35mm Push-Pull Density Audio Film - Saturday, January 15, 2011
A comparison of the audio characteristics of the two channels of a push-pull density audio track versus the combined track.
I was working away at the computer the other da and the scanning machine was idle, so I put in some unknown 35mm BW neg material, just because I love it so much and I found the pix film was wound with a density audio film that appears to have two tracks.
It was a short piece of film, few minutes and I thought nothing of it. Then I scanned the separate audio track film and thought I would process the stereo channels at just for fun.
At this point my extreme youth got the better of me. When I processed the channels as a mono combination, which I always do first, I got a bunch of distorted nonsense. Then after puzzling I decoded only one of the tracks in my software and it started to sound better.
A careful examination of the track images finally clued me in to the fact that the tracks were in fact out-of-phase signals. I happened to be talking to David Sheppard that day and he explained that I was looking at “John Maurer “ push-pull density tracks.
This is a form of differential signaling which is used throughout signals processing for a whole bunch of technical benefits. One neat thing is that each of the channels “sort of” looks OK and works OK on it’s own, but the distortion level is quite high.
However, when the differential processing is performed the result is as intended, quite superior. What I have done is used our processing software to decoded the left and right side of the density track separately to Mono mp3 files and then also converted the differentially processed density to an mp3.
My observation is that at low audio level, either tack alone will give tolerable distortion levels, but as the volume increases the performance of the differential processing is much better. Of course this is what was intended.
Caveat --. All of the audio you can here below could have used some low pass filtering and I may not have balance the channels perfectly before differential processing.
These were all coded to mp3 at 160 Kbps so the demo would not take forever.
All of these audio signals were decoded by analyzing the images of the sound tracks.
John Gledhill john@bitworks.org 905 881 2733
The images from the film are in ˝ resolution if the “original” size is selected.