Implementing Custom Output File Naming for MDX, MDXC, and VR Models

[Bebra777228]

Hello!

I've implemented a feature to customize the output file names for MDX, MDXC, and VR models within the separate method. However, for the Demucs model, I've left the naming as is, as I'm not entirely sure how to implement this feature for it.
#128 (comment)

Here's an example of how to use the new functionality:

from audio_separator.separator import Separator

separator = Separator()
separator.load_model(model_filename='model_bs_roformer_ep_317_sdr_12.9755.ckpt')

output_files = separator.separate('music.mp3', 'primary_output_name', 'secondary_output_name')

print(f"Separation complete! Output file(s): {' '.join(output_files)}")

In this example, primary_output_name and secondary_output_name are the custom names that will be used for saving the output files. For instance, the files will be saved as primary_output_name.wav and secondary_output_name.wav.

You can also rename specific stems:

• To rename the primary stem:

  output_files = separator.separate('music.mp3', primary_output_name='primary_output_name')

• To rename the secondary stem:

  output_files = separator.separate('music.mp3', secondary_output_name='secondary_output_name')

Or do not rename the files:

output_files = separator.separate('music.mp3')

---

Screenshots:

---

I encourage you to test this update, as you have a better understanding of programming than I do. You may be able to find a solution to customize file names for the Demucs model and fix other possible problems if they occur.

[beveradb]

Nice one, thank you for making this addition!
It's an improvement I'll likely use in my own tooling 😄

Released in audio-separator version 0.24 onwards 🎉

[Bebra777228]

Demucs stopped working. 🤔

ERROR:root:Demucs separation failed: DemucsSeparator.separate() takes 2 positional arguments but 4 were given

I hadn't tested it before, I thought it would work fine, but no, it's giving errors.