(1974-12) SQ Stereo Enhancement [CBS Labs]

[kfbkfb]

It'd be interesting to try this preprocessing with the Surround Master decoder in SQ mode.

I'm not aware of any SQ decoders that used this CBS Labs devised SQ based Stereo Enhancement method.


Kirk Bayne

 

[Soundfield]

It'd be interesting to try this preprocessing with the Surround Master decoder in SQ mode.

I'm not aware of any SQ decoders that used this CBS Labs devised SQ based Stereo Enhancement method.


Kirk Bayne

Odd that they don't give an actual circuit. There is a lot more detail (but still no schematic) in Bauer's US Patent of 1976 "Method and apparatus for quadraphonic enhancement of stereophonic signals" -
US pat 3971890

 

[kfbkfb]

I wrote to CBS about SQ and received this mail, it does includes a schematic diagram, but in this digital age, I didn't think posting the diagram was useful.


Kirk Bayne

 

[par4ken]

This is actually how the Space and Image Composer enhanced stereo, but by doing it directly rather than the stereo signal having to be pre-encoded. By swapping the front and rear decoder outputs and reversing phase we can produce this form of enhanced stereo.

 

[kfbkfb]

I haven't used "Scripts" for audio processing, it seems like it would be easy enough to process L and R with a 2 scripts (180 and 270 modes) and output L' and R' to feed to an SQ decoder.


Kirk Bayne

 

[Sonik Wiz]

This is actually how the Space and Image Composer enhanced stereo, but by doing it directly rather than the stereo signal having to be pre-encoded. By swapping the front and rear decoder outputs and reversing phase we can produce this form of enhanced stereo.

How was it made variable in the S & IC?
I wonder how Fosgate did the stereo to surround in the Tate 2 decoders. The Sansui synth mode and the Fosgate both offered nice 270 deg wrap around. But the Fosgate sounded a little bit crisper in direction but the Sansui also enhanced center back which the Fosgate SQ method did not. The Sansui approach gave a bit more depth than Fosgate's.

In the SQ world there's no problem that can't be fixed by yet another 90 deg phase shift...

 

[Soundfield]

I wrote to CBS about SQ and received this mail, it does includes a schematic diagram, but in this digital age, I didn't think posting the diagram was useful.
Kirk Bayne

I opted out of the digital age (!) and have a bunch of the Motorola chips - so I'd be happy to see it if you'd care to post it. Thanks.

 

[par4ken]

How was it made variable in the S & IC?
I wonder how Fosgate did the stereo to surround in the Tate 2 decoders. The Sansui synth mode and the Fosgate both offered nice 270 deg wrap around. But the Fosgate sounded a little bit crisper in direction but the Sansui also enhanced center back which the Fosgate SQ method did not. The Sansui approach gave a bit more depth than Fosgate's.

In the SQ world there's no problem that can't be fixed by yet another 90 deg phase shift...

Basically you would just take the front and back decoder outputs feed them to each end of a pot, the output would then be a mixture of front and back, at each extreme pot setting it would give (normal SQ) with the pot fully counter clockwise and full enhancement with it fully clockwise.
This works in the Audionics decoder (check the matrix schematic), likewise it's how you build an encoder with a synthesize setting. Left front and right front are converted to left back and right back while center front stays put.
I'm sure that Fosgate decoders did the same thing but didn't make the effect variable.

 

[Sonik Wiz]

Basically you would just take the front and back decoder outputs feed them to each end of a pot, the output would then be a mixture of front and back, at each extreme pot setting it would give (normal SQ) with the pot fully counter clockwise and full enhancement with it fully clockwise.
This works in the Audionics decoder (check the matrix schematic), likewise it's how you build an encoder with a synthesize setting. Left front and right front are converted to left back and right back while center front stays put.
I'm sure that Fosgate decoders did the same thing but didn't make the effect variable.

Didn't you do an article for the MCS Review that showed a DIY circuit for type of stereo to SQ surround? Or was that how to make an SQ encoder from a decoder chip?

 

[kfbkfb]

Here's the Schematic Diagram + Component List for the "Stereo-to-Quad Enhancer Encoder"

(I'm using a landline phone and had Dial-up until AT&T shut it down 2020-11, so I'm still technically in the analog age too )


Kirk Bayne

 

[Soundfield]

Here's the Schematic Diagram + Component List for the "Stereo-to-Quad Enhancer Encoder"
(I'm using a landline phone and had Dial-up until AT&T shut it down 2020-11, so I'm still technically in the analog age too )
Kirk Bayne

Thanks very much indeed - Very simple, I might give it a go!

 

[Marcsten]

This is actually how the Space and Image Composer enhanced stereo, but by doing it directly rather than the stereo signal having to be pre-encoded. By swapping the front and rear decoder outputs and reversing phase we can produce this form of enhanced stereo.

As I was reading Ben's paper, I thought the same thing - that the S and IC was the same in enhance, since this is how I play back almost all two channel source in my system. You get the extending of the stereo field, and out of phase info such as ambience is sent to the rears. I have used this for decades now and its my favorite enhanced mode. I should say that I have a QSD 2 but not the famous QSD 1 enhancement, so can't compare them.

 

[par4ken]

Yes, I did a DIY article about building an encoder from a basic (non logic) decoder. I figured that one out myself from looking at the decoder output phasor diagrams and what was required as an encoded signal.

I think I did do another article adding forward oriented encoding and stereo synthesis, which are essentially the same thing. My inspiration for that came from "Quadraphile" and information contained in the US patents. I found out about the patent information from "Sonik's" article "Uncle Sam's Department of Quad"! A stereo to SQ "Enhancer or Synthesizer" similar to the schematic shared by Kirk was published in an electronics magazine (I forget which one) but that article contained some errors (in the description I think), I didn't understand how it was supposed to work at the time, because of those errors.

I know that I constructed several encoders for people that included normal, internal, forward oriented, and stereo to SQ synthesis. I used a Quad pot so that the enhancement effect could be continuously adjusted.

 

[kfbkfb]

Skimming the patent, I didn't find any mention of how the "enhancement encoded" stereo sounds in mono (I listened to FM with a mono clock radio until I got a battery powered headphone only FM stereo radio in 1984).


Kirk Bayne

 

[par4ken]

The enhancement mode works the same as the forward oriented encoder. It uses the back inputs to the encoder, because of the added phase reversal the centre comes out in phase in the encoded output. The stereo signal is now converted to an SQ encoded one with left encoded as left back right encoded as right back while centre remains centre front.

 

[par4ken]

Here is a copy of the final version of my SQ Encoder. With the quad section pot turned all the way down the encoder functions as a regular SQ encoder. You can switch the top decoder board in or out with the N/I (normal/internal) switch. In the internal mode the phase reference network is not used (top decoder board bypassed) which randomises the phase relationship between the front and rear and helps prevent cancellations that might otherwise occur. It has also been used to simulate height I guess that's why the full centre position is referred to as center top on the Audionics test record.

To use as a forward oriented encoder the front and rear inputs have to be reversed and the pot turned all the way up.

To use to synthesise SQ from stereo, only the front inputs are used. With the pot turned up a left input is encoded as a left back and right as right back, while center front stays upfront. , 270° enhancement. The other extreme produces no enhancement 0° . With the pot dead centered we get 180° enhancement. So you can contiguously adjust the effect from zero to full using the pot. The "Stereo Enhance" function of the S&IC works the same way but does it directly without the pre-encode step.

The Radio electronics article shows the enhancement circuit using a switch instead of a pot but that only give you 180° and 270° settings. The article text is brutal I didn't understand it when I first read it. The author is talking about phase modulation (confused with Sansui?) and makes the odd statement that if connected to an SQ or other decoder that it may make a strange sounding output, then recommends removing the output buffer transistors, if that happens? It should be OK to remove them but why (how on earth) could they cause a strange sounding output? I've included scans of my encoder. M1 and M2 are SQ decoder modules, using the Motorola MC1312 Decoder IC. Also a link to the Radio Electronics Oct 1975 edition with the Synthesizer article.

https://worldradiohistory.com/Archive-Radio-Electronics/70s/1975/Radio-Electronics-1975-10.pdf

 

[par4ken]

I just wanted to add a brief explanation of the Synthesizer/ Encoder circuit just in case I lost anyone.

If you input a stereo signal into an SQ decoder (basic non logic unit) the rear outputs contain an SQ encoded version of the inputs. Left and right are mixed in quadrature (90°). Because encode and decode are inverse operations we just have to swap the left and right back outputs, to make a proper encoder. That was the basis for my original encoder design.

To make a Synthesizer or a Forward Oriented encoder you have to invert the phase of one of the decoders back outputs. That will put center signal back in phase so that it will be able to be decoded as center front, rather than center back.

Front encoding in SQ is simply stereo, so for a complete encoder the left and right are simply passed on and mixed with the decoders back outputs. The front signals can be run through their own all pass network (another decoders front outputs) for normal encoding, or directly for internal encoding.

For a synthesizer only we require just one decoder. We use the back outputs with left and right swapped, then phase invert the decoders left back output. That process by itself produces a 270° SQ encoded signal. If we were to use the decoders front outputs instead we would get just plain stereo. If we add together equally the decoders front outputs combined with the (phase inverted) back we will get 180 ° enhancement. That is to say that a left signal will decode from the left side and the right from the right side. The same effect that you get with QS/Involve. By using a 4 section quad pot connected to the decoders outputs front and rear we can continuously vary the amount front vs rear output and thus control the mix from stereo through to 180° (at mid setting) through to 270° at the extreme stetting. The pot sections have to be wired oppositely for front and rear or else it would just become a level control, instead it is more like a balance control in operation.

 

[Sonik Wiz]

Thanks for the further explanation of your encoder/synthesizer circuit.
You said:

I know that I constructed several encoders for people that included normal, internal, forward oriented, and stereo to SQ synthesis. I used a Quad pot so that the enhancement effect could be continuously adjusted.

Which makes me curious why/how are you & those others using the circuit? Just for stereo thru SQ? Do you do much SQ encoding from discrete? And you have the perfect stereo synthesizer in you S & IC.

 

[par4ken]

I haven't done much encoding for years. With digital discrete technology there is little reason to encode anything. I thought of getting the Involve encoder to try it out but again have no real need to encode anything.

I did try encoding with Audition both SQ and QS and it worked out well but was a lot of work. Someone else wrote scripts for that purpose, I would likely use the script if I wanted to encode anything. I also found an eBay bargain a circle surround encoder that I was playing around with but it only does L,R,C,S like DS.

Originally I used my encoder to make SQ encoded cassettes from my reel tapes and Q8's.

I built and sold a several encoders, one went to Fizzy. I sold my last one (the version with the adjustable synth) many years ago. I had intentions of constructing a new one using the 8-pole filter of the S&IC and collected up many close tolerance capacitors. I was going to parallel then to make up the required values. The Audionics used specially made capacitors in rather odd values. I never got around to that, Now I have three composers so I could rig up a couple up to do encoding, if I wanted to.

Because the S&IC has it's own syth mode I have no need of using a circuit like that of the RE article. It would be good for someone with a descent SQ decoder like a Lafayette, to get stereo enhancement with some logic enhancement behind it. From what I recall all those seventies SQ decoders used their logic for SQ only, RM/Stereo enhancement was all done passively.

 

[kfbkfb]

Anyone have a gain riding only SQ decoder to try with the CBS Labs SQ stereo enhancer (and maybe an SM decoder too), just to see how gain riding quad synthesis vs. SM SQ quad synthesis vs. SM QS quad synthesis sounds?


Kirk Bayne