Tutorials

[djtayz]

Hello! I followed all the instructions and installed everything but when I try to launch the vocal remover I get the following error:

Attachment 657

I tried reinstalling Python and all the files but it didn't solve the issue. I don't have any experience with Python so I'd appreciate if anyone could help me with this issue.

Do you have Nvidia GPU with CUDA cores?

[mihadaniela]

Do you have Nvidia GPU with CUDA cores?

I have GeForce GTX 950M and I installed the CUDA driver specified in the tutorial as well.

[Anjok]

Hello! I followed all the instructions and installed everything but when I try to launch the vocal remover I get the following error:

Attachment 657

I tried reinstalling Python and all the files but it didn't solve the issue. I don't have any experience with Python so I'd appreciate if anyone could help me with this issue.

Which version of PyTorch did you install?

I reference 2 in the main thread, one with Cuda and one without.

[Anjok]

***Update***

The GUI has been updated! I uploaded the new version to the main thread along with an explanation of the new model and new options. I will be working on some new models in the near future.

Enjoy!

[NewAgeRipper]

***Update***

The GUI has been updated! I uploaded the new version to the main thread along with an explanation of the new model and new options. I will be working on some new models in the near future.

Enjoy!

Giving the new GUI a spin with a song that is just bass, piano and some strings+vocals to see if the residue can be cleaned up. How well could this work for older Guitar Hero songs that were only 3 stems?

[Anjok]

Giving the new GUI a spin with a song that is just bass, piano and some strings+vocals to see if the residue can be cleaned up. How well could this work for older Guitar Hero songs that were only 3 stems?

I haven't tested it on tracks like that yet. I found that acoustic, orchestral, or other lighter tracks can benefit with more passes through the stacked model.

Let me know how it works for you!

[Anjok]

I have GeForce GTX 950M and I installed the CUDA driver specified in the tutorial as well.

I apologize for the confusion, I didn't see this response.

That GPU is a little old and might not support the version of Cuda needed to run this AI on the GPU. Did you also install the Cuda driver that links to the Nvidia website, or did you just install the version of PyTorch that includes Cuda?

If you keep having issues, try installing the CPU version of PyTorch.

[NewAgeRipper]

I haven't tested it on tracks like that yet. I found that acoustic, orchestral, or other lighter tracks can benefit with more passes through the stacked model.

Let me know how it works for you!

Oh I will let you know. i wanna help improve with feed back as much as possible. I'm gonna send you couple tracks for you to work with to see if the models can be improved more. If you can get these tracks to be entirely vocaless you'll have it made.

[chilinvilin]

***Update***

First off I just want to say thank you to @NewAgeRipper for doing an excellent job on the tutorial and for the shout-out!

I'm ironing out some bug in the update and it should be released sometime this week for sure! I'm trying to make sure that all of my future releases can be universally run with no issues before releasing them. I've had some help from a gifted Python coder to help bring my vision for GUI to life and I'm VERY happy with the options we were able to add to make it the best!

To clarify, the stacked model was trained on pairs consisting of conversions of mixes and official instrumentals. The point was reduce the vocal pinches and audio static in some tracks. The model works best when a track is run through it more than once and that is where the "Stack Passes" option comes in. The amount of conversion passes needed to reduce the vocal pinches will vary from track to track.

Regarding the possibility of having an AI that keep background vocals for karaoke tracks - I'm not entirely sure how effective it would be considering how I would train the AI to keep them in. The results are not consistent enough. This is something that will need to be thoroughly tested and will require a lot most research.

Ok this is weird I tried out the new gui with the multi pass option on GNR's Sweet Child O Mine and it came out terrible compared to the colab version. My settings were using stacked model with stack passes set at 2.. Did I do something wrong? Files below..

Stacked Model
https://www.mediafire.com/file/aqb1x...x/SCOM2pass.7z

Colab version
https://www.mediafire.com/file/2ds3k...k/SCOMcolab.7z

The differences are night and day

[NewAgeRipper]

Ok this is weird I tried out the new gui with the multi pass option on GNR's Sweet Child O Mine and it came out terrible compared to the colab version. My settings were using stacked model with stack passes set at 2.. Did I do something wrong? Files below..

Stacked Model
https://www.mediafire.com/file/aqb1x...x/SCOM2pass.7z

Colab version
https://www.mediafire.com/file/2ds3k...k/SCOMcolab.7z

The differences are night and day

Same for me with the track I tried. more vocal bits were left in the instrumental than with not using it.