#🗣️┃chill-chat
1 messages · Page 14 of 1
I have yet to find any typical music genres where Udio can make songs I want to listen to
the most obvious example is that suno will never do a dubstep WUB WUB, or even really any decent rich deep bass
Ok.. several extensions have only gotten to 1:38.
It's a songwriting issue not a technical one, thoguh I guess the technology is doing the songwriting too
https://www.udio.com/songs/2oYwbsgoDKARLkquKf1beK
Dang, its what I had in my head for 'synthwave country'
I also tried udio i will rate it 2 out of 10 it feels like making poems with random tones
Anyone having genre issues just flip that Manual mode switch
functionally, having manual mode turned off is the same as suno's simple mode
But being able to tell it lyrics outside of that, which is kind of nice
yeah actually that is nice
I do like that bit more subtle control of Uido
eurobeat electronic high-energy song. Listen and make your own with Suno.
got it to write a eurobeat song about how software development books are too long and i don't want to read them
the chorus at :47 tho
so good
I love how when you tell it to write eurobeat no matter the topic it puts references to driving
gets the genre too well
I already know I'm going to be disappointed but let's see how udio does when I tell it things like "electro-funk, country, guitar chops"
sigh
Man I was all set to have my opinion on Udio as being a cute distraction and a nice toy but Suno is all cool, but the way I'm getting Udio to give me what i have my my head when I throw genre mixes at it, as long as Manual mode is on is righteous
Hey can you guide me sometimes suno generates songs with US accent even if the lyrics are in my own language
lmao right? i got it to write songs based on japanese prompts but it has the worst american accent 😂
😂
Like I said. If we could transfer what we know about how to use continue-from to build stuff, and if we could more easily run experiments
No contest.
Suno can and will get better though, I have no doubt they'll get there
I just don't want to wait
classic first world problem
But that 33 second limit might indicate a technical limitation that makes it so they can't just change one line and let us have that v3 feature
Did this work out for you?
I'm just slamming stuff into Udio and letting it to everything and it seems to be doing just fine
I got one weird garbage and one really weird thing. I have a feeling both of these got ruined by "glitchcore"
I haven't used it yet but 'Remix' in Udio means if you have a sound you like you should be able to remix it and change the words
both really weird
those uh... those weren't good prompts
at least gens are faster than they were yesterday
yeah wow udio reacts way more violently when you're "fighting" it
sound like shit bro 😄
There seems to be some disagreement between the backend and the UI in Udio, it was saying I could generate 4 tracks at once and now it's saying 7 but when I try to generate more than 2, it says the servers can't take it.
udio is much more willing to make unlistenable insanity when you prompt for unlistenable insanity
all almost sound the same , if not jazz and soul from 70's can making only stange noises xd
It goes a little hard https://www.udio.com/songs/5MEzRGP5dzsUKRCqcJJ1gt
That is wrong: Start with instrumental and prompt your Synthies, you can specify all, Even keys. Than cut After some seconds and continue with your lyrics
give me an example of prompting interesting synth instruments?
try to make anything like that, second shoot on suno
forget about anything modern on udio
nice bass line
Wanna put money that I'll be able to make 10s pop?
Or 00s alt-rock?
Or dubstep?
i am bored of dubstep and alt rock
you said "anything modern"
people looking for modern sounds, flexible using
yes anything more than jazz, sould, hip hop from 80's
i made on youtube over 700k views on beat from suno
And I said I guarantee udio will be all too willing to do "radio hits from the 00s and 10s"
recall that you couldn't even make suno v2 STOP making radio hits from the 00s and 10s
and you think udio doesn't have "modern" sounds?
man whatever
i do also over 9 years beatmaking I have lot of tracks when I didn't use AI and over 1-3 milions views under beats and under mixed tracks
cognitive bias is a hell of a drug
right now i made many tracks on SUNO just for experimental tests , 3/10 sound like hits, even vocal is so amazing
Seems like I'm getting a lot of backend processing errors on Udio at the moment
no idea what's the differences between algorithms suno and udio
for sure udio is soo far away from suno technology
forgot about sound like it 😄
multiple different ways of prompting this have all ended up with complete disaster
not failures like when suno just makes something nice and ignores all your prompts. Disasters.
It's kind of entertaining tbh
But anyway, let's make some electro-funk
about time I start trying things in udio that DO work instead of setting it up to fail
Ran the Dune Udio prompt in Suno, not too bad.
lazy prompt: musical, musical, musical, musical, musical, musical, female, female, theatrical, stage, stage (stage might do nothing never tested that one actually)
https://suno.com/song/d946d6ea-9f27-4590-82dc-f3c7017813a1
tried to spice up prompt, not really better: Show tunes happy Soundtrack, >;;, theatrical musical showtune theater live happy female, oatmeal-stage, live
https://suno.com/song/ee4deb58-7cde-40a6-9f23-b592245bfe75
musical, musical, musical, musical, musical, musical, female, female, theatrical, stage, stage song. Listen and make your own with Suno.
Show tunes happy Soundtrack, >;;, theatrical musical showtune theater live happy female, oatmeal-stage, live song. Listen and make your own with Suno.
If you know what to use in a few steps you will get it on Suno, you can control the pitch, vocal tempo, singing, speaking, whispering, key, key, bpm many things if you know music, the only problem is when the servers are abused it manifests itself not only slow generation or problems but also reduces processing power and prompt results are imprecise
even when I am just being lazy with one or two words, i've fallen into a habit of making the style prompt longer by just copying and pasting them a few times. just a vibe that it helps overall
Seems like Udio just completely ignores the genre when I prompt it in Russian. It gives me Russian but the gnre is random
hard to tell but I think that actually has some of the "stage" acoustics too. We just never noticed it because we tend not to make broadway musicals in here haha
artifacts and noise sound like from a gramophone record, 90% of the vocals from Uido have noise, they are radio friendly
so who changed my name 🙂 on suno library:)
initial results are not promising
okay couldn't resist throwing even more random garbage into styel prompt, kind of interesting output lol
https://suno.com/song/56fa94f8-406c-411f-b761-3c3770c0b2d4
musical??>>happy, Soundtrack, theatrical, musical, , theater, live, happy, ;>SDF>SDF>WEF>F.SDf?>>>>>>> song. Listen and make your own with Suno.
they installed a system of profiles and their names, so you took the default name, you can change it in Edit Profile
thankyou, for a moment i fought i was botted:)
Even MORE garbage maybe it is helping. Who doesn't love the wonder genre we call >SDF>WEF>F.SDf?>>happy, Soundtrack, theatrical, musical, , theater, live, happy, ;>SDF>SDF>WEF>F.SDf?>>>>>>>?
SDF>WEF>F.SDf?>>happy, Soundtrack, theatrical, musical, , theater, live, happy, ;>SDF>SDF>WEF>F.SDf?>>>>>>> song. Listen and make your own with Suno.
I am starting to think that suno was trained on Hideki Naganuma and udio was not
because udio's definition of "electro-funk" is not the same
please give me a like im a smidge off trending now!
very early impression is udio likes to swing between extremes: extremely-normal-standard-song <-> is-that-supposed-to-be-music? seems hard to hit the spot in between
that is the impression I'm getting too. If I ask suno to do something unusual I get one or two of what I want, a lot of ignoring my prompts, and a handful of unlistenable shitshows. udio just gives 100% shitshows
I think one way to cheat your way into the sweet spot in the middle is remix, but since you can only remix 0:33 clips, hard to use for actual songs until that hcanges
Wassup, guys. Sorry if it was asked earlier, can't find. Is it possible to reuse generate music style ?
But some things don't change.
Udio is way more neon that Suno
I used to think 'synthwave' was the trigger for Neon but no
Just...it does it
inclined to agree
so much neon
man what were the prompts for all that cool Naganuma-like stuff I got from suno? I guess I could try the exact same
I made an entire album on suno about being a software developer
I made an album cover cause I was bored 
god i'm so entertained
a song about going to coding bootcamp
namecheap?
yes, but on a really tight budget and the godaddy 0.99 cents code seems not working anymore
no no namecheap is the name of a registrar
of course some domains you wont find cheap regardless of where you look
i've had domains in the past from namecheap by the way. just hustling for a pretty low deal
alright, multiple failures at prompting stuff like metal with a violin and EDM with a flute
I could try doing it with a continue, but that's no good when you can only continue at the designated 30 second mark, right?
Fool that I am for thinking that udio would be as good or better at this kind of thing just because it sounds so much better.
I really can't prompt for much of anything that isn't comfortably in a genre.
Spoke too soon, I got metal with strings
definitely not as smooth, but better than expected
my take is that udio is kind of rigid in it's output though the audio quality seems a bit over suno but suno is really flexible with the prompts and output
the thing is, udio isn't just "higher quality," it does instrument voices better on a deep level
yep also tried metal with violin no real success so far
I tried three different things, the last one worked. No telling which version of "violin" in those prompts did it
Why aren't you chatting up the Udio Discord?
because it's all people buzzing about their new toy, and I want to talk about differences and limitations and experimentation
wrong reply
i thought here was off-topic disc
bro, was not targetting you
It is, but @runic depot is the biggest Udio shill in here
it's new, it's interesting, I'm playing with it, therefore I'm talking about it. Block me if it annoys you, won't break my heart.
Yeah, I'm asking why not center in the Udio Discord
you don't have to bother yourself, it's off topic here, and it's suno competitor, so we can discuss about it here. just that it's not blatant promo
so the reason I'm interested in this is because it's exactly what we do all the time in suno, right
in suno I put violin in (almost) every song
yeah, the other day I had a lot of fun discovering the fact that suno is the same way with koto
i have tried udio, but i can't leave Suno for best reasons known to me
so why would you be hurt about it?
but of course there are things that suno REFUSES to put in anything except its designated genres, and for udio that seems to be basically everything
I'm not telling anyone what to do, it just seems that if you love Udio so much you should talk in their Discord.
metal + violin is too easy tbh
No one is, I have been in this chat for the past few days since Udio came out so I have seen most everything people are saying
just chill bro. breathe in and out
koto + whatever is a better test, because suno can do that even though it's unusual
There is a lot of very interesting discussion on the nuances and differences in Suno and Udio and it's perfectly fine to talk here
I never said it wasn't lol
Yes, it is
It was purely a functional question lol
Just waiting for improved audio quality in V4 coz Suno is the most flexible right now, udio is still rigid and has bunch of hallucinations also not friendly for custom lyrics imo
Cool just wanted to make it clear, I'm learning a lot myself about both. Off topic is apparently where it's at
If I was in here talking about Midjourney 24/7, someone might ask me why I'm not in their Discord.
You understand the logic @grim idol ?
lmao
my repeated failure at prompting unusual stuff that worked in suno is a definite bummer, because I would've been eager to jump ship.
I guess I'll play around with continuing in udio next
nice title
Love the name
did you try the remix function yet?
by accident, yes. What does it do?
Aside from copy over the prompts and such just like suno does
okay I actually love this
You can actually change a bit of words. Correct little mistakes and it stays close to original gimme a sec
whoaaa
The more interesting use case is to remix but change your prompt and tags. That is the weird cheat code
Make your music
Sadly only works on first gen currently not with continued
All right... I'll play around with it
the only thing i feel working for udio is the audio quality
and its lacking on afrobeats which Suno wins hands down
it's way deeper than just the audio quality. That's why I'm scraping for ways to make udio do the things I like to do in suno. It's starting to look like that's not gonna happen
alright, now I'm invested in Suno
Regarding Udio: There is no clear information rgeardiing copyright and if songs belong to you after downloading - like on Suno.
yes, once you get deeper, you'll see it lacks behind suno
Yes there is
It does, but you have to give credit
Where?
I mean that the reasons it "sounds better" are deeper than the audio quality
much deeper
The ToS - 6.3 and 6.4 @vestal frigate
Giving credit is not like Suno. I can't use music, where people have to give credit to Suno, Udio or anything else. So you actually can't compare Udio with Suno regarding copyright. It is useless.
I really like the GUI of Udio.
I hate it. Mobile site is awful
On Suno you own the music with a premium account - thats why I can offer it on my website filmmusic.io - with Udio this is actually not possible.
I also don't want to advertise to the world I'm using AI
I will admit it if questioned
suno's mobile site is alright. udio's is torture
Every song made with AI is clearly marked with "AI" on my website. All my music is marked with "HANDCRAFTED".
Well, then what's the problem? lol
People want to know if music is ai or not. I think for transparency this is important.
Then credit Udio and be done with it
People download music and can buy licenses if they do not want to credit me or anybody else.
post-anti-instrumentalized anti-hyper-gothllo trapelloized jewish nigun40404040404040, accidental song. Listen and make your own with Suno.
Not everbody can credit. Game developer cant.
thought this chat night like the weird shit
Not sure I am understanding what you are saying. If a Game Dev uses your song, they can put it in their credits and also mention Udio
off-topic has become the suno-Unio warzone
Game developer makers a game -> People make a recording of the game with music -> Credits not possible
Unio lol
love the layout of your site, seems you came at the right time coz i was looking for a similar build
The layout is a them... you can get it for 20$
show me how pls
Okay, it is 29$ ---> https://themeforest.net/item/valex-bootstrap-admin-dashboard-html-template/26645744
The copyright that Udio gives out is Limited, not Unlimited. But I don't understand all the Ins and Outs
Until they allow a paid plan with No Credit, I would not use their songs to give to Game Devs
or similar likes
Thank You
I can only recommend the Valex theme if you are developer, love bootstrap and a well documented theme 🙂
It took me 4 hours to adopt it to my site.
Reading the ToS will help you a lot @vestal frigate
There are also Laravel, React and many more versions - look at that.
I know the TOS, they do not allow commercial use without credit actually.
You just came in here saying otherwise?
Did you lose track of who you were talking to?
lol
Alright, nvm
??? I never wrote something else.
Did anyone else test the combination of
hyper binaural [normal style tag combo] ultra asmr not comma seperated
Yet? It definitely does 'stuff'
caught in 4K
A clear information is not protected behind 100000000000 sentences of TOS.
Suno has a CLEAR and TRANSPARENT information.
Read, understand, then write.
let me help you out
I haven't tried ultra asmr
6.3 "Subject to your compliance with these Terms, including, without limitation, Section 6.4, you may use your Output for both personal and commercial purposes, and we permit you to download a copy of your Output that consists of an audio file for such purposes."
6.4 "If you exploit or otherwise make any public use of any Output, or any content that includes any Output (in whole or in part), including the distribution of Output in the form of or as part of audio or audio-visual recordings made available via any third party application, website or platform, then you agree to include a credit, notice or other indicator in connection with such use (e.g., in the track title, in the credits section, etc.) that prominently indicates that such Output was generated using the Services. Such notice will be in a form and manner that is reasonable and practicable in light of the usage concerned. "
THIS IS CLEAR: https://suno-ai.notion.site/FAQs-b72601b96de44e5cacd2cd6baa985448#f0f1441180484d6094206e84e334ba36
Hello! Please email billing@suno.com if you are unable to find the answer to your question on this page.
Read above my dude @vestal frigate
Who owns the songs I generate using Suno?
If you are a paying subscriber to Suno, then you own the songs you generate while subscribed to Pro or Premier, subject to your compliance with Suno’s Terms of Service.
Dammit.
Udio you clearly have SO much money and talent behind you, the stuff you output has so much potential, so why can't you make anything half as cool as this
electro funk, funky, koto samples, record scratching song. Listen and make your own with Suno.
Udio requires attribution
I will not apend my time discussing here anymore. bye.
bro I know
It would surely be easier to understand the terms of service in the form of a song. Don't you agree?
I am not sure if it's placebo but it does 'stuff' but it's one of these combos that seem to range from actual noise generator to interesting stuff
somebody please
@vestal frigate
Udio ToS
6.3
Subject to your compliance with these Terms, including, without limitation, Section 6.4, you may use your Output for both personal and commercial purposes, and we permit you to download a copy of your Output that consists of an audio file for such purposes.
6.4
If you exploit or otherwise make any public use of any Output, or any content that includes any Output (in whole or in part), including the distribution of Output in the form of or as part of audio or audio-visual recordings made available via any third party application, website or platform, then you agree to include a credit, notice or other indicator in connection with such use (e.g., in the track title, in the credits section, etc.) that prominently indicates that such Output was generated using the Services. Such notice will be in a form and manner that is reasonable and practicable in light of the usage concerned.
@vale lake combination with chipped orchestra and polyphonic polyrythm opera was hilarious.. careful it gets loud
#🗣️┃chill-chat message
Apparently in some people's minds "you have to scrutinize the ToS for the fine print" = "the information is obfuscated behind 1000 lines of legalese"
If we're talking OWNERSHIP and COMMERCIAL USE then your ass had better be reading the fine print
gotcha, my bad lmao
udio understands me better than suno
a funk song about how this guy on the internet wants me to read the terms of service if i want to use the output of the website for commercial use, but it's too much work and i don't want to
why does it not give the audio preview sometimes?
If someone posts before gen is complete that happens
An then because of discod caching you have to add / to url or it wont update
Gotcha gotcha, makes sense
Well how the heck am I supposed to make anything fun in udio when it's way more of a segregationist than suno ever was
can I do a "style switch demo?" I have a feeling that'll be a disappointment too
I wasn't even able to impress myself and others by making foreign music
And enka is NOT OBSCURE
well it only has 33 seconds of memory it will switch on its own 😄
It was huge in, like, the 70s or something
Yeah, enka is postwar shit, 60s-70s
come onnnnn
Shit dammit I keep forgetting to change it to instrumental
i'm trying to make an enka song and it keeps making jpop
90% of the time when I give it lyrics in japanese it makes jpop no matter what genre i specify
what data did they use to train the model for japanese seriously 
yeah, you are 1000% wasting your time on enka
OH YOU MEANT SUNO
ohhh
yep I've gotten enka from suno
It's an enka song about people arguing on the internet over whether suno or udio is better
but udio quite assurely has no enka in its data at all
which is incredibly frustrating
imagine udio outputting enka high notes
this is so good
I love the name
why are the lyrics like that
admittedly i used a different website to do the lyrics lmao
lol
it was easier than trying to generate a song over and over to get enka
I couldn't get suno to write me an enka song when I wrote t he prompt in japanese
so I used lyrics from another website and then specified the genre as enka in english
yeah no I prompt enka in english and it works... maybe 30% of the time
... wait...
the lyrics being in japanese changes the genre that strongly?
send me those lyrics that made that happen, I'm curious
yeah i swear almost every time I do the prompt in japanese it writes jpop
maybe i'm just writing bad prompts
well what I'm gonna do is prompt some extremely not jpop genres and see if I hear jpop influence
well, it gives up on japanese about 50% of the time
send me the lyrics that did this
oh, when you write the PROMPT in japanese
that... is... interesting
well
I used the same lyrics as the song I posted, but put the genre name in japanese instead of english
when I used enka in english it worked
but when I wrote "60s enka" in japanese it gave me jpop
perhaps the AI that reads the prompts doesn't know a word of japanese and says "moonspeak again, give this weeb some anime shit I guess"
heyoooo
sucks that my go-to example of enka singing is now one of the most hated vtubers ever
because she's also quite possibly the most powerful singer in the vtuber sphere
man vtuber and enka are two things i never thought i'd hear together
Enna sings Amagi-goe (天城越え), originally by Sayuri Ishikawa.
Original source: 10/16/21 First Karaoke https://www.youtube.com/watch?v=ss7Y0kXlORM&t=2568s
Watch Enna’s streams here: https://www.youtube.com/channel/UCR6qhsLpn62WVxCBK1dkLow
skip around if you gotta
you'll see
don't ask me what the song is about I have no idea
Can someone help me figure out what the pre-chorus on this song reminds you of?
https://suno.com/song/2fd15b91-f7af-46ed-9c44-b523ac1dee21
steampunk, mechanical orchestra, deep male voice, synth, slow, epic song. Listen and make your own with Suno.
It's at the tip of my tongue
Sounds super familiar
It's from a popular song at least
I know I've heard it on the radio lol
Probably a few years ago
I'm no musician, I can't articulate it, but that just sounds like a sequence of notes that you'd hear in pop
in fact that first line definitely appears exactly as-is in... all kinds of stuff... dammit, I can't name it either
I'm thinking female 10s pop singer
katy perry, sara bareilles, and friends
lol no trouble, I'll most likely figure it out one day, thank you for trying
god dammit this is gonna bother me
I know some of you said Landr i didnt like what they had so does any one else know any other or use any other distributor site for there music?
Hello
[Continue]
[Verse]
Start from here
Do you mean that I should add a [Continue] before an insertion part?
Would be amazing if someone knowledgeable would compile all the [...] commands with explanations of what they do and how they work 🤔
there is no actual system for [commands]
they only work because the AI sees them and, you hope, correlates [guitar solo] with a guitar solo
it's very consistent with some ([chorus] mostly) and others are almost certainly placebo (like [drop] no matter what others may tell you)
I see... AI is a cat 🐱
yes
Kinda sad realization considering you have limited tries.
making a full song takes a bare minimum of 300-400 credits
and if you're writing original lyrics and retrying until the AI does something with them that sounds right, whoa boy.
Check out the first 4 songs of my original album: “symphony of synergy” https://m.youtube.com/playlist?list=PLODVHnh6zmA5qhEx-dNRoFhq3qjM6vbq8
Prompted Udio for "Kanye West Singing About Goku Defeating the Illuminati" --- Let's see how this turns out
pretty bad lol
https://suno.com/@twig
Hope you all like what I've put together so far!
idk
Did Udio just crash
you know I think half my problems with udio may be caused by having no foreign music in it whatsoever
I tried to didgeri the doo with udio but it was just as uncooperative as suno
is this yours ? pretty epic
Yes and thanks. But note that the vocal changes where unprompted that was just udio forgetting the voice.
oh. yeah, that's going to be an issue. i think everywhere. I just trying "taylor swift singing about the bayou in 'Taylor Swift River' " over on an AI that in theory lets you do that, but it gave me three gens, all of which are dudes
hm yeah i tried yesterday to nudge it to metal from what is was doing it it was like "more of epic opera sure"
Anyone beat maker here?
Maybe.. theres some difference in prompting for changes or general "secrets" we don't know yet I did not gen too much with udio yet.. made my obligatory lorem ipsum test. tried remix. Servers crashed and that was it basically some 20 more test gens maybe..
right now I am trying out extending into a new song
it can work i think the biggest issue is no custom continuation point.
two more gens, "female vocalist" is a dude on udio
not being able to choose the continuation point and the 30 seconds of "memory" are basically a death sentence
but...
knowing that AI music can sound like this is killing me
4/4
i cannot seem to control anything on udio
buch of tiny little musicians trapped and forced to do music. explains the gen time on udio as well..
You know what this is, this is suno from The Bad Future
the timeline where AI music goes in the iphone kind of direction, and the developers' goal is to give the best possible experience to everyone with as few clicks as possible.
For datascientists out there https://paperswithcode.com/dataset/magnatagatune
MagnaTagATune dataset contains 25,863 music clips. Each clip is a 29-seconds-long excerpt belonging to one of the 5223 songs, 445 albums and 230 artists. The clips span a broad range of genres like Classical, New Age, Electronica, Rock, Pop, World, Jazz, Blues, Metal, Punk, and more. Each audio clip is supplied with a vector of binary annotation...
like, do you see those kids in general right now with their "soon we won't need spotify because AI will make whatever music we want to hear on the spot"
same energy as "haha those elitist furry artists are scared"
aight imma make my own musical AI with didgeridoos and inpainting
... yeah
that's what we want
but is it what people with the resources to make these things will want to sell
Sooner or later a decent open source(ish) one will appear. I believe at least
i mean there already are they just don't to lyrics at all. And i don't know the quality
mind you, I'm probably wrong about udio in particular, because they specifically said that making features geared towards musicians is on their list
but you can probably understand what I mean anyway with the iphone metaphor, right?
no idea why this was wordfiltered
This open source project plans to release something before end of 2024 https://github.com/open-mmlab/Amphion/issues/170#event-12387808127
Yes i get it. But I also see the potential in udio They already have that remix thing going for them. I will stay positive and hope they have already a model with larger context window in the works
There is one other notable open source project in the works I've heard about in passing but they don't have a project page yet
always excited to see people carrying the torch of open source
Cool.. I Only saw MusicLM and audioLM so far.
I keep hoping Suno will publish something, anything; they truly have State of the Art and should get publication credit for their work
imb4 it's just bark with more params and 44.1khz stereo output
Lol it's bark, but using bark to annotate the Real (read: proprietary) music dataset
And the using that to train another model
*then
i looked into bark and I was kinda set on trying to train it on music... then i discovered theres no training code implementation
So I wanna post the song I just wrapped up in because it has some really strange "glitches" that mmmmmmmight actually not be glitches?
no, no that's definitely a glitch
I've gotten this kind of thing in suno but extremely rarely
23 seconds and...
2:42
I really should buy headphones for my phone so I can listen to this stuff at night...
I can get you an alternative training method/link if your willpower can reignite on this topic, just let me know
Oh sure I will look into it
https://github.com/theodorblackbird/lina-speech
This was text-to-speech with Apache 2.0 Rwkv 4.0 (linear transformer)
https://github.com/BlinkDL/RWKV-LM for the latest rwkv 6.0 (may not be a simple drop in replacement for the text to speech project there)
Not bad.. best I heard from Suno, I still prefer Udio's guitar solos. They just sound cleaner for some reason 🤷♂️
thanks I'll check it out over the weekend and gauge if it's useful for me
This isn't it, but I did think it kinda maybe sounded a bit like, "How far I'll go" from Moana
Hi
added this suno song to udio, it's a little bit too fast for my liking and some of the words are super quick!.
Is anyone else getting a lot of credits out of nowhere?
they were part of the sorry from suno.for there server crashes.
bubblegum-punk song. Listen and make your own with Suno.
I made this, does someone like this? And can you give me tips? https://www.udio.com/songs/bXgS4JroE3o5EspgHppTdf
Is "udio" the studio without the "st" or the audio without the "a"? 🤔
Neither... udio = you -dio = your audio = your music = your sound ... arrh... sorry that's stupid! lol
udio sounds sooooo ( ...) bad at this stage, i know they will get better. but all my results are background noise instead of songs or tracks
what kind of songs are you making? i.e. what's your prompts?
emotional, sad, piano and its really fked up
also it's important to note that at this stage, if writing custom lyrics, only write 4-6 lines at a time as it generates in ~33 sec segments
share an example?
ok let me try less text maybe
its german xD
w8
sad, piano, emotional, female thats are the prompts, and 2 verse german text
xD
but as you said i will immid...things, try less text
tis you made with udio ? yeah , i can hear the potential...for me is big " BUT..."
udio?
oh sorry i thought we had discussed udio, the new star in AI Music creation, but for me its not
good morning mate whazaaaa ^^
I'm on a rant at the moment 🙂 I wanna rant. About how people shit in a prompt box and think this is art. About the next guy telling me that Bob Ross is a great contemporary painter. He's the sweetest person and a nice teacher but that is not hq stuff it's painting by numbers. We're all used to an industry shitting in our eyes and ears, but common man.
pass the old el crasso ^^
This was Sudo. Making a few acid metal songs,
yeah i feel you Mate
here listen to my song it can definetely keep up with the top charts on spotify but it's still shit 😄
how can all these people even have a life, when their tastebuds and judgement are beyond mad?
to me it feels like it was few months we talked about Chatgpt, now whole Music industrie is going about to blow away
sory my engilish
Honestly, when music AI gets set up enough that you have a lot of control over it, it won't be much difference between it and fruitloops.
german dude i can tell ^^
xD
I'm german too, but it's good to practice imo
yeah lets keep english here
I love AI for doing what it does. It serves everybody the idea that whatever they thought was brilliant music on their spotify is basically pc crafted in 3 minutes
i'm obsessed with suno ! this totaly insane
here's your new little teenage wunderkind performing her great song "I shake my ass"
i try to make a song for my soccer club, this will epic
🤣
ok budy was nice to talk littlebit, need to go to kitchen make some breakfast and take look at my room mates here xD
aight have a nice day
Still gotta work on the lyrics for peanuts
ty cheered me up
https://suno.com/@stitchez got some alright tracks made so far. Been doing my own lyrics so i at least feel involved lol
is that a commercial? people here are already hooked. Post on social media maybe.. Everybody here has music coming out of all ends :o)
Out of curiosity but does anyone here know of the JoJo's Bizzare Adventure anime?
I was trying for some instrumentals that sounded like League of Legends K/DA but ended up with one that sounded like an opening for the JoJo's series. lol
k-pop, pop, dance pop, electronic, hip-hop song. Listen and make your own with Suno.
its a suno link, people can choose to listen or not @west hull they are indeed posted everywhere here so never thought id be harming anyone lol
That is actually pretty damn close to a jojo intro/outro
Just for giggles to make it even worse, look up season two opening and play the two at the same time and it almost fits perfectly. XD
Bloody stream?
you don't Imake jokes bro
wha?
I said, I make jokes. It's a joke, don't be offended please
i see, no problem! I'm not
I could say that every 2 minutes and we all post music me too ofc
no 🙂 no problem mate
exclaimer: this is a commercial, too ^^
When all you wanted was a instrumental song similar in style to K/DA from League of Legends and you get this gem even though you can't understand the lyrics since its SIM speak. lol
k-pop, pop, dance pop, hard rock, hip-hop song. Listen and make your own with Suno.
It was a little heavy tho I see that now. It's when I'm a little fired up I tend to provoke. sry for that 🙂
Its really no problem! I was just confused is all! Sarcasm by text can be pretty hard to read 😋
I think your reading skills aren't that off to be fair ^^
hey can anyone guide me what to write in style of song if i want both male and female to sing together?
been messing around with custom a bit - https://suno.com/song/6f98ad2b-2d09-4201-8082-1d9462191fb2
liquid drum and bass song. Listen and make your own with Suno.
Suno is surprisingly good at making drum and bass
(Bloomberg) -- When Adobe Inc. released its Firefly image-generating software last year, the company said the artificial intelligence model was trained mainly on Adobe Stock, its database of hundreds of millions of licensed images. Firefly, Adobe said, was a “commercially safe” alternative to competitors like Midjourney, which learned by scrapin...
As of today all my suno songs are now live all streaming services! Next to be released as single: https://suno.com/song/f0240d5e-b2d4-4a72-84cc-2424ac65092e
Cute Pop Rock, Pop Country, Happy, Uplifting, Upbeat, Melodious Vocals, Indonesian Female Vocalist song. Listen and make your own with Suno.
Seems Udio can't extend my song that is 4:22 in length any further. Not a processing error, just flat out refuses because it says the song is too long. Thought it'd be interesting to note that in case anyone else tries a long one. It's nice to not have to stitch parts together separately but it's also kind of annoying to have to seek to the part of the song that's new.
in term of creativity and musicality i still prever Suno..
been experimenting with udio with the same lyrics and prompts
I just hope some of the best tools in udio will be copied, such as the ability to add parts before and after a song
well ..that's like 9 or 10 continuations ... the 33 seconds/clip is slightly frustrating. As for creativity and musicality, I disagree. At least Rocksong wise udio seems to be more akin to V3a than V3. 'worse' /rougher sounding but with a highter temp
that would be killer...
also MUCH less choir-heavy
yeah maybe on my genres suno is better, but udio may be better on rock etc
I had to keep continuing because it was making nice things at the same time as refusing to go in the direction I was giving it 😆
Suno excels at electronic stuff (inclusive pop and hiphop). It REALLY sucks at rock since V3
I just cannot be satisfied with what udio gave me after 500 credits for a prompt that would get good result in Suno within 10 or so generations
So I wanted to stick an intro and outro on it and it went "no."
Stuff like that is impossible with Suno
is it just me or is Udio really good with oldies?
For now I will stick with suno. I believe in you Suno!
yeah .. .rock or jazz [which you might define as 'oldies'] are definitely better on udio
but you can hear the "roughness" of the result in udio ... it sounds like a bad radio
also, I still cannot commit on Udio with their copyright policy
Suno's sound is clearer and crisper ... and full of Highschool Musicals surprise choir on faster songs with guitars. which makes rock tracks unlistenable
I know they are still in beta, but still.. having to split royalties even if you subscribed? wtf is that? or has it been updated or been greatly misunderstood?
I've made something that I'm going to use as a test track to confirm whether Udio's MP3 bitrate is a placebo or not. I'm convinced that the audio quality is lower, aside from maybe the vocals. Drums sound obviously watery to me. Will also do the same with Suno.
Udio quality for me is way too good like im finding it hard to use suno again
The differences I hear are in instrument fidelity and expressiveness. Udio renders out very expressive lead guitar, for example. But it does come at a cost - often simple musical ideas are being buried in too much expressiveness and decoration and there doesn't seem to be an easy way to tone it down. Like it's a bit of a diva. Or like Midjourney with stylize set too high - initially the images are just "wow", but eventually you get bored of seeing tiaras and filligree jewellery on every fantasy character.
Its a new world 🙂 https://www.udio.com/songs/c7RdEEBfDYcQ39xiWDV1N6
They are not yet a live service - or at least not in its long-term form. It's a free public beta and unclear where it is heading.
😩
https://suno.com/song/bdf02f4b-7449-4bf0-b129-de5106ece7e1
I was going through old stuff and found an alt version of a song I put on my first album
v2 was still so decent
midwest emo, math pop, slacker rock, chillwave, lo-fi song. Listen and make your own with Suno.
that song became this
Music video made using Stable Diffusion Video 1.0, images generated via Stable Diffusion.
The album (Yolkhead) available on Spotify or basically any other online storefronts.
Lyrics:
A familiar morning
in your coffee and pills
you merge into traffic
the day got away again
like a bird cage inverted
you aren't sure what you did
a muscle remembe...
also v2
Am trying both UDIO and SUNO and see UDIO has nice terms for now considering it's free for now, but still a bit of bugs here and there and lacking sought features by advanced creators who use a DAW... what am paying attention between both are features mainly and the copyright license... very promising and am sure others are gonna jump into the arena with their own offerings...
i wish udio didn't have the weird 32s segment thing
or at least
gave you the option to just have one track
that's longer
but otherwise I much prefer the music it produces to suno
udio's terms are they own your music...
not sure how that is 'nice terms'
from the term of service 6.3 Your Content
In connection with your use of the Services, you may be able to post, upload, or submit content and other information owned and controlled by you (such as your Input Content) on or through the Services (“Your Content”). As between the Company and you, the Company does not claim any ownership in Your Content, including any Output generated by you in response to the submission of your Input Content to the Services, provided that the Company or its affiliates or their respective licensors own and will continue to own the Services and any and all other software or technology that was used to generate any Output. Subject to your compliance with these Terms, including, without limitation, Section 6.4, you may use your Output for both personal and commercial purposes, and we permit you to download a copy of your Output that consists of an audio file for such purposes.
guysss, not question about suno but about music on platforms like spotify, yt - what genres are most liked, most viewed by people and which are most popular?
whoops look like someone deleted there lastest creation:)
any modern rubbish that use repeat beats and drum loops 🙂
eg rap, hip hop ,trash in other words written by talentless street urichings that can't be bothered to learn to write/compose music who running around in street gangs to make them self look cool 🙂
lol, you mean like almost all poets throughout history?
Like, Poe famously died basically strung out on a park bench
drinking lean
i can't put my udio version here has it only a pg 13 server , and they don't know nothing about the birds and the bees yet 🙂
Anyone tried Udios beta yet? I mean I'm not paying pro monthly anymore on Suno I'll tell you that. I'm just so amazed by it so far
I'm using udio
and Its in a lot of ways better xD
@tight nest what do you think?
the important one: quality
in my opinion suno creat always very crips voice, destroyed a lot of times songs
I mean it's got a way to go yet with hoe Suno is presented but if this is just the udio beta then I'm really excited for what its going to bring
How' not hoe 🤦♂️
Made the jump myself over to Udio and while I much prefer the options given to us by Suno I am finding the Quality of Udio is better than Suno on the music I am creating. But having these two great sites I see as a bonus for us all.
Ok I figured everyone saw udio... here's my 2 cents.. holy sh*t.
My two cents was holy fucking shit
pretty insane
It just sounds so much crisper cleaner. God why can these two sites not just combine into one.
thinking for Suno to get there is gona have to be at version 8 or something.. hopefully competition makes suno team get into the requests a little bit more aggressively cose no way they can compete with that sound right now
here is what I am working on atm on Udio.
Haha very true!!!
One of the first gens I did on Udio. Tina Turner?
that sounds bananas.. in a complete dif stile I got this.. still a little dirty but those vocals arrangements are fucking pristine
sorry.. i got the wrong file
I should make a Tina Turner album having two songs sounding like her lol
I like that
My first Udio gen (nerdy shit, I tend to exclusively do stuff relating to my tabletop games). The style isn't what I was going for, but I ended up really like it.
I also tested Symphonic metal and well it gave me this. Suno can do it but not as close yet.
You guys tried flipping the 'Manual Mode' switch on?
That seems to be THE way to get it to follow your style requests more and steer it out of bland pop-like stuff
Yer tried it on a few I like it. Sadly Udio has gone offline.
It just came back up for me :v
Messing with it last night and that's how you can get live performance style out of it
Oh, I have not, I completely forgot about it, even though I meant to use it, I just kinda forgot. I'll have to when gens are back up.
If you don't flip manual mode Udio will refine the live sound out
Did some testing earlier. Weirdly Udio output SOUNDS like it's 192kbps MP3 or lower but a spectogram shows it is what 320kbps is expected to be
so you do just select manual mode and write the same promt you would have in not manual mode?
That's what I've been doing yeah, since Udio will change your prompt it gives to the model if you don't
this is true for Suno as well, in case anyone wasn't doing that
always go manual
neo-bachzarthoven indie post-anti-instrumental, anti-folk song. Listen and make your own with Suno.
I did these two gens today
to show how Suno goes toe to toe when you focus down fidelity
neo-bachzarthoven indie post-anti-instrumental, anti-folk song. Listen and make your own with Suno.
always be chasing clarity is my suno motto
these ones you get the string sounds
etc
like, the whine of hands level of acoustic fidelity
Suno's manual mode isn't as subtle in UI, so it is super easy to forget the tick in Udio in my experience.
yeah
Though admittedly user error, an option to keep it ticked would be nice in the future.
Man it's just so easy to get clarity out of Udio
not so easy to get compelling audio though
I'm not a big audiophile, but I find it plenty to my tastes. Though that is the subjectivity of music.
epic hard rock with electric guitar song. Listen and make your own with Suno.
Hopefully when Suno doubles down on clarity, I can more fairly judge the two. But for now all I can hear going back to Suno is the muddy sound, and not how much more compelling the music is, but I'll take your word for it.
I find that my Udio songs sound better, but that may be user error, Udio may just be easier for the layman to get songs they want, since it has better (in my opinion) lyric generation.
At the very least I go into less work to get something I want, given my inability to write compelling lyrics.
and going just with music I'm having easier success with Udio despite it being weird and janky
Though I will remain a financial supporter of Suno, since I think they have a good business model, and I am hopeful the competition helps them retrn to the top of the list.
I have not had jank in my experience, what parts? Or do you just mean service and the occasional UI hiccup, had plenty of those.
Their team is definitely floundering under the success.
so, I thought the same thing at first. But the more I've used it and listened back on them, the more I see a huge difference. Like, I hear them and I'm like "that sounds good!"
I never find myself compulsively listening to my own udio music though.
like, the suno music is good enough that I get it stuck in my head and want to go listen
but maybe that's just me
it also may very well be the genre
Just done a few generations coming from UDIO for a bit, and Suno is definitely more creative, if they can fix the quality to match UDIO it would be amazing. I find myself head nodding more with the Suno creations, though they definitely use more autotune sounding voices.
Yeah, same, I mostly don't go back to Udio music, like. If I am in the car, the only time I have found myself playing AI music is when my wife asked to hear a joke song I made at one point while driving.
Can't generate Udio tracks fast enough, simply don't have the music to listen to HEH
yea, the suno music though? On Spotify averages 12.111 streams per listener
meaning
they like it
and I find myself going back and even listening to some v2
gens
now and then
Oh I can't do v2 anymore
same, it's inferior imo
but they were better than I remember
I mean even listen to gens I thought were great
like 1 or 2
slam poetry, spoken word song. Listen and make your own with Suno.
I will admit I al pretty closed of in genres that may be bad on Udio. I listen to a lot of stuff (mostly folk-punk and classic rock), but do not listen to much modern pop, and don't like rap. I imagine Udio may be bad at those, given just how it feels when listening to the songs.
classic rock Usio is probably better at, I haven't done much of it on Suno, I've done mostly experimental Goth type stuff and mixed genre
I certainly can't find myself going back to Suno music on easy listens. Again, not much of an audiophile, but the quality hurts me, and it often doesn't make something I like.
you said you like folk punk?
what, like Rasputin?
rasputina
or where does that land
It can make some pretty get folk-rock as well, or midwest emo folk. I think you had a song on Udio with the same lyrics as hollow on your Suno list, it was what made me try Udio actually.
yea, i did
wasn't complete
but I remember that one
idk, i made a pretty good metal song
More like AJJ, or Mountain Goats.
anti-pop-punk lofi cassette, anti-sound anti-fx, anti-recording style, trap song. Listen and make your own with Suno.
I think that might be too poppy tho
I also quite like your Unrefined Echoes cause it sounds a lot like MAl Blum.
for you
classic rock elements in here tho
intricate math dance-pop post-anti-indietronica lofi cassette, midwest emo song. Listen and make your own with Suno.
however, it's mixed genre
and takes advantage of decay
in the model
to create a kind of musical drip
drop
It has a good beat to it, I'm not super against it, but it is a little off my tastes. Mind I don't hate all pop. I just have a lot less that I find to me tastes.
I think my most "disliked" genre is rap, but I still have like, 5 songs that are the exception. Mostly older 90s rap though.
Yeah, this one is nice.
you like older rap?
personally this one is the best I've made on here, but it's slowcore
or closer
to slowcore
but like, nothing comes close to some of those, from udio
And to be fair, it's hard for us to properly tell WHAT people are making
on suno
bc the explore really only showcases oddities
and Russian
whereas Udio has a lot more showcasing
like, there has to be other people on suno
making like, postmodern indie rock
Oh, prolly one of yours that's up there for me. I like this kind of thing as well.
true
or rather "old timer"
guys want a laugh? https://suno.com/song/a90acc78-1482-4579-a0b9-dbe214f88ee3
mexican salsa style rap, fuggin, shidded song. Listen and make your own with Suno.
I made like, one song that isn't just Tabletop game related, or a joke song to mess with my wife, and it was pretty nice. My lyrics for it are ass. I thought my ability to write would translate well into lyrics. It does not. But I did what I could with some ai assistance, and it came out to almost something I'd listen to in the car.
folk-punk, post-anti-indietronica, midwest emo lofi soloist song. Listen and make your own with Suno.
yea, the ballad of gentle orc one
I'd have to agree. I don't really use Udio's homescreen to be fair, but it is a lot better than Suno's at least.
yea, this is what it's for imo, getting music that tells a story YOU like
It really does though, as demostrated by the only other songs I really finished on Udio are of a similar (and nerdy) vein.
There’s over three of us!
I will likely use ai assisted writing once I run out of material
maybe earlier
I don't think it's good enough to do it itself yet
but if you could have it spam basically possibilities
from a given point
and then build the lyrics the same way we do suno songs
j could probably make something better than my own ppems
poems
since that's the actual method I essentially use
when editing
for the last 20 years
except it's easier to appraise when its produced by the machine
I will say, I appreciate that in Udio I can just dump things into the prompt. This is an example. I use one of my personally played character's backstories - compared to the rest which are my player's character - which is a good handful of google doc pages, and just dumped it into Udio, and got this. Was my second song there.
But the ability to do this is nice, for someone not great at lyrics, and pretty shit at AI prompting. It follows the timeline of events pretty well, and only took me a handful on gens to get something I was happy with.
But this is yet again rock, though operatic, something Udio does well, as we established.
yea, you can just add more and more which is nice
Hi comrades!
Is anyone else having trouble with the text?
The text is black and is not visible until you select it with the mouse.
copy text -> press "ctrl + shift + v" and see if it pastes as plaintext
I added it all at once here, which is probably the most impressive part. But I agree adding it in waves would have been better and likely used less gens. But I wanted to see what the program could do.
I think it's a great idea, half of what I do is testing/finding the edges of the program
I do though, appreciate the way extensions work in Udio though.
yea, that's nice
The ability to change the prompt, or add more to it, if you don't want it to use something too early.
I really think general inpainting
would be the best solution
where you can select a section, and it regenerates
After this, the text is inserted, but is still invisible until selected with the mouse 😦
Yeah, that would be a very useful tool.
The amount of times I get a great gen, except that one fucking part.
i got nothin else that comes to mind
Id kill for that, especially in multitrack format.
Thanks anyway 🙂
Another great tool would be style locking, to keep a voice or song motif for another gen.
Refresh and do the ctrl shift paste. Sounds like you pasted some text with color info
Yeah
Or pause inserting, when you want the singer to wait
Can you not do that? I have found a good amount of success.
With stuff like [Pause], [Silence], [Paused Silence].
I am writing text in a text field on a website
Though I have only tested it on Udio.
Oh, I didn't know that was possible. Thank you
Paused silence 😂
Look, you get the ai to listen anyway you can.
Silenced pause 🙂
I get it, its just a terribly illiterate AI 😂
I think we'll get it at some point, I don't see why not. The tech is there.
I wonder if the training data is there or only the studios will be able to make something good, proper multitracks arent very common
And some genres would be heavily over represented
Another thing I think both Suno and Udio would benefit from, but especially Udio: For extensions, they need to be auto-placed into a folder or something. Again, the is especially egregious on Udio. Due to the lack of automatic cover art, and similar names, makes it so you eventually can get quite lost if you end up doing several extensions of the same song.
not sure, but inpainting, minimally, should be possible already
you really just need a controlnet equivalent
for audio
now, it may be there are additional hurdles to that
I don't know about
but it seems possible
Udio/Suno dont use the architecture that supports controlnet type guidance. Only Sonauto does, since it uses diffusion. Udio/Suno are built on transformers IIUC
So im not sure controlnet is possible without them making an architectural breakthrough
yea that and it being very hard to tell what you liked
like, it doesn't update properly
oh interesting, I didn't know that
I mean it's my understanding diffusion is the future
Yeah, I have found this part always updates fine for me, on my creations.
But this part updates like, 2 days later.
Which is exceedingly funny, and frustrating.
Hey guys... question... im guessing you can collaborate making a song using the "Extend from" if I share the Chirp?
Eh i think it's OK, especially for a beta.
well that is your opinion mine is more the merrier competion is always welcome.
Udio sound is so overcompressed I'm starting to think people are confusing loudness with sound quality
I have the same feeling about Suno when it comes to epic music. Udio is just on another level. Suno can generate an interesting epic cinematic track, but you need luck, and even then, Suno tracks sound 'thin' compared to Udio.
I think it really depends on the genre you want. For everything I like to listen to, Udio sounds leagues better, but I have seen examples of genres that Suno prevails in.
Exactly
I'm waiting for Suno v4
It's illegal to upload music with this, right? I'm afraid
For what?
Type of digital stores as if they were for artists
From your experience, how good is Udio with creative genre mixing & recognizing some more obscure genres/instruments? If that’s something you’ve experimented with.
It's all in their TOS, if you are a free user the copyright is Suno's, otherwise it is yours. I recommend reading it for exact details, it is all very concise.
I mean they aren't monetizing anything yet, is it just for creativity or not?
But there are only free users
in tha beta
You didn't specify, I assumed you meant Suno due to this being a Suno Discord.
They also have a TOS though, which explains their copyright.
I've not experimented as such I've kept my stuff fairly grounded within a handful of genres, none on which are bizarre.
I've mostly done operatic rock, folk-punk, and bardic styles so far, as I like to recreate tabletop experiences in music for fun, and to goof around with my players.
Worked a lot on these lyrics, check it out:
https://suno.com/song/f38f7673-fb9a-4f38-a8a0-484e55a18297
syncopated kids music, vocaloid song. Listen and make your own with Suno.
Ah okay thanks. And good choices!
well the suno does better music in some cases and udio in othersm suno can make better beat but struggle like crazy with quality 😦
I'm starting to think it's impossible for AI to recreate deathcore. It's the only genre I literally can't find anything for
Suno is weird with metal in general when it comes to distinguishing between sub genres.
True, from my experience typing anything related to metal just makes it bring up power metal. Which I genuinely hate, but I get some people like it
That and forget trying to make it do blast beats 🤣 thing will implode
Yup that has happened to me a lot too. My best guess is because power metal generally has cleaner vocals which it handles better if you have anything in the lyrics section.
It does Emo Grunge extremely well though
I’ve tried extensively to make “pure” Rock songs for like 3 months now - with no success - Until now - https://www.youtube.com/watch?v=J0rz8RoKDCU (I was going for a fusion beginning… it landed a little too R&B… but that’s my signature…so)
DallE2andMe2 studio production
Is this the first "real" Rock song generated by Ai?
Lyrics By David J Ritchie
Music was Ai generated and sound engineered by David J Ritchie
Like the music? Hire me to get your very own DallE2andMe2 song produced now!
https://www.upwork.com/services/product/marketing-unleashed-musical-vision-studio-production-...
I see what you mean, honestly it's not that bad
https://suno.com/song/1884bdc5-57c2-4751-b284-72bc4fcaab91 here's my example of how great this works with Grunge (lyrics will be confusing if you don't play don't starve lmfao)
Beautiful.
Just when I thought it couldn't get better, now theres two.
I'd make it three but these two are enough trouble as is 
I love the sound of udio but two things are holding it back for me.. I'm trying to make musicals of things like Elden Ring, and suno comes up with great tracks but udio 100% has only speaking voices when I use the keyword musical, no matter what other keywords I use. I can't get it to both sound like a musical and have mostly singing. secondly, not being able to continue from a timestamp kind of ruins it for me since I'm so selective about what parts I keep. I'm sure in a year from now both of these services will be amazing though and have little downsides
just need to go see miyazaki in person and ask
🥲😂
as are yours
I see this is a cat channel now
not if i throw a shlawg in here too
Thanks 
One more for the new cat channel
Bad lighting included
Forgot to turn off the flash and now he has one green ey3
Nobody answered you, but yes. This is completely possible
Heh... Udio doesn't even work in Firefox.
huh I was using it earlier in firefox werid
i kept trying to reload in various ways and what not. no donut. but it works in Edge.
too Tiger Lillies-esque?
https://www.udio.com/songs/npXsiShBR9SKz6itaoak2h
how in the world did you get those vocals!!!!! this is perfection! NAILED IT
sure, but those vocals are so accurate! It's impressive
It scared me when it made it....but it couldnt manage to do wretching!
I love it! My sister used to put Tiger Lillies on in the music store when she wanted people to leave
I've seen them live in London - amazing. They are an acquired taste though
Yeah for sure! I wish I could have seen them live!!!
I'd like to see Gogol Bordello live too... has anyone made anything with a gypsy punk genre?
Udio doesn't know any of the genres i tried... until i went for "Blues". that one it knew at least.
Sorry I missed this
Yes have a look at https://discord.com/channels/1069381916492562582/1221135437230379048
it is almost 25min
wordless comedy:
hm seems like udio knows russian hardbass... if that's a useful information... I don't know
Suno v3 turning voices into a choir half-way through the song is so predictable that I might as well anticipate it in my text: https://suno.com/song/ebc170f0-b04e-4f7d-b86e-67fee54be18b
Griot, Djeli, Mali Highlife song. Listen and make your own with Suno.
Any tips on getting a Halestorm sounding song out of Udio? (i.e. a voice similar to Lizzie Hale) or even on Suno
Dubstep Kawaii Future Bass, glitch fx, VOCALOID song. Listen and make your own with Suno.
its so fucking good
and in first try
Ah when we still post udio in here I'm in. Got to make some lyrics that make sense tho... it's been a while 😄
A song about how I own too many clothes but I still want to go shopping and can't resist a sale
The ability to throw whatever random grievance I have into suno and get a song out of it is incredibly therapeutic
Eurobeat song about how at my new job the morning meetings are at 7:30 and I've never had to wake up this early so it's really hard, but at least no one turns their camera on so I can wear my pajamas
Bam, song
2024 folks
found that if your prompt is lyricsrelated you get text anyway, had New Orleans Carnatic and instrumental, it would always talk or sing
did you get special effects some time? tried a lot but not much. [gunshot] can get close and a few times it showed stuff. For vocal sounds I tried a lot like Wow, yeehaw, bowchickawowow, and I find that wow in rockgenres rules, but make more vocals and it will be ignored. like woooow. Yodel is a total fail, it's just not there yet I guess. Overall I'm totally in love with the vocals, but make it a decent song? work... now I made lyrics for the chat propmt 😄
someone please help how can u make song longer than 2 minutes
How do you create a song longer than 2 minutes r u using paid version
This isn't off topic, buuuuuuuuuut all you do is click "extend"
Listen to this song guys 🤭 https://suno.com/song/b4d9e87e-728d-4ad7-b1ea-eb78236c29e3
90s, 90s R&B, pop, attitude, ego male voice song. Listen and make your own with Suno.
A rave song. I thought it was pretty good. Just wanted to share. https://suno.com/song/702e9fff-bb1d-48fb-978e-4bbe688b2ee0
Hard trance, hardstyle, stutter techno, scene-core song. Listen and make your own with Suno.
*Is it just me, or do the generated songs tend to get views right after I publish them, but then they taper off?
Also, if you could, would you consider giving a like to all my songs on my profile to help increase their visibility? Thanks! https://suno.com/song/b4d9e87e-728d-4ad7-b1ea-eb78236c29e3*
Just liked it, brother.
Hyperpop about loneliness. https://suno.com/song/e844e12a-00df-49af-b76a-5c66e761eff4
Thanks so much, brother:))
Suno all the way up
why is everything on suno neon
ask the Choir 😉
neon?
I'm lost
I'm here to get new tracks suggestions
Any takers?
Yes
Check gala music streaming platform
it'll be worth it
I'm more into the token though
$MUSIC
I'm all in with $MUSIC tokens on All Access! The idea of using tokens to prove my loyalty and potentially get exclusive access is thrilling.
you might wanna have a look guys
the word neon just seems to come up a lot
in case you ever had wanted to listen to a rap song written in japanese about system design theory
Track 6 from Bob Jerkins's upcoming album!
Verse 1:
I was sitting on the toilet, doing my business,
When I saw something strange, I couldn't resist.
I poked it with my finger, just to see what it would do,
It was poop, but I didn't care, I poked it again, just for you.
Chorus:
I poked the poop in the toilet, just to see,
I poked the poop in th...
nah, its just one button here
and then you have this
try a bit playing with this, good understanding to have
Bye Bye Movie Score -Composer 😸 https://suno.com/song/454d2690-76f2-47d8-b8fc-8c9132e72e92 https://suno.com/playlist/899aa4b4-d76d-4a12-8f24-d9176ef480b7
cinematic, gladiator, arabic voice, fx song. Listen and make your own with Suno.
Hi, I made an album dedicated to a D&D game I used to play with some friends so the genre is usually power metal although I also experimented with other genres. In case you like to listen to it and give your opinion.
https://youtube.com/playlist?list=PLSqE9hNGDoliuOLivgKLcQNVgCejHesvZ&si=Ck5vUeHalp3KMupP
Holy moly
I asked it to write me a disco love song in SQL queries and it really did it
the transformation is perfectly timed in that one, love it.
btw just in case you aren't trying it, the simplest choir adaptation I have found is probably the most effective, which is just to avoid [chorus] tags. often does not work but it overall can help and sometimes a lot. Tried it on your prompt and it overall much less choiry-y even when it didn't fully remove it:
https://suno.com/song/8058cda3-ba93-4495-bd66-e54d0bcda36f
(sample size 8, 4 each way: https://suno.com/playlist/6622f7c6-edc0-41d3-af67-27aee7afb022)
Griot, Djeli, Mali Highlife song. Listen and make your own with Suno.
Yeah absolutely. It's pretty chaotic in the sense that even small changes can have huge effects, like even just one second difference could make things generate very different.
I tried "refrain" instead of chorus, but it didn't make a difference. I even burned credits making a song which didn't mention other people, choirs, extra volume, or anything else that might be interpreted by the engine as a "oh!oh!The User wants a choir!! Let me Oblige!" trigger and it didn't make a difference. It's genre, gender, lyrical volume and speed specific though. fast songs with clear stringed instruments with relatively long lyrics (jazz or rock songs) sung by a male voice are impossible to get without choir. Synth heavy song sung by a female voice are okay'ish. Still a lot of reverb and chorus, but it doesn't detract too much as it fits the style and the reverb. slow to mid-tempo Pop or hiphop songs, female voice work relatively fine. Again the occasional chorus effect fits with the genre and the instruments used (also, we're all used to heavy autotune in commercial pop songs anyway)
Yeah it doesn't always work for sure. But it is always worth trying (or just starting your song with [refrain] instead) because sometimes it works quite well.
I have no found a specific style word that helps very much. The only style thing I know helps is 'more unique longer prompts' in a very general ways seems to make it less likely.
yeps. What MIGHT work (which I didn't try yet) is to split the lyrics in short (4 lines) parts and generate only those 4 lines, then extend from that. THe first 4 lines of lyrics are generally super clear (much better voice than V3a to be honest) ... you can be that on the 5th line an extra voice will be added and that it will go full choir by line 8.
I think I tested adding a fake chorus at end (after all your real lyrics) so the song would never get to it, but it didn't help much. Or at least using [refrain] seemed better. But not sure
Swing song. Listen and make your own with Suno.
(wet circuits was telling me that my prompts were too complex)
When you want to do something in photoshop that you figure is 1-2 steps, it's usually 10-15 steps.
I'll test it quick, as I am procrastinating on really boring work...
How dumb does photoshop have to be? When you make a selection in one image, you can't use that in a different image tab.
melodic hard rock ballad song. Listen and make your own with Suno.
@grim idol wow, thanks for the heads-up. Oddly enough you still had a [Chorus] tag left in several places, and yet you got a lot of non-choir successed. I'm definately gonna try using [Refrain] instead of [Chorus] and see if that helps.
Sometimes it makes literally no difference, but sometimes it swings things quite a bit, but I have no idea what is different when it works or doesn't
yeah it's all statistics
Hmm, right now i just get this:
Maybe the server is overloaded. And I wrote a song about that (only the voice is AI): 😄
Oh, i got it unstuck now.
hmm. getting choir/reverb/fuzzy voics still, despite using [Refrain] instead of [Chorus] in these first few tests.
ohh... got a good one
but it sure takes more tries than i did with v3alpha or v2 for that matter
A few simple tweaks but indeed, no tweak stood out as making a difference. Very tentatively the best success was extending the song from a clear voice specifically using a blank style prompt in the extension, but with a small sample this isn't really evidence that blank style extensions improve things. (all tests https://suno.com/playlist/06bb8e84-f0b1-4fa4-b564-5ceaa8253db8)
https://suno.com/song/8667a5f2-ad54-446d-a5ee-964d10dbbfdd (reasonably decent blank extension)
Swing song. Listen and make your own with Suno.
i'm definately gonna stick to using [Refrain] for now, because it makes sense that it would pull the bot less in the multi-voice direction
i've tried that too, but the choir effect usually increases during the song, so even starting out well is really not a guarantee.
in fact, must best work-around so far has been to do the first part in v3, then extend using v2 (which does better voices), and then maybe finish up using v3.
it doesn't seem impossible that blank style, lack of style is vaguely less like a standard song structure so less likely trend towards a big chorus like if you ran 1000 tests maybe starting a song with your style prompt then immediately continuing with lyrics but a blank style could reduce odds.
For a rap voice, the choir effect is really a no-go, and i'm having much more luck by simply sticking to v2 entirely. Example:
i haven't experimented enough with a blank style btw.
oh, and that's danish rap btw. 😅 not a case of wrong pronounciation.
i just tried two times [blank style], and got a choir in both cases.
(pop-rock both times)
on that rap example, it does say [Chorus] btw, and yes, it does a chorus. but only for those parts.
"realistic" soundstage prompt test (careful it screams)
the stereo is much too wide. it's not mono compatible.
on that soundstage test there
awesome drumming though 😄
I mean that was kind of the point of the testing to get something wider than normal suno output
ohhh! well you sure got something wide there 😄
slightly wide
people listening in mono will hear the drums being louder than the guitars.
people listening in stereo will probably hear the drums being softer than the guitars.
At first I thought it was german and that I was having a brainstroke because I couldn't understand anything 😛
ha ha
Cryptomedic: The Sea Ballad is my favorite of those. 👍
спс
vˈas ˈɪst dˈɑː lˈoːs??
you can use German directly ... works fine
catch the fish?
that moment when Youtube suggestions are actually good :https://www.youtube.com/watch?v=xp-8USQQc1Q
Official music video for “Fresh Fur,” plus intro track “Realm,” from Castle Rat’s Debut LP “Into The Realm.”
Pre-Order “Into The Realm” here: https://castleratband.bandcamp.com/album/into-the-realm-2
Produced, Directed, & Written by Riley Pinkerton
CREW
Director: Riley Pinkerton
Camera: Olivia Cross Cummings
Costumes: Riley Pinkerton
Makeup/H...
That's the complete opposite of Songs generated with Suno or Udio, put on a RunwayML/Sora/Noisee MV ...
The MV was recorded on VHS (no idea how they sourced tapes in 2024) and they are actually playing their instruments 😉
So, been messing with Udio a lot. I see a lot of headlines "Udio is now THE BEST"... I just think Suno still has something special going on. Suno's model generates songs that are way more "catchy", "Addictive", "get stuck in your head" "ear worms". While Udio definitely has some things that are better (sound quality not getting muddled), Udio's songs are just not that catchy (for the most part). Maybe after 50 generations, you get something ok-ish. But in just a few gens on Suno, there are some really great songs with good intros, hooks, structure, etc. Suno can still be the top music generator. The core of what they have going on is great. Keep up the good work. Improve the quality and all that. You guys will be killing it! The competition is great. Both services have things that are stronger. We are living in some fun times. I am looking forward to the next models (from both companies)
Udio doesn't know very many genres.
I tried multiple different ones, ended up trying simply "Blues" before i got something recognizable.
Whereas Suno knows pretty surprising things like even "soukous" - a style from Congo. Example (sorry to anyone who speaks french):
not a perfect result (it's a bit too "schlager" like), but it's not far off.
try "Zouk"
oh, good point
I lived a long time in Martinique 😉
hah!
also : I HATE Zouk 😛
LMAO!!!
i don't know why i have a soft spot for that stuff. maybe the exotic and tropical feel makes me forget the cold scandinavian weather.
when it's all that everyone around you plays and your "normal" mood/mode is more akin to semi-dressed Frazetta lady up there, Zouk can be particularly grating 😉
but here ... to heat up your cold scandinavian heart (technically that's Guadeloupe Zouk and not Martiniquais ... but /shrug https://www.youtube.com/watch?v=0G02YNxLk7g
🎵 "Laisse Parler Les Gens" est le premier extrait de l'album "Dis l'Heure 2 Zouk", interprété par Jocelyne Labylle, Cheela et Passi.
🎧 Streamez l'album ici: https://spoti.fi/3hjeDjC / http://bit.ly/2Zouk_Deezer
👉 Retrouvez tous les classiques de Passi: http://bit.ly/2DwuZ1O
(one of the few Zouk songs I actually like 😉
probably close enough for someone like me 😅
it's slightly different from what i know as Congo "soukous"
slower
yeah ... which is probably why I like it 😛
but my guess is that Zouk might be more present in the engine's training as it's listened to a lot in France
yeah, i wonder what the data set was.
Hello, the other day I asked the group about recommended music distribution companies and was told about ditto music. I've been looking into them but I also wanted to ask again if anyone else has any other suggestion for music distribution companies that play well with AI generated content or that they're successfully using with Suno generated music. Thank you.
Off the top of my head, there's tunecore, distrokid, lablworx
