The Legacy of Dr Bauer

[Soundfield]

I believe that to be incorrect. Look at pan pot S5, it connects to both the forward and backward encoders, that's the only way it will work.

Regardless of the internal coupling of the two sets of pots via the one marked S5 the whole array still only seems to connect to the SQ Encoder via two four-way signal busses (one for the Forward and one for the Backward encoding module). This certainly ties up with the following description of the input arrangements of the 4211-

i.e. Two sets of four inputs for the F & B encoders and one set of four inputs for the 'normal/ standard' encoder (a total of twelve inputs).

 

[MidiMagic]

The problem is that any single SQ encoder makes mistakes for some panning selections.

Because SQ is a matrix with points of inflection in its encoding set of modulations, pannings between different speaker pairs require different phases on the channels. At most three pairs of speakers can be encoded correctly with one encoder.

SQ ENCODER	LF-RF	LB-RB	LF-LB	RF-RB	LF-RB	LB-RF
4-Corners Matrix	BEST	BEST	Odd	Odd	BEST	No
Acroperiphonic	BEST	BEST	Odd	Odd	Odd	Odd
Diagonal Split	BEST	No	Odd	Odd	BEST	BEST
Forward-Oriented	BEST	No	BEST	BEST	Odd	Odd
Backward-Oriented	No	BEST	BEST	BEST	Odd	Odd

BEST - Optimum encoding
Odd - Odd but usable encoding locations
No - Encodes in the wrong direction

The 4-corners encoder is the original SQ specification.
The acroperiphonic encoder applies no phase shifting at all to the front channels.

Note the following:

- All phase shifts other than 180-degree shifts add a psi component, which makes phase frequency dependent.

- When we use psi phase shifting, we use the phase difference between psi+0 degrees and psi+90 degrees.

- Adding an extra 90-degree phase shift to a signal adds an additional psi phasing. This makes the frequency-dependent phases on that signal different than the other signals in the encoder, so they do not sum and cancel properly.

- It is impossible to remove psi phase shifts from signals.

These are why it becomes necessary to have more than one encoder. There is no way to encode all positions correctly with one encoder.

 

[Soundfield]

As we've said on this thread previously - SQ was a mess!

 

[Sal1950]

Thank God that so many of the original discrete multich masters still existed to reissue on our DVD-A, SACD, BD, etc. A shame so many are lost but we can rejoice in what we have.

 

[par4ken]

As we've said on this thread previously - SQ was a mess!

With proper encoding (and Tate decoding) SQ was amazing. Sad it was largely discontinued after all the bugs were out! With the release "Serious Moonlight" in 1984, I for one was optimistic that more releases were to come but sadly they never materialised.

 

[kap'n krunch]

I personally have gotten some REALLY good results from AA3 scripts...so much so that my conversion from many moons ago (in 96/24!) of Santana's first LP was extremely close to the Japanese SACD released recently... of course there was some pumping here and there but the imaging was VERY close to the discrete master, and besides, the fidelity was great...of course the discrete SACD is the real thing but SQ was VERY close!

 

[Marcsten]

I confess a lot of the technical stuff in this thread is far over my head. What I get is something we have all known for several decades. In order to get more than two channels on a disc with only two walls, you have to compromise. Whether its the shortcomings of SQ, which as others have pointed out were nearly all overcome by the end, although things were pretty terrible at the start, or discrete which to this day don't work right much of the time due to the high demands of the ultra-sonic signal and the limitations of the analogue format, both had lots of problems. Its hard to say where things would have ended up had both formats continued to evolve and tech continued to improve. But with what we have now, both are flawed systems, although as kap'n krunch points out, at the end, imho, SQ was king. OK. Start flaming.

 

[kfbkfb]

For the then near future (late 1970s), (IMHO) we'd probably have had discrete quad on the Laser videodisc (from day 1, not having to wait till 1995 for 5.1 AC-3), the CED videodisc and the Compact Disc.

Even Beta and VHS could have had discrete quad by using more FM carriers with the depth/azimuth tech.

SQ might have been ok for stereo FM and stereo AM, since most radio listening is done in vehicles and a simple 10 40 blend (no logic) SQ decoder could have been used.


Kirk Bayne

 

[Owen Smith]

In the UK lots of radio listening is done in people's homes and similar places eg. garden sheds.

 

[kfbkfb]

http://www.radiodiscussions.com
I got my listening info from here (it's US centric though, I don't know much about radio listening elsewhere).

Discrete quad FM had an added noise penalty although not as much as stereo FM compared to mono FM.

(due to the audible [to me] FM audio processing, I rarely listen to FM at home, I sometimes listen to HD radio [less audio processing])


Kirk Bayne

 

[par4ken]

For the then near future (late 1970s), (IMHO) we'd probably have had discrete quad on the Laser videodisc (from day 1, not having to wait till 1995 for 5.1 AC-3), the CED videodisc and the Compact Disc.

Even Beta and VHS could have had discrete quad by using more FM carriers with the depth/azimuth tech.

SQ might have been ok for stereo FM and stereo AM, since most radio listening is done in vehicles and a simple 10 40 blend (no logic) SQ decoder could have been used.


Kirk Bayne

I never ever liked SQ with blend, it ruins SQ's greatest virtue. I know that without blend the vocals are everywhere at once, some say nowhere at once but that was always a minor concern for me. Years ago when I constructed a few SQ encoders using surplus SQ decoder boards they had the MC1313P chip on them (the 12V version of the MC1312P). With many/most if not all of them I replaced the chip with the MC1312P. Anyway I was going to use one of those decoder boards in the car (as it was originally intended for) I just never got around to it. I still have some of those 12V decoder chips though.

It always disappointed me the the CD was designed for Quad as well as Stereo but that feature was never used. Again my hopes were raised only be quickly dashed.

 

[kfbkfb]

For vehicles, the listener(s) are generally way off center, I don't know if SQ decoder logic system action would be audible under these circumstances.

The Involve Audio people found that more than 12dB channel separation isn't needed.


Kirk Bayne

 

[Owen Smith]

The Involve Audio people found that more than 12dB channel separation isn't needed.

So they claim. I haven't seen their evidence, and I disagree with them on this point.

 

[Sonik Wiz]

So they claim. I haven't seen their evidence, and I disagree with them on this point.

Fortunately in practice the Surround Master goes waaay beyond 12 dB separation. Some testing I did a while back bears this out:

https://www.quadraphonicquad.com/forums/threads/2020-sm-v2-testing.29536/
In olden days of matrix quad I think the MFG'ers reached for -25dB separation bragging rights because that was the common separation specs for a decent TT stereo cartridge. Certainly they were using test tones not records that would have been less in the real world, and they stated average or maximum separation, not minimum.

If not -25dB separation, then what is meaningful? There is precedent to the -12dB effect. I don't know how Chucky tested but I have a few booklets published in the UK by The NRDC, an office that helped promote the design of Ambisonics. Using Calrec Soundfield mics was the most common way to produce Ambisonics in the "B" format and then encode to UHJ. But the Ambisonic people wanted to make it possible to take regular discrete surround in PWM & convert to "B" format, and then to UHJ. It was pretty straight forward and if instead of going to UHJ, should you feed that encoded "B" format signal to a straight off "B" format decoder you get.... a totally symmetrical sound field with -12dB separation all around.

The Ambisonic tested this in focus groups with various types of music & ultimately came with saying " 12dB separation in any direction is sufficient for sharp localization." They didn't say it was the maximum needed or that any more separation was un-needed, just under those conditions it was enough.

One of Involves main concern is at the encoding side where their Smart Encoder maintains a variable but maximum width of 12dB between the output chs. Being able to produce a stereo compatible program source that gives good width in stereo & good surround sound is their goal.

And how people listen is important as well. Music through headphones is going to sound pretty wide with 12dB separation and if your listening to Bluetooth party speakers or soundbars talking about minimum separation is funny enough. In a more dedicated environment I have a very acoustically dead room. I'm gonna say I using a mono signal between the front speakers & dropping the right front, I am going to detect a change increasing above 12dB separation. On the other hand I had a friend in New Mexico with tile floors & stucco walls. I have a feeling room acoustic would seriously degrade the signal before -12dB sep is reached.

So you raised an excellent point to consider, IMO. The question of that -12dB criteria depends on many things & tough to stick with a blanket statement one way or the other.

 

[par4ken]

For vehicles, the listener(s) are generally way off center, I don't know if SQ decoder logic system action would be audible under these circumstances.

The Involve Audio people found that more than 12dB channel separation isn't needed.


Kirk Bayne

I would imagine that a full logic decoder would of sounded fine in a vehicle but it would of required (back in the day) that FM stations broadcast an SQ encoded program or that your cassettes were of SQ encoded material and that azimuth errors didn't ruin the encode. In that situation a synth/enhancement mode would also be desirable/necessary for regular stereo.

I think that with 12dB or more direction will be readily apparent, but more separation is desirable. Lower separation has the effect of pulling the speakers closer to you.

Motorola never made any 12V logic chips, so we don't really know how well a car decoder would sound. One could experiment but you would need a power converter to step up the voltage.

I remember waiting patiently for the Fosgate car Tate SQ decoder that never materialised, and I was willing to pay top dollar for it at the time.

 

[Electric Moo]

Well, I know who Dr Bauer is now, after doing some research/reading.

noob!!!

 

[Sonik Wiz]

Well, I know who Dr Bauer is now, after doing some research/reading.

noob!!!

By all means, when you have the time, google/research the Father of (matrix) Quad, Peter Scheiber.

 

[BJMarsh]

And Lynn Olson. Best dang SQ decoder - IMO. Check out the SQ Shadow Vector thread.

 

[BJMarsh]

Where is Lynn? Haven’t head a peep in months!

 

[MidiMagic]

I preferred the 10-40 blend for SQ because it was better for hall ambience,