there's another free background vocals remover?

[dawn badger]

There's alternative to MVSEP and x-minus.pro for remove the background vocals? I mean because the premium of x-minus.pro is for paid users now and the queue of Mvsep is too long ✌️

[granite mango]

Yep, UVR 5 Google Colab https://docs.aihub.wtf/rvc/resources/vocal-isolation/#cloud-uvr

[dawn badger]

Thanks

[granite mango]

Yw

[dawn badger]

But the problem that queue there is probably long too like on Mvsep or we should just waiting or maybe because I'm not uploading the file immediately after I opened the Gradio right?

[dawn badger]

Pls do screenshot or screenrecord how to use that Gradio to remove the backup vocals to understand more or learn more?

[dawn badger]

Yes but it's seems not working when I marked the GPU

[chrome plaza]

Ayo? @dawn badger level 7 !!!

[dawn badger]

Maybe but i'm not familiar with this Colab and that Gradio yet so idk yet how to use this

I already know that I mean I'm not familiar yet is that Colab and Gradio for separate the vocal and instrument and I've been using Colabs and Gradios for a months now for infer and train the models

Yes but only for finding/searching models

I think more than 30 minutes or 1 hour

More than 3:00

Idk since I'm not familiar with that Gradio yet and I can't remember how many minutes but maybe I will try again

[past glade]

Can u send on what page are u using that?

[granite mango]

no thats just the other things loading, not a real queue of people, as long you are using the google colab one, also be sure you are using the free t4 daily gpu

[granite mango]

sorry but i cant do that, everything should be written on that guide, im honestly busy with other things like school

[dawn badger]

Okay

[dawn badger]

It's from Gradio from this Colab https://colab.research.google.com/github/Eddycrack864/Ultimate-Vocal-Remover-5.6-for-Google-Colab/blob/main/Ultimate_Vocal_Remover_5_6_for_Google_Colab.ipynb

[past glade]

Oh, so you're probably out of daily GPU

[dawn badger]

I hope someone recommend me another background vocal remover that has no long queue or short queue bc I think that Gradio has long queue too just like on Mvsep

[dawn badger]

but you can remove the background vocals on '/separate' command on Astra Labs tho

I mean I already tried the '/separate' command on Astra Labs the vocal downloader there didn't include the background/backup vocals and they only separate the main vocal one

I marked this as solve because Astra Labs helped me with this although they may not 100% remove the background/backup vocals or background noise

[granite mango]

nope gradio shouldnt, just be sure to use gpu

theres another site you could try which is x-minus.pro but its kinda paid

else more you could try

-karafan

[cloud bison]

KaraFan (Improved Version)

🥇 Awarded to 1st place by the community as the Best software in the World, for having the less vocal residues in Music track!

Note: Improved by AI HUB Staff members. Eddy and Leo_Frixi.

[granite mango]

i dont think theres other options than these

[dawn badger]

Yeah this is why I'm not using x-minus pro anymore

[granite mango]

then could try karafan or astra labs but honestly never tried it

[dawn badger]

I'm fine with Astra Labs but okay I will try this if it's works to me but pls do screenshot of the tutorial to understand more?

[granite mango]

sorry but i never used it, could try to see maybe in the github theres more info

[past glade]

Run the 2 cells and an interface will appear, so u need the settings this ones:

You can select MDX23C 8KFFT - V2 (VIP Model) and Instrum HQ 4 if you wish

Now I realize there is no Karafan documentation, I'll work on it

[dawn badger]

Okay but how to upload the music (mp3,WAV,m4a etc.) there to separate the vocals and instrument of the songs?

[past glade]

yep, only vocals and intruments, works with mp3, wav, flac but no with 48KHz audio

It's like a Vocal separation ensemble

[dawn badger]

Okay but pls do screenrecord how to upload the songs there and I was wondering how to upload/attach the songs/files there ✌️ but maybe we should just type the title of the song or copy an paste the name of mp3,wav etc. right?

[past glade]

Make a 2 folders, one for the input audios and one for the output audios. So u paste the path of that into the Karafan GUI

Karafan will look for every single audio into the input folder

[dawn badger]

Okay but I think I should move on already bc this KaraFan thingy seems it's hard and I'm fine with the vocal and instrument separator on Astra Labs bc Astra Labs is the easiest one

[lilac spindle]

karafan seems wont get BS roformer unfortunately

[dawn badger]

I agree with you but what's that BS Roformer mean? I've been making AI Covers for a months but I'm still not familiar with that