FORUMS: list search recent posts

Voice Recognition Solution for mono vocals/background separation?

COW Forums : Adobe Audition

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Christopher Erickson
Voice Recognition Solution for mono vocals/background separation?
on Dec 4, 2012 at 9:26:17 pm

Hey y'all. I'm working on the edit of a multicamera edit of a theater production I did in the summer and I recorded second system audio off of the board to my H4n. The sound designer failed to tell me that he only had one available AUX out so instead of getting vocals on one track and band on the other, I got them on the same track. I ended up needing to shoot on multiple days and because every performance is slightly different, editing the different days together makes the vocals jump in time.

I was thinking about solving the problem and looked it up on the web for how people have handled it in the past. I didn't find what I was looking for, so I came up with two hypothetical solutions. The first, as the title of this post suggests, is that there might be a plugin that utilizes voice recognition technology to process and isolate vocals from the background. The second is that there might be a plugin that isolates only frequencies that are at or above a settable dB level (Think of the spectrographic view and imagine only effecting stuff of a certain color or hotter, not the other stuff, if my description wasn't clear).

So does anyone happen to know if either these things exist on the market for Adobe Audition or just generally for the Mac OSX platform (especially Soundtrack Pro or another third party)? I'd perfer freeware solutions, but anything will do, really. Thanks for your time.

Return to posts index

Angelo Lorenzo
Re: Voice Recognition Solution for mono vocals/background separation?
on Dec 6, 2012 at 5:06:01 am

No, there isn't anything to solve your issue perfectly. The human voice is frequency rich which means there is a lot of cross talk with instrumental frequencies (the whole reason singing to music sounds so good is the same reason you can't easily separate the individual sounds)

Here is what I recommend:

1) Rubber lip it. Use the audio from one night and then edit pieces from the other nights to line up as best you can. You'd be surprised how forgiving this technique can be

2) You can try taking a noise reduction profile in Audition, cranking the noise reduction line to 100% and then drawing a dip in the graph to keep some vocals. This works.... ehhhh. It also tends not to affect much in terms of percussive instruments. Noise reduction is just a frequency selective noise gate. Last resort.

Angelo Lorenzo

Need to encode ProRes on your Windows PC?
Introducing ProRes Helper, an awesome little app that makes it possible
Fallen Empire Digital Production Services - Los Angeles
RED transcoding, on-set DIT, and RED Epic rental services
Fallen Empire - The Blog
A blog dedicated to filmmaking, the RED workflow, and DIT tips and tricks

Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2018 All Rights Reserved