Pull music files apart into instrument stems.
For decades, DJ's and producers have struggled to find "stems" of various songs that they wanted to mix, remix, or mash-up. Stems are the individual recordings that make up a song. Each instrument (or group of instruments) is a stem. So you might have the lead guitar of a song as one stem, along with the drums, vocals, etc. Stems have always been hard to come by. They typically were guarded very closely by the artists and producers that created the music and were very rarely released.
But today, AI can be used to un-mix a fully produced recording and do a pretty decent job extracting a song into its component stems.
That's what the Audio Dissector does. A frontend for the latest AI separation algorithms for Windows that will allow you to import an audio or video file and re-balance its instruments, or extract them into their own stem instrument files outright. (Specifically, Audio Dissector has built in modules for both Demucs and Spleeter.) Furthermore, it goes beyond Spleeter's current capabilities and can extract stems from recordings of any length (limited only by your system's memory). It also has the ability to save the analyses into project files so they don't need to be subsequently performed again and can be quickly reopened. Here is a video demo of the software in action.
Here's an example of what the Audio Dissector can do:
ORIGINAL SONG:
DISSECTED INSTRUMENTAL:
DISSECTED DRUMS:
DISSECTED VOCALS:
Audio Dissector is still in private beta. If you have a beta key, you can download it here: