Ripping music from TV shows - Removing voices?
 

There are some music cues I wanna rip out from the show House (and various others like Heroes). What is the best method and program to REMOVE those voices though? Does getting the DVD help since the DVD versions have 5.1 surround sound mixes? I used adobe audition to removed voices from the downloaded episodes that are TV rips, and its ok, but doesn't completly remove it and still has faints sounds from the voices.

Here's what I ended up with.

Putfile - House and Camerons theme music

Anyone can help me out?

Adobe Audition & Yogen Vocal Remover

I said already used Audition's vocal removal. Doesn't take away the voices completly cause I can still hear the faint sounds.

Unless you're saying I use Yogen's as well on top of that?

As far as I know, there's no perfect way to remove vocals from an audio track. The one that you came up with sounds pretty good as it is already, doubt it can be improved.

Removing a signal A from another signal B usually works quit good if you have the signal A in it's original form.
In your case it means you have a separate recording of the voice in the track.

Without any knowledge about the unwanted signal you won't a chance to completly remove it (not to mention that the filtering removes also wanted signal parts).

Well, maybe the Season 3 DVD will improved the quality, that was ripped from the TV AVI rips found on torrents... Since after all, the DVD does have 5.1 surround... Thanks for the tips!

I'm with shadoweave on this one, there's no one perfect way to remove vocals from music unless you have all the original sounds separated into audio channels. If the sounds have been merged just into one track, it becomes the sound equivalent of flattening an image in Photoshop, all the data underneath the top layer is either lost or mixed into one.

I doubt having 5.1 surround would make much of a difference honestly. As S_K said, the only perfect way is having the two audio tracks separate in the first place.

I know of a matlab program developed by a grad student that can remove voices via frequency analysis, especially if you have existing samples to compare against. I don't think there was any kind of windows-runnable code at all, let alone an audio editor plugin, it was just a neat matlab script.