Is it not possible to generate instrumental tracks without lyrics or vocals?
No matter what I do, it comes out with vocals layered on top, but I need an instrumental track without lyrics or vocals.
+1 to that! This is very odd!
Can try this for lyrics:
[Intro]
[Instrumental]
[Break]
[Instrumental]
[Outro]
[End]
or just [instrumental]
Sometimes it works ;- ) but depending on the genre and prompt it tends to add vocals
[Intro]
[Instrumental]
[Break]
[Instrumental]
[Outro]
[End]
or
[Instrumental]
I tried it like this, and as expected, the vocals come through in some tracks.
Since the developer says there are no planned songs, it looks like we’ll have to use Suno for the time being.
have you updated nodes or custom ?
have you updated nodes or custom ?
ComfyUI is always kept up to date, and I'm using the default YuE node.
Other YuE nodes don't use ComfyUI-specific models, but rather general ones, and they won't even install, so I can't use them.
Are there any nodes that support instrumental music?
This video says you can do it by describing the sounds of the instruments in square brackets in the lyrics textbox: https://youtu.be/9RtywbN--QE?si=rBQh4tZlhOQ6q-68&t=1170
[steady buildup of staccato strings and drums]
No, that doesn't work either. It still starts singing after about 20 to 30 seconds anyways.
+1. It makes pretty nice gens, but really needs some ways to make only instrumental compositions.
Wow.. forcing vocals severely limits this model.
yep, mentioning instrumental only seems to make weird vocals(that match the song, but dont make verbal sense).
would be nice to force the model to not have vocals or removing vocals from the final generated song
For me i never use vocals for the music i use or is useful for me its always instrumental so i hope there are fixes for this.
