Let's assume the same background music is used at different places in the same video and also in different videos (e.g. documentary series).

AudaCity has a vocal removal plugin that works well sometimes, but still leaves some background noise and degrades sound quality.

How can the information of the same background music from different sources be combined (preferrably using AudaCity), so that the common information (background music) stays and the uncommon information (voices, other noises) get canceled out?

  • 1
    Will they all line up to sample-accuracy?
    – Tetsujin
    Commented Jun 2, 2019 at 16:15
  • @Tetsujin Yes, they line up.
    – neverMind9
    Commented Jun 2, 2019 at 18:57

2 Answers 2


Without going into too much detail, the result you seek cannot be achieved in this way. The technique you are referring to is often used to recover "uncommon" information - you are seeking the reverse.

  • How can it be achieved then?
    – neverMind9
    Commented Jun 2, 2019 at 23:29
  • @never - it cannot, unfortunately
    – Rory Alsop
    Commented Jun 6, 2019 at 10:36

Not sure if I got your question correctly. But if what you're trying to achive is the sole background music without the vocals why don't you just extract the vocals by cancellation as already described by you and afterwards cancel out the vocal from the original track with the vocal that you extracted?

Due to the left over backgound noises and degradation of quality you might cancel out some things that you didn't mean to cancel out of course, however trying this approach should be worth a try.

  • I am trying to get a soundtrack out of a documentary. Sometimes, Audacity successfully extracts the instrumental. Sometimes however, Audacity turns the audio into distorted MP3 compression noise or even silence. That vocal removal does not always work, especially not when the stereo sound tracks of a source videos are identical, effectively making it mono.
    – neverMind9
    Commented Jun 12, 2019 at 17:16
  • 1
    @neverMind9 I guess I misunderstood part of your post, so my mentioned approach would probably not work as you say that the vocal removal tool doesn't always work properly. However this tool is based on some algorithms that have the potential to remove vocals in some case. As I've heard a presentation by an university about a course of study for audio engineering they spoke about a technique how you can extract vocals very precisely by applying different algorithms. Hence extracting the vocals is possible but I don't know how to achieve it technically.
    – Samaranth
    Commented Jun 14, 2019 at 11:59
  • 1
    Still, thanks alot for the help.
    – neverMind9
    Commented Jun 14, 2019 at 17:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.