[Discussion] How do you edit and use json files?

I finally installed spleeter in Windows using Anaconda after a few hours of trial and error and tested with the 5stem model. It works.

I notice in the documentation that you can use a json file and customize the behavior a bit. I downloaded the 5stem model and ran it. It seemed to take longer and produce very, very marginally better results.

Is there any documentation about how to edit it and get particular results? I'm especially interest in separate bass parts as good as possible to make it easier to learn them.

EDIT: I found some documentation of command line options themselves with "spleeter separate --help". Looks like you can do more with JSON.

Start with the existing documentation and the `spleeter separate --help` entry point to determine how JSON files are edited and used with the 5stem model. Done means documenting the relevant JSON options and how they affect separation results, including whether they can improve isolated bass parts.

Written by the indexing model from the issue text.