>> Note that you can't play these recordings "properly" without an advanced
>> SQ decoder. My Tate II seems to have "bit the dust", and can't be
>> repaired.
>> This is unfortunate, because some Columbia SQ LPs have excellent sound.
> Surely an universal genius like you should be able to repair the
> decoder. They are fairly simple devices. Circuit diagrams can easily
> be found on the net. It shouldn't be too difficult to build one, either.
On the assumption that, despite your sarcasm about my intellect, this is a
sincere remark, I'll give it a proper answer.
A "basic" decoder for any matrix surround system is fairly easy to build. It
requires all-pass phase-shift circuits, plus a matrix to generate the
speaker feeds, all of which are easily put together with inexpensive op
amps. The problem is that subjective separation is poor, * and further
enhancement is required.
This enhancement is provided by dynamic crosstalk cancellation. QS used
VarioMatrix. The comparable system for SQ is Tate. These are not simple.
Circuitry is needed to determine the intended direction of the loudest sound
(this can change many times each second), and variable decoding that
positions the loudest sound correctly (even if it results in the wrong
directionality for all other sounds). The attack and release times of the
variable decoding have to be adjusted to minimize IM distortion.
The patents explain how this is done. But the Tate System implements the
process as ICs, which are no longer made. (At the time I purchased my
Fosgate Tate II decoder more than 25 years ago, the source was drying up,
and Fosgate could no longer get the chips. Not surprisingly, a working
decoder goes for over $1000.)
The logical solution would be to implement SQ decoding in software, but
there are probably no more than 100 people in the world interested in
playing SQ LPs, so no one is willing to do it.
* The exception is Ambisonic decoding, which provides stable imaging over a
wide listening area without additional enhancement.