Flubly said:
DigModiFicaTion said:
Flubly said:
Can anyone tell me the circumstances wherein they were able to effectively use the "invert waveform" method of dialogue isolation?
Initially I used this method to turn my 2.0 stereo audio track into a 5.1 surround for my Shipwrecked edit. Ultimately I used the vocal removal effect.
Hmm, yeah it seems like the best route for getting better results is just getting into more and more sophisticated noise reduction. I have an engineer friend who owns the Izotope RX bundle (I'm not sure if he has the egregiously expensive "advanced" bundle). I've been thinking of asking him if he'd be interested in taking a look at some Ladyhawke stems.
I've seen phase inversion recommended once in a while, but it didn't make much sense to me since it's the dialogue that's in mono and music that's in stereo. You could do it with the soundtrack, but often with older movies there's so many variables against them matching. The movie is often just the capturing of the optical strip which makes it a different sound signature than the cd, and if it's a remastered copy the mixing and mastering can be different enough despite them both being clean.
If someone has ever been able to phase invert using a CD copy of soundtrack against a movie, I'd love to hear how that worked. I've never been able to do that effectively because of the above reasons.
I've never gotten phase inversion to work for film audio cleanup but also haven't experimented with it as much as I'd like to.
I think if you're lucky enough to be able to do the following it may work:
Rip the video source from its disc with all audio tracks
Demux the stereo audio track from the ripped video source
Rip the stereo soundtrack audio cd
Line up the soundtrack file with the demuxed audio file in your editor
Change the volume of the soundtrack file to match the volume of the music that is mixed in the demuxed audio file
Invert the soundtrack file.
There is only one broad reason this might not work, and that's of the waveform on the soundtrack doesn't match the source.
But there are many specific reasons they may not match up.
If the original sound editor of the film or show had access to the unmastered tracks from the soundtrack (like all the separate files straight from the recording engineer), they may have applied new effects (like reverb or eq), changed relative volumes (like maybe lowered the strings and raised the brass), or separated the channels into a surround mix that was then downmixed into stereo or they may have mastered the tracks differently somehow. If they only had access to the final mastered audio (like from a CD, the same as what we'd buy from the store), they still may have applied volume fades or envelopes that may be hard to match, changed pitch or timing, or may have recut or rearranged songs. Even the bitrate may affect results, I don't know.
Really interested in what others have learned trying this out.