(1972) Sony SQ ad (SQD-1000, SQA-200 decoders)

[kfbkfb]

1st [Sony] SQ decoder(s) with logic circuits.


Kirk Bayne

 

[par4ken]

1st [Sony] SQ decoder(s) with logic circuits.


Kirk Bayne

I have the box in front of me. I'll check it out in a bit. It doesn't look to have enough components to have any logic in it but we'll see. Any such logic would pre-date the front to back logic decoders. No mention of any form of logic is on the box itself. Only 6dB of enhancement should not be that disturbing vis-a-vie audible pumping. I suppose that vocals could shift around a bit, without any rear blend.

 

[Soundfield]

Odd that. I’d always thought that the SQD-1000 was a completely basic decoder with no enhancement whatsoever. Tab Patterson even described it thus-

But, the Sony advert is quite definitive in it having some form of front –rear enhancement and I guess they should have known! I can’t find a service manual or circuit diagram on-line anywhere so am unable to confirm one way or the other.

 

[kfbkfb]

Only 6dB of enhancement should not be that disturbing vis-a-vie audible pumping.

The way the Sony SQ ad is worded, it seems like the front-back gain riding logic senses CF (and probably CB) content and adjusts the gain to give a net 9dB channel separation (kinda close to what Involve Audio discovered ~40 years later - only 12dB channel separation is needed).


Kirk Bayne

 

[Sonik Wiz]

The way the Sony SQ ad is worded, it seems like the front-back gain riding logic senses CF (and probably CB) content and adjusts the gain to give a net 9dB channel separation (kinda close to what Involve Audio discovered ~40 years later - only 12dB channel separation is needed).


Kirk Bayne

Just looked at the Sansui QSD-2 specs for SQ. It states 20dB separation Left/right and 12 dB front/back. Left/right direction sensing is fixed for SQ decoding in the Sansui unit. So this is real front/back enhancement where SQ needs it. And it is truly push/pull complimentary where as Vari-Blend only affected the rear channels.

And the current Surround Master has even higher L/R specs and even better F>B or B>F specs than the Sansui gear at 19dB.

 

[kfbkfb]

https://www.hifiengine.com/manual_library/sony/sqa-200.shtml^^^
SQA-200 (also has f-b logic) service manual


Kirk Bayne

 

[Soundfield]

The way the Sony SQ ad is worded, it seems like the front-back gain riding logic senses CF (and probably CB) content and adjusts the gain to give a net 9dB channel separation (kinda close to what Involve Audio discovered ~40 years later - only 12dB channel separation is needed).
Kirk Bayne

Nothing so sophisticated!
I couldn’t find the SQD1000 manual but Sony say in that advert that its decoder circuitry is the same as the SQA200 whose manual I have found. It confirms that there is indeed some form of front to back logic. Sony describe it thus-

So it’s a pretty crude blending of the front or rear channels if a mono (centre) signal is detected in the other pair. Their claim of this circuit ‘making SQ more close to discrete’ seems ludicrous.

 

[kfbkfb]

So...an early form of variable matrix SQ decoding, somehow I got the idea that Sony was using gain riding only logic in these SQ decoders, I guess I was wrong about that.


Kirk Bayne

 

[Sonik Wiz]

Nothing so sophisticated!
I couldn’t find the SQD1000 manual but Sony say in that advert that its decoder circuitry is the same as the SQA200 whose manual I have found. It confirms that there is indeed some form of front to back logic. Sony describe it thus-

View attachment 78285
View attachment 78286

So it’s a pretty crude blending of the front or rear channels if a mono (centre) signal is detected in the other pair. Their claim of this circuit ‘making SQ more close to discrete’ seems ludicrous.

Interesting info, indeed. Now then in the case of say a SQ left front signal, that would not produce a sum or difference control signal for blending. Would this put the circuit performance back to simple decoding: 3dB separation front to back?

 

[Soundfield]

Interesting info, indeed. Now then in the case of say a SQ left front signal, that would not produce a sum or difference control signal for blending. Would this put the circuit performance back to simple decoding: 3dB separation front to back?

Yes. The blending is only applied in those two very specific Cf and Cb occurences. Otherwise it maxes out at 3dB! But I’m really not sure why Sony bothered with the F-B Logic, it seems to have very little effect. Even the service manual doesn’t attempt to quantify it, rather feebly suggesting that when centred signals are present-

But despite clearly being pretty ineffective, it is quite complex and represents a fair proportion of the overall circuit (two sets of comparators, detectors, AGC and mixers). They could have reduced the overall price of the thing by 20% by leaving it out and no one would have noticed the difference! It would probably have actually sounded better if they had just spent the money on more accurate phase shifters.

 

[par4ken]

The way the Sony SQ ad is worded, it seems like the front-back gain riding logic senses CF (and probably CB) content and adjusts the gain to give a net 9dB channel separation (kinda close to what Involve Audio discovered ~40 years later - only 12dB channel separation is needed).


Kirk Bayne

That ad took me aback with mention of logic. Obviously it was a precursor to actual front-back (gain-riding) logic. I think that F/B logic was actually a step backward, introducing unnatural and unnecessary pumping. With the variable logic or blending they solved the SQ problem of the vocals coming from everywhere at once, or as the critics said nowhere at once. The more common solution was fixed blend, usually 10% front and 40% back. IMHO that (fixed blend) is a case where the cure is worse than the disease! The SQD-1000 keeps the vocals up front like the fixed blend decoders but without the left to right sound stage narrowing!

Such a scheme would've made the decoder work decently on EV-4 Dyna and even RM/QS. It would be good for Dolby Surround as well. When the vari-blend idea was brought back it was simpler to forget the front blending as it was felt that the real problem was Cf getting into the back. A mistake IMHO, dual blending would've been better.

Now I wonder how it would sound if the SQD-1000 was used as the headend for a full logic decoder, just leave out the fixed blend that is often used on the output. Much room for experimentation!

 

[kfbkfb]

Have you had a chance to try the SQD-1000 on some of the more difficult to decode SQ content (I own very few SQ records and have only heard the SQ decoder in the Pioneer QX-747 on a regular basis [not for about 30 years though])

Maybe try some of the early SQ discs w/CB content (IIRC, the Indian Reservation SQ disc has specific CB content) to see if the SQD-1000 is able to better localize CB?


Kirk Bayne

 

[par4ken]

Have you had a chance to try the SQD-1000 on some of the more difficult to decode SQ content (I own very few SQ records and have only heard the SQ decoder in the Pioneer QX-747 on a regular basis [not for about 30 years though])

Maybe try some of the early SQ discs w/CB content (IIRC, the Indian Reservation SQ disc has specific CB content) to see if the SQD-1000 is able to better localize CB?


Kirk Bayne

I'll give it a try but I'm used to the S&IC which has no problem at all localising Cb.

The SQD-1000 should be able to do it just as well as it can Cf. Listening to SQ source material through the small speakers surrounding my work bench/computer desk the decoded result is amazing I have no problem localising separate sounds from each speaker. I'm thinking that if the decoder is hooked up to my main system the subjective results will be not nearly as good. Just as the QS-1 can sound amazing with closely spaced speakers but seems to do little in a larger room.

 

[barfle]

I still have one. It’s been in a box for a decade or more. I doubt that anyone would want it.