Indeed [Bob Carver] was a true bull****ter...

I would rather say hyperbolic and self-serving.


In Carver's case, he had an amazing and inquisitive mind...

But one lacking discipline and true curiosity. See following.


I was personally a very big fan of his Sonic Hologram, the first
consumer interaural canceller unlike any other device offered
at that time except for a very weak competitor from Sound Concepts
using a very noisy bucket brigade delay line.

I reviewed the Sonic Hologram for "Stereophile". I was pleased to find that,
with my own coincident-mic recordings, the result was much closer to what I'd
heard standing at the mics. The results with commercial recordings -- which
are rarely simply-miked -- were more variable.

The Sonic Hologram was hardly the first consumer crosstalk-canceller. JVC's
BN-5 did an excellent job. The effect with binaural recordings was nothing
short of spectacular.

The Sonic Hologram attempted to get rid of delay lines by substituting
constant-group-delay phase shift. Such circuits cannot actually delay the
signal. Rather, they shift the waveform envelope. * Given how well the Sonic
Hologram worked, I suggested to Bob that /perhaps/ the circuit did "good"
things neither he nor anyone else was aware of, and he should look into it.
"No, it works the way I say." And that was the end of it.

* I'd better explain this. Imagine you've designed an op-amp circuit to
produce 1ms of constant group delay over the audio band. If this circuit
"truly" delayed the signal, then its output would be zero for the first
millisecond after a signal is applied. But this is impossible, because a
finite input (with zero output) would drive the op amp against the rail. What
is actually output is a signal that -- after passing through the feedback
network -- cancels the input signal. (This is The Basic Principle Of Op-Amp
Analysis.)

You can easily show this with Laplace transforms. I did it 30 years ago, and
am not in the mood to do it again.