Blumlein

Just thought I would share something cool I just did today…

Alan Blumlein, a freakin genius, patented stereo playback from two loudspeakers in 1934
http://www.google.com/patents/about?id=RAxoAAAAEBAJ
and already understood the HRTF involved in recorded sound, and how to correct for it in the recording. EMI, where he worked, used it for a little while but wasn’t doing the kind of stereo recording that we know and love today… combined with analog filter networks having phase distortion, it never caught on much.

The problem is that the stereo image becomes too wide in the treble because the recording isn’t pre-compensating for what the human head is going to do to the audio.

Well… I advanced Blumlein’s correction ideas today in a way I haven’t seen done before. 😀

If you consider his formula for sound recording that should stay compliant with proper binaural reproduction from speakers…
? = ?y sin ? / 2c
where y is the side amplitude (above 700 Hz), and c is the sound velocity… y being the variable we can control… and
(L-R) / (L+R)
is the ratio of the width.

We can use this setup (which i’ve extremely over-simplified the description of) to make sure the high frequency stereo width never goes above Blumlein’s formula evaluating to true. This will prevent the high frequency image from becoming too wide, without much perception of a reduction in the treble spectrum.

The red part of the circle meter/knob shows the gain reduction of HF Side. When maxed out, the red is all the way to the right, and there’s the maximum of 3dB of gain reduction to the sides, per the standard Blumlein shuffler design for standard HRTF pre-correction. So yes, this is basically an automatically relaxing Blumlein shuffler. 😀

"Speed" is a limiter on the rate of gain change to prevent audible distortion. Since it’s only effecting sound above ~700 Hz, the rate of change can still be plenty fast enough for great control of high frequencies without much lookahead. Actually, I’ve found I prefer no look-ahead since it leaves a tiny bit more micro-dynamic of the treble transient’s width.

—

EDIT: I just modified it so the change in power gets compensated for in high frequency L+R (other than what was already happening by reducing L-R) which mostly prevents a tone shift or loudness change from happening… and it should keep the image more solid with Decca tree mic techniques too. 8)

—

Anyways… just thought I would share my neat little invention with you. Cheak out Alan’s patent. It’s a trip to imagine him working all of that out, for the first time ever, in & before 1934.