in future you can clean that sound with ai
Final version with ''video'' and noises removed with RipX, quite happy with result https://youtu.be/oFEYpPeZYBo will do one more tonight with Udio will see
Ever feel like your TV remote is fused to your hand? This song is for all of us trapped in the binge-watch cycle...
The TV whispers your name, the remote feels glued to your hand... we've all been there! This song explores the battle between wanting to break free from endless binges and the irresistible pull of the next episode.
Did this song...
I reeeeally wan't a more vulgar llm ... -.-
flag for weener or fap or sth it's pathetic. language is free. it's art, what the fatwa
Why going vulgar? They is plenty of way to express the same without being vulgar
and if I want to try some beavis and butthead type stuff or jack it in san diego?
They i guess you record your own voice and you work it after ;P
OK, so it appears that help
ya i just ask for slightly different words and correct that afterwards but it's work
This isnt my style of music at all, but its something that i would recognize as good vocals for the most part from Udio
Make your music
I am quite Happy with AI music, was waiting for something to open a Youtube Channel so i did, i am not two weeks into it an already have 2.3k views and over 36 hours of view lol
Have to say that i am 100% sure now that audio quality is better as we speak than SUno
yea, audio quality def is better
that is clear, there is a LOT more information, the files are massive
AI music is def making things interesting
I haven't seen a ton on Spotify yet, mostly just metal
I've been wondering when I'll run into another me
Anti AI group on FB are getting Livid
What are you guys listening through? what type of audio device?
Cause like i dont really hear much difference in audio quality, the vocals are a little clearer thats about it
fr though, AI drf has potential
but it isn't disrupting yet bc most of the people who can make music already were
and they are institutionalized opposed
so it will take a bit before it moves beyond jingles and dance
and hey, I like some dance don't get me wrong
but edm happens to be a genre that I imagine will incorporate it into traditional workflows
since live shows are basicslly pretending to twist dials
I mean i never was making music before, im nowhere near that area of expertise
well, 2.4k views is great
Everyone in the live audience write one word on a whiteboard. Take a picture of the words, randomly pick 10, that's the style prompt. DANCE.
Even better if you can change the style prompt over time as the signs change... (it's got to be almost fast enough already for that)
yea, it's pretty close
I think if it were local, it would be fast enough
I mean for me im not really posting anywhere, except here to share them individually or something, even if i was a subscriber i wouldnt put them on spotify or anything, thats just me personally, im here to add music to my playlist for gaming lol
for sure I get that, I think that's a great use
I think that, writing for family/friends/jokes
like, it's trivially easy to crank out some stuff that si just fun
plus honestly it would be a great party game
if someone designed something around it
Thats what i did with one of my helldivers songs honestly, its a joke song from when i was little about "bug guts on the windshield" https://suno.com/song/12365596-da12-4bea-afc1-f344b1d714ab
satire aggressive hard rock song. Listen and make your own with Suno.
my dad used to sing "bug guts on the windshield" in a like bluegrass tune when a bug would hit the windshield
it's great w kids as well
post-anti-folk post-anti-indietronica anti-instrumental post-revival, midwest emo, female vocalist song. Listen and make your own with Suno.
easy way to kill an afternoon
I still tend to like what i get out of suno more than Udio
same. When udio works, it's really good
but some things it's like beating your head against a wall
Suno however, could make music out of a hairdryer
Yeah ive generated over and over and over and i like none of the vocals on Udio
but i guess its a good thing i got more free credits on Udio
Think where I've arrived at is:
- Suno to get better quality data so the outputs are higher quality, like Udio. Clearly this is going to happen since it's the natrual 'thing' to upgrade to.
- Suno to get infill/inpainting
Suno is better, but I really miss that quality from those generations where Udio works
Basically I expect Suno v4 to erase this
I have for Udio at the moment
agreed, I got an amazing Neil Young-like vocal out of Udio but wrangling it to make a complete song that stays on track is a slog. But, conversely Suno (for now) could never give me a vocal like that.
I think i just have to figure out the style prompt for Udio, cause the same words that get me exactly the type of sound im looking for on Suno gives me trash on Udio
and I've really just tried to re-make songs I've made in Suno in Udio to see how it handles it
same
Thats what im trying now yeah
I found that "porting" songs to Udio did not give me results I was happy with. Kinda need to start from scratch
Well its not a whole song its the first part of a suno song i havent continued yet
I've had a bit of luck
https://www.udio.com/songs/wMwEXQByFZXVKdG7qy8Qa7
Make your music
Musically I'd give it to Suno, but quality-wise and uh, spontaneity? To Udio, but once Uido decides on a thing it won't move from it
and it has a real hard time keeping choruses
Is there a 10000 emoji? LOL
Initial creative busrt from Udio is excellent, but hope you like it because that's where its staying
and as Yolkhead found out a couple days ago, Udio is far better at 90s slacker, alternative rock than Suno is
https://www.udio.com/songs/7TH5PnoaEKfXfVAFc6Dd8b
Make your music
Each portion of the song feels like a whole new song. Just roll the dice until it's similar enough that it might be the same song.
the chorus to my songs on Udio have been terrible compared to the magic Suno puts out. I stopped using the [chorus] tag altogether on some songs
Yeah same and instead just kinda let it go where it wants and see what happens
The more I mess with Udio the more I wonder if a quality increase with Suno will 'fix' lots of things when I compare them
I wish I didn't but I can hear the difference so clearly
totally, just waiting on that sweet sweet v4
With no evidence and no special skills, I feel like it would be easier to add higher quality to the Suno model than to get the Udio one to do sensible things. 🙂
I suspect, based on how they talk about it, that each 'model' has a kind of data that goes into it
as in, v4 will have a different training set than v3
We saw it with the moves from 1 to 2 then to 3
Even gained stereo between 1 and 2
So basically until it becomes irrelevant we're probably going to see quality increases with each Suno version
irrelevant = good enough. I don't imagine trying to do this AI LLM stuff with flac is realistic
I think thats the thing like for me, being someone who's not really in the music industry and who's goal is music for my gaming playlist, which is mp3 files played through computer speakers, suno is plenty good enough
It isn't quite there for me. I thought it might be but now I'm poisoned with Udio
I know there is better and that thought sits on the surface of my head
I listen in headphones, on a bluetooth speaker and in my truck. Some of the Suno songs in my truck have mega bass that doesn't come through via other listening devices. I basically sound check my songs these 3 ways to get an idea of the quality I'm satisfied with. With my headphones i can pick up background static w Suno that I haven't found with Udio. But Udio is taking mushrooms and hallucinates all over the place
dang yeah the bass
I have headphones too, but also a buttkicker (https://thebuttkicker.com/) and Suno does have a maybe 3x bass to it I can only hear out in the living room
I drive older 80s and 90s cars so I wouldn't hear the bass in those
The living room though with it's typical sub has exactly what you describe shizzle
i have a cousin who is really into car audio systems and i've sent him songs just so he can hear the bass in some of the generations lol
Soooo. Seems 4/20 is the only upcoming holiday. Is there going to be a v4a make your friend a pot themed song web page anytime soon? 😛
my parents are coming to visit me from out of state on 4/20, so ill be busy all day so thats DEFINITELY when the next update happens for sure
just got 500 free credits
cool
Does anyone know what sort of style prompt i could use on Udio to get music similar to what i've got in suno?
cause that same style gives me trash
If you haven't enabled "manual mode" for the prompts, try that
(otherwise Udio rewrites the prompt)
i had no idea what that switch did
That did help a little
For some reason I did not know there was an off topic chat here.
Goddamn, but the process of making songs on Suno just feels so much better than Udio. Sorry, random fanboy moment. But it is just a more pleasant process. At least partially because I know the rules, I can just pop my tweaks in and it does what I want!
I so feel that too, Suno has much more fine tuned control
This is the 'Udio' chat at the moment, though shills come in and spam whatever, then it quickly evaporates~~~~~~~~~
thanks mods/staff
The Ui is just kinda bad yeah, not knowing at a glance what is playing unless you mouse over each generation
ironically off-topic tends to more chirp-technical than chirp-technical, just how the conversation flows
And the way the page does a like refresh as the generation is occuring
would be cool if Suno incorporated the Track Cover editing
I want lyric modifying
slamming fists on table in fill In Fill IN FILL
cause like, i had one song where it messed up after the first verse, so i continued from before the second verse, including verse 2 and the chorus only, which caused the subsequent combined generation to have the lyrics as:
Verse 1
Verse 2
Chorus
Verse 2
Chorus
which of course only the first Verse, Verse 2, and Chorus were actually in the audio
Or how i've had to type made up words like Malevelon Creek as Muh-level-on to get it to say it correctly, i'd like to be able to correct that so its normal in the lyrics display
all of my lyrics are entirely unintelligible because of how I continue from the middle of clips while there's still more vocals to go
With Suno, I just feel more like I'm participating in the music creation instead of Udio which feels like a slot machine totally. I've joked that Suno was a slot machine in the past but in comparison, I very much know what's about to come out of each Suno gen.
Intersting, i specifically make sure to continue during an instrumental only portion
Thats my thing with Udio yeah its like i prompt something and sometimes its right on, most of the time its just WAY out there
Let's say I have the verse and the first verse and first chorus in my current segment. If I like the first half but not the second, I'll continue from the middle, resulting in my lyrics looking like this:
[verse]
[chorus]
[chorus]
Sometimes the lyrics even change between generations
dance-pop, future bass, uplifting progressive trance, syncopated, transformative theme song. Listen and make your own with Suno.
Udio at this point is something to mess with when I've exhausted my Suno credits. Its taken me 10x generations to make something coherent w Udio than it would with Suno. But the competition is good
Same here yeah
Udio just sounds so much better. It's not even a comparison.
You can't do in udio what you can do in suno, and I'm not even trying right now because of the server load, but I mean... the difference is just staggering
Udio sounds better as far as audio quality wise, like the clearness in the voice and such, but the music itself is worse to me
Agree. But when I come back from Udio, I'm relishing the "coming home" feel LOL
The main difference to me is that udio is more willing to create disasters if you prompt for disasters
Hearing that clarity, the depth between different instruments and things
dude the instruments
with Udio, it's missed when going back to Suno
I'm pretty much exclusively trying to make it do weird things at the moment lol
the instruments is maybe an even bigger difference than the vocals
It really is. Anything with distortion seems to have more 'room' to it to fill out, so it just doesn't become compressed noise
ya I spent a day testing how udio does when I give it completely unreasonable requests... it does not do any better than suno when I ask it to do the impossible, sadly
I guess for me i just dont totally hear all that, im not really musically inclined, i just know when a song makes that thought in my head of "hey i like that" but i couldnt give an in depth analysis
Yes. Unfortunately I can't fairly assess Udio without saying it has a depth and clarity I can't get on Suno. But just the feeling that what I make there wasn't mine diminishes the joy of that better sound (for me)
Im still trying to finish a song after working on it for over a day lol.. on Udio. Because its so good.
80s rock/metal .. Not final but I like the base of it.
absolutely, lost count of how many generations i had to use but I just couldn't let go of the uniquely Neil Young vocals i got so i kept at it
I miss Suno's fast 15 second gen though
Ive used almost 100 I think because I dont wanna give up on this song.
BUT, Suno topped off my credits so I'll be back making stuff on Suno this evening lol
I messed it up because I used a combination of auto generation and manual lyrics. I learned my lesson. Had to restart the whole song.
I have found one (1) genre that suno has and udio doesn't, which is a real shame because suno does it badly and I wanted to hear udio's take
LOL that is a fact. If I could gen as fast there as I can here. I might be able to figure out if there is some method to their madness.
My experience with Suno:
Make 10 generations - "ok i like maybe 6-7 of those but let me see how many other good ones i can get that i might like even better"
My experience with Udio:
Make 10 generations - "I like none of these i'll keep going to see if i'll get one thats good"
I have A LOT of really good unfinished songs on Suno though, dunno when I'll get back to it ... cant get off Udio lol
I tried making that type of song I just posted on Suno for an hour, didnt get what I wanted. Tried it once or twice on Udio, boom. Got that.
multiple gens concurrently w Suno too. I hate having to wait while i try one thing when I could be trying multiple things at once
Mines been the opposite, i have a few of my most recent song on Suno that i liek, trying to pick whati l ike best, but Udio ive done so many of the first verse and nothing
They need a better work flow like Suno... dislike and the song is gone only showing the liked songs. Its a bit messy on Udio right now. I get lost many times on what I was doing.
Opposite for me. Almost every gen on Udio I really like so I cant figure out what one to keep lol. Then I make different versions and well.. Im going insane.
It was easier on Suno to make a full song. I didnt get many good ones so it was easy to pick the best.
Could be style of music im doing maybe but my Udio stuff is just crap to me
They are different
Udio's quality I love though. 320kbps.
really different
That wins it for me.
Doesn't matter how good the segment is though if it doesn't attach to the rest of the song. Those weird 5 second hiccups at the joins!
Suno sounds like 128kbps.
Is that a new thing that's happening 'cause I don't hear any joins at all
ahh yeah Suno kinda better at stitching parts together.
I know exactly where they are too, every 33 seconds
I haven't tried to make a longer "project" song in udio yet, because of the time sink. But the difference in how good udio sounds is too much. It was hard to go back to suno
Also must be a music style thing
I also miss timestamp where I can change any part of the song. That sucks. You are stuck with what you get.
I have not done multi-part today. I forced myself to do a whole song yesterday, and I was having where it just would not join together. There would be 3 seconds of silence or 3 seconds of clashing guitars. And don't get me started on trying to get a consistent chorus
I also miss multiple gens 😦
I'm fascinated by udio's tendency to make singers go off-key. Why does that happen?
To me, the 320 is a win win. I cant go back to Suno because of that.
Too much artifact sound in Suno.. but I will still finish some songs on there. Too good to ignore some of them.
Yep, confirmed udio has no idea what enka is. That's sad.
keep missing it, they are likely going to have to increase the gen time to increase quality
I'm not convinced that it is truly 320kbps. The MP3 you download might be this, but the actual audio quality seems a lot lower than this. I'll do some testing sometime. Suno's audio quality seems to be more to do with the generation containing a lot of artefacts/noise rather than the MP3 encoding.
Still sounds better than Suno though.
This is how I feel too, it's a data thing
But yeah it does sound better
Thats the thing that annoys me with Udio.. its too slow right now.
It does sound better in some ways, but most of the music I've tried doesn't seem as interesting.
It depends on the person I guess. I couldnt get any good results with Suno for 80s hard rock metal trying for an hour. Then with Udio I got it straight away.
🤷♂️
you can't have your cake and eat it too
if you want a better Suno we will have to have higher Gen times
Yeah, could be artificially rendering out at 320kbps, might only be 256kbps, who knows
I'll take that trade in a heartbeat if it means more diverse and real-sounding instruments and voices
but that's not just "quality"
that's a huge advancement
Yeah, I am going to guess the majority of the Noise and Distortion issues are due to limited render time
so even removing those is big, and then getting us to 256kbps with that is bigger
How much have you played with or listened to udio?
I am talking about Suno
but I have listened to plenty of samples
it definitely sounds better for clarity and lack of noise distortions overall
ahahahahaha udio ABSOLUTELY has "touhou" in its training for the same reason suno does
But, Suno could bridge that gap fairly easily IMO just by doubling Gen time to like 3 minutes, increasing to 256kbps, and removing MOST of the noise and distortions
... You don't understand
a lot of what you're hearing isn't audio quality, it's more realistic instruments
and more differentiation between them
yes, it's both
I could never get these results on Suno 🤷♂️
Yes. And once Suno gets more render time, they can deal with those extra variables all the same.
The vocals sound more realistic on Udio
Like, I can't stop listening to that Dune: The Musical, because I can HEAR the fact that it's Broadway acting, I can even hear the acoustics of the stage
It hit, specifically, the sound of five Broadway actors belting at the audience
unmistakable
Like I said, double the render time, your debatably getting double the product, but realistically, probably like 50% better than before
Can also never get guitar solos on Suno like this
And then the even clearer difference is something like dubstep or, yeah, a guitar solo
Yes buddy. We are talking about a future improved Suno
which will come lol
not the current Suno
I agree that instrumentation and vocals and all is great in Udio and better than Suno, already said that.
And I'm saying the difference isn't just the noise and audio issues. There are also some advancements that can't be explained by simply better/broader training.
I guess what we at least have now is two separate tools that seem to have their own strengths and weaknesses at least
We will just have to wait for the next update to see
As much as Udio has clarity, I feel that a lot of the outputs lack character and ambience. Hard to describe. But Suno does pretty well where Udio falls short there IMO
so there are Pros and Cons in my eyes
Keep getting Backend Processing Error in Udio
That song was starting great https://www.udio.com/songs/4aGGboNhxoUvJuet3yiEMG then i was blocked because it's Spongebob 😉 but music was FN amazing
this one sorry Echoes of the Abyss, was the best Spongebob remix i could imagine
after extensive testing, it does worse. It's "willingness" go act creatively imo is a product of overtraining ie it only moves towards a musical structure when it is close enough to its training data, and gets kind of lost if you veer to far from that. This is why all of it's voices, at least to me, sound eerily similar in many cases to famous singer, Eddie Vetter (I think that's his name) being a prime examples I've heard several times now.
it's sound quality is way better
@amber hemlock can I dm you, i got a lot of music data will to share
and this makes it feel "better" ie it changes thr users expectations
so we are sort of primed. It is "shiny", but objectively, it has only 30 seconds of context, continuation can't be picked, and the progressions are almost always less interesting
I have some songs I really like from it
In that case, we should see how udio compares when it comes to the "banjo test"
so when it hits, it really huts
oh that it'll win
or, my preference, the "koto test"
has anyone been able to get in touch with billing@suno.ai? I reached out 7 days ago and followed up 3 days ago and haven't heard back
on acoustic sets and things with that underlying twang, it wins
due to the audio. It just has way more info. But the underlying model is inferior to v3 imo
ah, that's not what I meant
ah, explain
So I decided I'd try promoting individual instruments to see if I could get a really contradictory mix of stuff that doesn't have any genres or cultures in common
guys for suno if i put instruments in the style of music will it only use them instruments? or will it use its own aswell?
ah that's a good test
Mixed results but I found that suno is more willing to put banjo and koto in things than you'd think
this would be expected
based on my results
They're suggestions. It'll add and or ignore what you ask for.
banjo and koto were the winners, marimba and steel drum were the losers
ie udio, although it SEEMS more creative
is less so
the things I've made with suno at times don't really have clear analogs
in fact, most of them you wouldn't hear it and say "oh, it's Ripping off X"
but Udio, it's like, every other song
I hear chord progressions with even the same twang at times
that I recognize
The only thing that i would want from Udio for Suno is the better sound quality, the rest i can work with while the improve the AL
I've only had ONE such occasion with suno, which was a modest mouse progression, but it didn't fit the genre at all so I believe it was pure coincidence
Six gens in udio... five backend failures. It's like opening week for a new multiplayer game
it'll improve over time
however, the good things you get out of it are good
but still, weirdly rip offs
like, that's Ben Gibbard
at times
It's funny because with both of these you know there's just no WAY they were trained legally
I actually believe suno was
if it wasn't, it would have been easier to get to v3
I also think that's why chiptunes
appear a lot
mmmmmm well
A while ago, someone actually managed to reproduce a glitch where their outputs would sound exactly like the song they prompted
and then just built it up on user preference
interesting!
I'd love to know which songs
would give an idea. But i believe msft bought a substantial share in suno
unless that was just a runor
rumor
btw you guys should quick claim your usernames
in Suno
You don't want to be Yolkhead1

I'll go grab it tho
It was Alliecat, they said they got "gimme the beat boys and free my soul" among a few others in the exact correct tune
this does make sense for how an AI could be trained legally
Suno
LOL I had trouble deciding which name I wanted.
niiiiiice
https://www.youtube.com/watch?v=uaQo04X51NM - that took a while to get something "consistent" at that length 😄
🎧📻 📱LISTEN TO MY SONGS EVERYWHERE YOU GO! ALL MY MUSIC HITS ALL MAJOR STREAMING PLATFORMS. SEE LINKS BELOW! 📱📻🎧
This operatic exploration delves into the evolving relationship between humanity and Artificial Intelligence. "The Song of the Machine" confronts the perception of AI as a mere mimic, questioning if true creativity can exist within l...
Suno
Will claiming the name become an issue if it is found that the song produced is distributed under the name of another artist?
I need the Suno multi-band plan.
There are three things that made me say "oh okay so this is actually leaps and bounds ahead"
four if you count the Paul Atreides song
The first is getting a stereo panning effect
The second is hearing that not every damn synth instrument sounds the same in EDM
and the third...
Make your music
Holy. Shit.
yea, it's the quality
no no
when it hits it hits
oh nice!
That is. That is nowhere in suno's data.
Working on that track but site is unstable 😦 https://www.udio.com/songs/phxy7N9a8WvxxfDBRfWEw2
bc their servers must be hammered
I'm with you there, although Sound Quality includes reducing Audio Distortions, Clicking, Hissing, Echoes, etc.
Like I said, opening week for a new multiplayer game
in any case, the issue comes down to: it sounds better in clips, but it's composition is really really not grsat
in comparison
like, we're all biased bc we can recognize a suno song like within 10 seconds
so we can't blind test
Overwatch, Halo Infinite, Endwalker, Helldivers, surely countless more that I'm not thinking of, and you best believe Dawntrail is gonna be a shitshow
but I think most people will put down an udio song, in most cases, bc they get weirdly boring at times
like, I've had a few that are gray
great
but Suno makes some stuff that just grabs you
and doesn't really let go
but part of that is continuation selections
but more broadly, the fact that the other one has one 30 second context
means it has to be dumber
you might be right in that something about suno makes it less likely to stick close to established songs
like, it's gimped
right, and the 30 second context is bound to be a problem
and can't do complex progression. The fact that it holds the voice is reflective of the overtraining
and it's likely their solution to keeping some consistency
but it still diverges
Most of the stuff I hear from Udio doesn't trigger any emotional response. Though some do, just to a lesser degree. Most people were making Meme songs with Suno. So not really a good estimator lol
suno makes music, like, unique music. It's just not amazing audio quality.
Personally, I think a seperate, small refiner model could be interesting as a solution, since the problem affects extreme noises, like heavy bass and screams don't translate Tully
fully
yea same
like, I play it for people because I'm like, this is cool
and nothing
sunos stuff
I had people calling me
I literally have a groupy and I have like 10 fans, my wife thinks it's hilarious
tailing the prod db and watching the suno handle goldrush rn like
I can't speak to this because My GeNs ArEn'T fUcKiNg WoRkInG
Basically, I'll use the other one when I don't have credits, but if I have credit I'll take suno any day
but hey, I'm glad they're there, it is nice
and they aren't a total loss like sd
Yes, the competition is good
so I got ONE successful gen where I tried to prompt for a touhou song
SonAuto has long way to go before I consider them competitive
and I was like, oh damn this is funny, let me get a really good one to show people as a "Hey look! Touhou music is in udio too!"
six failed gens
I've baiscslly had 1 song people liked, but they all just take too long to get started
oh and this one, but that's bc it's short and funny
agreed on the refiner. SUno generating the base track with the Udio refinements would be the goldilocks solution.
My most viewed song on Youtube was the one I did for fun, that I generated on accident (then extended), and that I didn't plan to distribute. --- Now I can't decide if I want to do a full EP of a similar genre and distribute it to my main profile, where it's out of place. Or if I want to create a new artist page and distribute it there where it's new and mainly focuses on no lyric songs/soundscapes or very minimal lyric songs. This is the song for context: https://www.youtube.com/watch?v=cImmoMLt984
Ever wondered what a Cyberpunk Nightclub would sound like? Now you know! Follow me on Spotify: https://open.spotify.com/artist/6w2lZ7GNxR6wiHdg9981Xi?si=cIzUBOcLTBWJAEKnA50hNw
This is mostly just for fun. I also make other music, with lyrics. Subscribe if you like my content. Follow me on Spotify! : https://open.spotify.com/artist/6w2lZ7GNxR6w...
I mean, I only say it bc I assume it would be cheaper, compute wise, similar to how stable cascade has several steps that reduce some of the models complexity. But like, I guarantee they have never ideas than that.
bc they're scientists
and I'm a guy who didn't go to school haha
is it suno or udio?
Suno
Yeah, I mean if it's feasible for Suno to create a "Refiner" capability where you just select the Clip or Full Song, and comb over it to Master it/Remove Audio Distortion/Reduce or Eliminate Echo, that would be dope. It would be better if it actually just took the melody and everything and copied it, but made it better, so regenerates it with improvements.
I will suggest this
Was able to finish my song, i did the same prompting and everything on Suno and it was far from that quality. You have to like that style but here it is https://www.udio.com/songs/hQvGaaWz9a4rGAEHTLd68S Eyes of the Machine (They Watch)
like, the voice is good, quality is good. But like, it is so much less interesting in how it conpoaes
composes
I know this sound p well
and Subo does it too, just less detail
but the quality compositionally is better
paranoid experimental goth trap lofi cassette 4-track homemade amateur, post-creep, autism song. Listen and make your own with Suno.
one example
the buildup
and progression
then numb beat
it's solid
I don't know much about guitar solos or prompting them specifically, but in general I like to throw a bunch of garbage in the lyrics with instrumentals, seems to help. this isn't the same style but a bit closer maybe
https://suno.com/song/ad908653-82d9-40f4-af4a-636d1db3a808
solo, ;;;;;;;classic-rock;;;;;, virtuoso-guitar, hair-rock, guitar-solo ;;;;;;;;;;;;; melodic, instrumental, liverpool song. Listen and make your own with Suno.
but the udio version just lays there
I wanna poke it, but I can't get into the 30 seconds
depends on the type of solo
but usually people mean hendrix
can do plenty of solos
can you use a non download?
believe it or not, this joke one ends in a wicked guitar
ASMR ASMR ASMR ASMR ASMR ASMR ASMR ASMR ASMR whispered tingly song. Listen and make your own with Suno.
I don't think anyone else has gotten that far though
02:54
https://suno.com/song/44811e63-4a22-41fc-9605-7afc31929767
2:40 ish blues guitar solo
acoustic neo-delta blues revival, black 1940s anti-solo soloist song. Listen and make your own with Suno.
with guitar throat
dude literally spits guitar halfway through that one
in any case, if you want a solo, just post the sample where you're up to and we'll slap one on there
these two were ok:
https://suno.com/song/956b87b0-fedf-4cb1-b06e-5934c07c7b52
https://suno.com/song/47b86dda-883c-4921-a9f0-cafed5279758
solo, ;;;;;;;;;;;;;;;;;;;;;;;;, liverpool, guitar, rock, rock, solo, solo ;;;;;;;;;;;;; melodic, instrumental, liverpool song. Listen and make your own with Suno.
technical, guitar instrumental, instrumental, rock, hard rock, instrumental, heavy metal, melodic, energetic, song. Listen and make your own with Suno.
very anthemic, they def fit a lot of the stuff I see
I've never done a full guitar solo actually, ima go put one together
Yeah Its kinda tough ngl. I tried a bunch of different prompts but this is the closest Ive gotten.
https://suno.com/song/2f8092da-a3f3-4f57-8516-0eadc5a317fa
Classic Rock Guitar Solo, going off song. Listen and make your own with Suno.
probably easier if done in smaller chunks, asking for a pure 2:00 guitar solo is a bit atypical so they tend to transition into a normal ish song after awhile
I think you may be right
psychedelimath anti-stratocastuosic guitar prodigy, technistery intimate guitar solo, steel guitar, weepingnotes song. Listen and make your own with Suno.
trying to dial it down lol
Vibes are cool with this one
that is my exact feeling too. The Udio music seems like it is "better" but I have never generated anything I want to listen to again. It all sounds super generic. Suno makes more interesting sounding songs.
I'm so damn impressed
with suno
I've spent like 10 hours over the last few days perfecting an album ahahaha
the lyrics are so on point
yeah the outages make me sad
They know
I've gotten a 1000 free credits from them these past couple weeks, so that's cool
lol got another 500 credits
but lately I'm more on Udio xD maybe not the style but the damn quality is amazing!
You have to imagine part of that is a nice excuse to give people a reason to generate more with Suno than Udio
competition is always good, but I need same quality on Suno that's what's missing now by suno
I thought I was crazy at first but I'm happy to see more people agreeing that Udio makes better sounding music but not necessarily better music. But it's just V1 so I'm sure it will improve as will Suno with V4. The real winner is the consumer.
true, but the time matter now, if Suno will not uptade fast they will just lose
Yeah, they're pretty good at the updates given how long they've been around but I'm sure they're looking to get more out there now
half of our users think we have no idea what we're doing and the other half think we're playing 11-D chess, feel like you guys should really get together in a room and hash it out 😅
That's the classic thing though, right? One moment no idea, the other THEY KNOW ALL
Lots of people have that outlook about lots of things so sure, why not here
You are a bunch of clueless monsters implementing a master plan to rule the world!
maybe becouse there is like no roudmap? We need to guess?
Isn't there that thing about how the enemy is both incredibly incompetent and an existential threat? I think the Suno devs want to get as many people as possible using Suno, that's my theory.
Here is their master plan u ready: "make a thing that they wanna make"
Maybe make some money and some cool career things
Always in the political space yeah
When things go wrong, the enemy is sneaky and smart and in everything giant conspiracy. When things go good, the enemy is a MORON IDIOT CLASSIC ENEMY MOVE
Being bought out for a bazillion dollars probably in the plan somewhere but have fun along the way?
They already have money investors
lol that's the point of every company ect that's sells anything
I don't think anyone is going turn down billions
But not a bazillion dollars!
But I bet their number is higher than we'd all think
oh it's a lot easier to make better sounding music when you are just straight cloning famous artists voices 😂
in this essay I will
eh
why so?
LLMs like this are hydras
cause they are cloning famous artists voice
hi i know you guys are arguing about platform stability and whatnot but can you make it so i can put playlists on my profile
not unlike the GREAT MP3 WARS of uhhh
I don't think Suno brought down the servers on purpose but it never hurts to have a reason to spread some good will, right?
of the 2000s
my 6th generation i had them generate me a gangster rap song and it was just straight up ice cubes vocals with random lyrics
like they don't even hide it
Napster is dead, oh no there are 4 more in its place
xD Voice and? There is no law yet
why do you think the vocals sound so fucking amazing? these are actual artists
Can't copyright style
The laws will shake out how they shake out
But if you don't claim it's by whoever it sounds like
Then as far as the law is written, RIGHT NOW
Sure you can. Style(c) <- See!
it's fine
I am unsure if a voice ''tone'' is copyrighted i mean it's a tone, it'S a pitch.. I dont know
I tested Udio today, 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
This is the result of about 60-90 minutes with Udio, in case folks want to compare quality: https://www.udio.com/songs/ovJy5pbG9jMSGdHSHsoPuA
So what everyone is doing? Udio might be more recognizable with the higher audio quality but I've gotten plenty of Suno generations where it was obvious who it was. That's the nature of training a model and prompting it to a narrow selection of possible tokens
It isn't, but law is a movable, stretchy thing
this, i'm making eurobeat music and suno is even imitating the accents of the singers
nah udio is on a different level with the vocals
never had one generation on Suno that sounded 100% identical
it all actually sounds legit AI generated
what makes me mad by Suno is when i want a violin dubstep i got 9 od 10 times no violin or only at end xD, did try on udio and 9 od 10 got violin, i feel like udio is better in term of keeping with prompt also with lyrics while suno did miss words or make random ones
on Udio it doesn't sound AI generated at all it sounds like a real artist
cause it is lol
it somehow sounds like a real artist while still being void of proper human emotion
I find Suno can be quite emotionally expressive at times - I like how the vocals came out in this one, for example:
Music generated with suno.ai, art by Stable Diffusion, visualizer via MusicVid.org, lyrics written by a human. Public domain. Souls not included.
Lyrics:
Behind the curtain is a man
Who operates a machine
Look closely, there's another curtain
With mysteries yet unseen
Pull back the curtains, face the truth
Behold the wonders of the mind
When ...
Yeah, Suno wins on the emotion factor
The audio sounds like shit, it generates 90% of the songs in a dark deep edm atmosphere, even church songs, with suno it's an innocent toy
But how do you REALLY feel
Udio i mean
yes, many of the Udio songs somehow sound similar
they both have signatures
Suno and Udio, but Udio seems even more uniform
100%. The bridge to the end of this one... Suno blew me away. https://www.youtube.com/watch?v=k_7e8oaRZX4 Udio is very flat in delivery in comparison.
Dim lights, crowded bar, your girlfriend's by your side
But your eyes keep flicking over, and it's something you can't hide
She sees it all, that hungry look, yet no rebuke you hear
The way you watch me move, (and) your eyes linger on my rear
But here's the secret, darling, The truth behind her smile
She's dreaming of me also, just in a differen...
just be like me and spend cash dollars and generate 5000 versions of the same song, you'll find that eventually there will be a masterpiece
Exporing the ~latent space~ with Suno is like that, or any generative AI thing. you are digging through a lot of eh, some garbage, but there is a gem in there
one song typically takes me 1000-2000 creds, unless I get lucky
Fighting bad data more than anything it sometimes feels like
yeah I spent 2000 credits in a day 🥲
this is the way
easy, when I get my monthly 10k it's worth approximately 1.2 albums
but like, come on
that's well worth it
I call it spearfishing, personally, go put yourself in the area you think you want to be, than just stab and see what comes back. Eventually it's something yummy. 🙂
How many parts are you putting together for a 2000 credit song or is it just dozens of generations per part?
how many songs are considered an album?
honestly, the process is almost as much fun as the result - shit's addictive
Lots of analogies like that work, yeah. You are spending time to find the good thing
usually 2-3 parts, just dozens and dozens of generations
This is why I'm realllllly looking forward to infill/inpainting. Since if I find a part in a song I like, I can save that part
For me, it's regenerating stuff a lot, to not only get the sound you want, but also get a clip/song without errors.
Right now I can only save it if I like what came before
I have a playlist full of generations like that
Suno dishes up stuff I didn't even expect / totally off-prompt sometimes, but it's often very tasty still
But then again watch v4 be the same sorta jump from v2 to v3 and I'll just throw out all my v3 generations
The first segment takes like 3x the rest. Sometimes I can do the subsequent ones in 1 or 2. But getting that magical first part...
I usually generate until i get lyrics i like (or hit gold and get lyrics and music both) -> plug the lyrics back in until i get a first part i like -> extend that until i get a consistent sounding part 2 -> generate an instrumental outro until i find one that's consistent
I imagine inpainting will get those generation numbers down a lot, yeah, far fewer good generations will go to waste.
Let's hope so
It happened when we could pick the time to continue from
Udio is better in that regard but it's still not proper inpainting
Suddenly a lot of tracks were viable
I've started making an instrumental first clip, then cut a lyrical one on around 7-15 seconds of that.
and sometimes the problem is the Audio quality or echoes/distortions. So being able to "Refine" a song/clip will be big too
I really like Udio's ability to slap an intro on whatever you are working on at any time
ALSO this
But a 33s intro is too much
Just replace the stuff you don't like
the 33 second rigidness is a thing to work around yeah
Not being able to continue from a specific point seems like a dictation on how their model works
back on the solos lol
How it works must require that 33 second thing
steel guitar{dynamtechnicalt post-anti-math revival, anti-solo mizrahi guitar vocalization}, lofi song. Listen and make your own with Suno.
I've been working on getting them without . .......
for funsies
just bc I wanted to better understand some aspects of prompts in this regard
it's less of a solo as much as idk
guitar singing
that song give me serious elevator music vibes
elevator music rules
@vale lake
Absolutely, i missed cutting after 10 seconds, instrumental beat, testing udio. Suno is way better here
yaya
reminds me a lot of this song that they played on the trolley between the hotels at circus circus in Reno
the start of it
sophisticated, emotional, soulful, sophisticated, complex, lush, Fairlight CMI, carpenter, best quality, minimoog, scifi song. Listen and make your own with Suno.
you talking about Maniac?
i haven't managed to test much, since most times i get errors, what i did manage it interesting although suno to me sounds better. the sub-bass makes it feel warmer, richer. although seems based on others that there's things that udio does improve on.
ya, for sure
elevator music for sure lol
expressive guitar riffoloist wahwah-shredding, crying in the strings of the steel amp song. Listen and make your own with Suno.
I said it before but Uido is absolutely unable to do a live musical performance
i feel like i'm playing a casual puzzle game
She's a maniac, maaaaniiiaac!
so I finally got it lol
that last one is a solo
in the classic sense
with solid progression
and no prompts
Oh yeah? That's a bit surprising, considering how realistic that Dune: The Musical sounds
I swear I can hear the acoustics of the theater
It can do live, but when you ask it for music it falls apart
and I can hear that some of the actors are farther from the pickup mics than the others
ao now I'm going to not forget
to save the damn prompt
bc I've developed like 5 different ways for songs
and I forget every time
I tried a few times. It can do stand up with the theater and crowd and everything
so I don't waste 500 credit on getting a bombastic guitar outro
holy shit
rename the song with the prompt you like to "prompt I like" and save it to a playlist called "prompts"
For those who have never seen this
I listened to this and said "suno killer"
Straight up.
Suno killer.
Maybe I can get it to work by stealing this prompt and prompting in the lyrics
Cursed
You hear it too, right? The actors sound kind of distant
the sound stage yeah
It sounds like it's in a theater
Yeah you can hear the fact that they're in a theater!
I'll take that compliment thanks :))))
It shows sparks of something special, you just need to find where it shines but it seems better at these edge cases with sound effects and stuff than just making solid rock or pop with solid hooks and progressions.
So what kind of live music sound weren't you able to get?
Something like this
https://suno.com/song/b67d8cc7-be49-4b74-b630-55f0e0c53c30
intense,hair metall,80s,live performance,live audience sing,audience singing,hardcore,electric guitar song. Listen and make your own with Suno.
I haven't done much with udio but I can imagine myself prompting for "live recording" and "cheering crowd"
It's probably workable but I couldn't find an edge into it with my Suno tricks
Oh yeah, this is definitely trained on live recordings
this sounds fantastic by the way
"Shrek: the soviet school musical"
Not super soviet and we get a bit of Sinatra up front but it kicks in eventually https://www.udio.com/songs/6mwqmQieQNUdXfMjXEQfb7
Thanks! It's a limiting thing keeping it live but it works pretty well. @sly meteor was the one I first saw that figured out the live trick
Plus when it's 'live' you forgive a lot of the low quality
sneaky sneaky
so "live performance, audience singing" didn't work in udio huh?
I'm going to just wholesale copy the theater prompts and see if I can get music in it
that's almost the exact same thing I would've tried, "live recording, cheering crowd"
I don't see a single thing in the Dune song that suggests theater acoustics. udio must've just done it because that's what musical theater sounds like
... "musical theater" might be the move
Maybe its like Suno and the lyrics shift things
maybe "original cast recording?"
yeah, look at those prompts, "musical theater" is the only one that would remotely suggest that acoustic quality
So maybe you could try and prompt a rock 'n roll musical?
https://www.udio.com/songs/pVQ7X8H43BDbAtAXmEmrog
Tales of green, a world unseen, in a land
Of a comrade's bold ideal
Ogres roar and yet yearn for
The warmth of a real domain
Mud was fine, in this plot line
Now it's concrete gray, we strive```
man made the perfect outro and it disappeared off suno before I could merge it 
THose are lyrics
It's not bad, but it's wrong
The other one:
https://www.udio.com/songs/eaizMcTmrsP37SDWoXVa8w
This one sounds pretty slick
Going to try manual mode
Kinda like that one also, https://www.udio.com/songs/jfd7FtU9KhdqvuRpUDG81T i feel that Udio will give me more pleasure playing with
ah ok, that's the trick
"manual mode"
Yeah, that's it. Not flipping on Manual Mode sabotages live recordings
Well there you go
Yeah it's not what you asked for but this is neat
What?
wow udio absolutely failed hard on making eurobeat, back to suno
Which means Udio will mess with your prompts
But if you flip that switch it will do it 'raw'
Right, right. It sounded like you said the opposite
Oh no, I meant NOT doing manual mode is what breaks the live sound
udio is in its equivalent of "simple mode" by default, you have to tick "manual mode" or else it'll fill your prompts with guesses at what other words might be relevant
It probably helps people's weak-ass terrible prompts 
suno also sucks bad at eurobeat so let's see if I can get some in udio
ooh check this out
https://www.udio.com/songs/vZn8CseSFAjgFCxqnvRmt8
It doesn't have a crowd but I got the "stage" sound
and relatively "unplugged" vocals too
Yeah this sounds live but without a crowd
It's like I'm actually watching my drunk uncle's band perform live
I don't know eurobeat but ???
ok udio literally just copied dave rodgers voice
Hey what do you think which is better udio or suno?
that's eurobeat all right
Note that I let Udio make the whole generation, a single word prompt: "eurobeat"
still neon
Flip Manual Mode on
udio would be leaps and bounds better if we could transfer all our skills from using v3's continue-from to make proper songs in suno, and if we could rapid-crank generations for the sake of experimentation
Udio only creates music from the 1960s, I didn't find anything modern there
but neither of those things are true or even close to it, so for now suno is better for actually making stuff instead of pressing the make music button for fun
Suno is better, it's a better product, it has more thought put into the UI, the generations, the angles for organization. Udio has a clear lead in raw audio quality and seemingly depth of what it is capable of. 'Manual Mode' seems to be the real trick to get what you ask for
I want Suno but with whatever Udio's model is, but with Suno's 2 min generations
Just drop it right in :))))
thing is, "depth of what it is capable of" goes pretty damned deep
yeah
Yup
Deep but limited in that some genres are just not good but then random others are great. I do love it's musicals.
the other day, some a couple of guys were saying "suno can't do EDM at all" and I was like what are you talking about I add bass drops and trance vibes and house oonts oonts to stuff all the time
but they were talking about how suno seemingly only has one "synth" instrument for all of EDM to shre
What genres are you not having luck with in Udio?
So is the max time for Udio 1:38?
Indeed
Now I'm kinda wondering if 'Manual Mode' is the reason