What happened to all the AI music threads? Don't tell me you guys got bored with it already? The Bing AI image threads went on for months
AI Music
What happened to all the AI music threads? Don't tell me you guys got bored with it already? The Bing AI image threads went on for months
Bot music doesn't have the same capability to serve the cumbrains who repost the image generation generals twelve times a day.
Ah, I see, no anime girls = no interest
Hopefully soon
bullshit. I've spent hours making music.I'm currently generating a concept album about a train ride.
Ok, and? How does that translate to the topic of music generation getting five or more full threads a day, like the cumbrain-fileld image generation generals?
Udio will in the future let you upload your own AI Art covers so then you can put whatever anime girl you want on it.
In terms of weeb shit, its because no one knowa that both Suno ans Udio can do japanese vocals and lyrics. If they did, they would be flooding the place yesterday to make anime openings.
Wake me up when a local model exists
this
frick censorship
what would you even want uncensored??
vulgar shit or stealing real artists voices? AI music is doomed to fail any way, you wil never replace human art
This. I want to finetune on music I like, you quickly noticed how generic stuff like Udio is and since you can't mix and match artists you lack control when you want a specific sound
It will be possible eventually. I remember being delighted how dall-e managed to pull the found footage of king kong fighting in bosnian wars.
you actually need to learn music lingo to have more control
https://www.reddit.com/r/FL_Studio/comments/1aleyui/music_creation_app_sunoai_is_actually_getting/
analyze this prompt and notice
that's child's play compared to the shit I've made in suno
https://suno.com/song/1cb5a647-8643-4f43-b069-a27264d37d2f
Neither model is trained on a real understanding of music theory.
Same issue with image gen. Its made by CompSci individuals that only have an outsider's knowledge and need people in the field to help contextual the data they have so they can get more out ofbwhat they already have.
yeah, this, there's obviously no real understanding going into either of those, its what you would if you get if you spent ten thousand years just copying musicians and then trying to make something original without even taking the pains to figure out what a scale is
similar issue with imagegen, even with current models you sometimes randomly get some unholy abomination mixed in with a batch of like 4 normal looking images
>What happened to all the AI music threads?
Were there any? I only saw random posts with songs.
Isn't this based on SD's open source model?
music? no
completely different model
I just meant stability sorry
https://stability.ai/stable-audio
thought it was based on this or something
This, this is what i really want. And the ability to train it on new music
right here
https://www.udio.com/songs/tLJapLpPD4xSfDhR4d6Ukc
This goes hard
goes slightly less hard, starts out as simple as possible but comes together nicely
https://www.udio.com/songs/87SsS1Uqk3XBqqERqycLJf
I'm going experimental
https://www.udio.com/songs/mgo55nQ4aNkuzwqDDF6ByD
https://www.udio.com/songs/foVqjXtP1jHxc7QpkM1khD
Working on something right now, everyone got 200 free credit refill this morning and I was out.
Any love for Yogurt Prophet?
https://www.udio.com/songs/kugmDw3foEuMX7CJqDocUf
This one is good
https://www.udio.com/songs/dSQks7EteVAKfZoTp4dpDA
I love AI music.
become da scrum master
https://vocaroo.com/18WgSIWDbAZE
Someone post baskin roberts
any local models to run? havent followed the scene at all
Not yet, but the available platforms are free to use right now
how do i train a model off tablature or midi files