[BC] Re: CD vs. LP (or PL .sv DC?)

[Jim Wood, C.P.E.W.]

> What it did was reverse the polarity of the peaks.
> Highly asymmetrical waveforms such as voice
> had their positive peaks inverted to negative peaks.
> Some voices are highly asymmetrical and would
> otherwise saturate the tape, or make the recording
> more noisy if the overall level were decreased to
> accommodate the positive peaks.
> This avoided tape saturation by virtue of the fact
> that you cannot exceed 100% negative modulation.
> Since the tape bias is amplitude modulated, reverse
> duplication actually produced a better sound without
> the need for audio processing.

Almost.  Actually, HF bias isn't amplitude-modulated by the audio, it' 
simply combined with the audio at the recording head.  Although 'phase 
flipping' is a trick used to some degree of modulation advantage in AM 
broadcasting, it really doesn't apply in this case.  There is no -100% 
modulation limit with tape recording as there is in AM radio; negative peaks 
produce the same flux level in one direction as positive peaks do in the 
other.

What re-recording 'backward' does is actually help to restore the audio 
waveform to its original shape.  The analog magnetic recording process 
imparts a sort of 'phase scrambling' not unlike the all-pass filter in front 
end of an Optimod 8100.  Re-recording in the other direction reverses, or 
undoes, this, usually without adding a satanic message.

However, the old practice of over-recording a tape for more 'punch' on the 
air did two things.  First, the phase-scrambling effect reduced the 
peakiness of voice waveforms, and secondly the soft compression function of 
tape saturation decreased the average-to-peak ratio of the voice.  All this 
was done with some pre-emph / de-emph EQ going on, so THD in the final 
result didn't sound nearly as bad as it might have.  The added IM distortion 
added a 'ballsy' quality to male voiceovers.  Someone ought to make a box 
that simulates this.

Jim Wood
Brea, CA