Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This is a longstanding problem in audio engineering, along the lines of "how does one un-bake a cake to get the eggs?" There's always going to be artefacts and distortion ranging from unpleasant to extreme, hence audio engineers when mixing from stems/channels will do a stereo instrumental mix and a mix with vocals.


Yeah, it's impossible to do it perfectly as long as human voices' frequencies overlap with instruments, basically.

It's like asking to recover layers from a PNG/JPG File.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: