SQ Shadow Vector Soundfield Mapping

[fredblue]

Hey Folks, I wonder if someone could help me get an SQ channel test file so that I can test a decoder that I am working on. I think I have an issue with the decoder and I need a known good test file. Any help would be greatly appreciated.

Blaine

i might be able to help, i'll be in touch soon

 

[bigbillquad]

Hey Folks, I wonder if someone could help me get an SQ channel test file so that I can test a decoder that I am working on. I think I have an issue with the decoder and I need a known good test file. Any help would be greatly appreciated.

Blaine

Hi. BJMarsh

I think Sonic Wiz may be able to help you with that
BBQ...

 

[Sonik Wiz]

Hi. BJMarsh

I think Sonic Wiz may be able to help you with that
BBQ...

I have QS test tones ready to go but no SQ. I could probably whip them up when I get home.

 

[fredblue]

I have QS test tones ready to go but no SQ. I could probably whip them up when I get home.

i sent Blaine some already

 

[BJMarsh]

Thanks to all of you. Did some quick tests last night and found the decoder working fine on the SQ tones. Some kind of issue with the source LP that I am working with- DSOTM. Will dig in deeper

 

[Sonik Wiz]

Thanks to all of you. Did some quick tests last night and found the decoder working fine on the SQ tones. Some kind of issue with the source LP that I am working with- DSOTM. Will dig in deeper

Sounds like your making progress. May I ask what type/brand decoder uou using?

 

[BJMarsh]

Pure software implementation Lynn Olson’s Shadow Vector. Once I get back to the computer, I’ll explain some of the issues and observations if anyone is interested. At this point, I am thrilled with how well it works. It’s way better than I dared hope.

 

[Lynn Olson]

Good for you, BJMarsh! Wishing you all the best from Colorado!

I’ve been in “idle” mode for the last several weeks, as it has dawned on me that we’ll be self-isolated most likely through early 2021. (The earliest possible date for a vaccine through the Kaiser healthcare system, although I have asked to be signed up for early trials.) Colorado’s performance has only been so-so, nothing comparable to Australia or New Zealand, which has been orders of magnitude better.

The two ways of looking at local Colorado stats is that death rates for the age 70 to 79 cohort is 18% for those with confirmed CV19 (horrifying), or looked at another way, only 1 in 1000 Coloradans in that same age cohort has died thus far (not quite as frightening). Most of the deaths were in the initial tidal wave in mid-April; I surmise the recent decreases are due to new clinical procedures in the hospitals as doctors figure this thing out, as well as a threefold decrease in new cases. (Nowhere close to zero, though.)

Still, no reason to go out without a mask, keeping social visitors out of our home, and keep taking 600 mg/2x day NAC and 3000 IU of Vitamin D3. No shopping, restaurants, bars for the foreseeable future, just at-home deliveries. No visits to local hifi friends, either. Maybe next year.

So I’ll be on the sidelines, watching from afar, and hoping you have good results. Since the state of quad synthesis on mass-market receivers is so woeful, I think a QS/RM/EV4 dynamic matrix is the best way forward, with full SQ for legacy quadraphonic recordings.

(Although this thing occasionally scares me witless, as my doc and dentist have both told me, I’m doing everything correctly to minimize exposure, and I’m in very good health for my age group.)

 

[GOS]

Good luck staing healthy! Yes, those numbers are a bit daunting....geez.

 

[BJMarsh]

Thanks, Lynn. Take care of yourself. I certainly appreciate your efforts educating us (me) on your invention and matrix encoding/decoding in general.

 

[gene_stl]

For us oldies but goodies this thing is very scary.

 

[Sonik Wiz]

For us oldies but goodies this thing is very scary.

Yup I live in the hardest hit Kansas county (Johnson) & across the next state you live in most problematic county in MO being St Louis county. Strange days have found us. Take care.

 

[BJMarsh]

Here's a deep dive into the Shadow Vector directional sensing. There are 6 fullwave precision rectifiers (using op-amps as precision rectifiers) for each cardinal point of SQ ... LF, CF, RF, LB, CB, and RB. There's a bit of lowpass filtering for the output of each fullwave rectifier, followed by 3 differential amplifiers for LF/RF, CF/CB, and LB/RB. There's a single tight AGC loop with at least 40 dB dynamic range wrapped around the array of precision rectifiers, lowpass filters, and differential amplifiers. The logic only relaxes for low-level inputs that approach the noise floor of an LP, around -50 dB or lower.

Following that, the three bidirectional (+/-) control lines go through three identical nonlinear circuits (in practice, an array of resistors and diodes) that shape the control response so the variable-gain amplifiers track the Scheiber sphere precisely. The nonlinear shaping circuit was determined by connecting a pair of analog oscillators 1~5 Hz apart, which generates a spinning-phase signal that goes from LF to CF to RF to CB to LF again. By observing the output of LB and RB on one axis of the scope, and using the LF/RF control line on the other axis, the nonlinear circuit can be trimmed so the each of the back channels precisely follows the ideal SQ pan-locus. LF to CF to RF to CB to LF isn't a standard SQ pan (in fact it is an ideal QS pan), but it calibrates the response of each back channel exactly, so separation is maintained as the signal sweeps across the front, and then smoothly moves back to the default LB or RB position as the signal sweeps across the back quadrant. Just to check, the spinning LF/RF input signal can be re-matrixed into LB/RB, or CF/CB, to confirm the other axis of the decoders have symmetrical responses.

Hi, Lynn. Hope all is well. I was trying to repeat the calibration of the shapers that you referenced using signals generated in software. In the left channel, I had a 300Hz sine wave and the right channel I had a 301Hz sine wave. I ran these through my SV decode algorithm and did not see the decoded output sweep from the expected LF->CF->RF->CB->LF. I must not be understanding what you were saying in the above. If you have a moment, can you clarify? Blaine

 

[Lynn Olson]

The spinning-phase encoded signal should be CF to LB to CB to RB to CF again (or in the other order). It will never enter LF or RF, since the proportions of Left and Right are always 50/50. The only thing that changes is phase. A signal like this won’t make much sense to a QS decoder ... it’ll just move from CF to the middle of the room to CB and back again. (Yes, I got the original LF -> CF -> RF -> CB -> LF pathway wrong. Oh well, it’s been a while since I juggled a Scheiber sphere in my mind. Sorry about that.)

Out of curiosity, what emerges from the SQ software decoder when the spinning-phase signal is sent to it? Slowed down, of course, but a 1 Hz phase difference is a very slow movement easily within the range of a competent decoder.

Psychologically, I’ve come to terms that I’ll be under quarantine until 2021, with lots of nervousness if a repair guy shows up (wearing a mask of course), but so far we’ve had a dishwasher replaced by a local plumber, and two routine dental visits, and everyone has been very, very careful around us old folks. Plus our overall health is actually very good ... surprisingly, I’ve lost 25 to 30 lbs since this all started, by not eating at restaurants any more, and very little late-night snacking.

 

[BJMarsh]

Thanks for helping out. The output from the decoder sure looks bizarre but I need to set up some software test points properly. I also misread what you were saying so with the info you supplied, I am off and running. Once I check the balance of the "circuits" I will do a screen shot of the decoder output. I am in the same age range as you (68 this year) so it is concerning. Getting outdoors is really nice. Fortunately, we have a farm that allows us to wander the fields.

 

[kfbkfb]

Re: all software Matrix (SQ) Quad decoder:

How about a "mild logic" decoder configured as follows:

Fixed 12dB LF to RF and LB to RB channel separation,
"mild" Front/Back logic (similar in concept to what was
used in the Sony SQD-1000 SQ decoder), but providing
a maximum of 12dB (logic assisted) F/B channel separation.

Reason for 12dB:
https://www.quadraphonicquad.com/fo...rround-master-sm-465.17227/page-3#post-171125
Please consider offering a "mild logic" SQ decode option
(my audio signal processing philosophy is to try to minimize
processing, I prefer Dolby B NR to Dolby C and S NR).


Kirk Bayne

 

[Lynn Olson]

Depends on what you mean by “mild logic”. If there is Sony SQD-2020 gain-riding (for corner enhancement), even a little will be noticeable as a flattening of dimensionality. I frankly think this was a big mistake on CBS’ part ... it’s just not a good way to build a decoder, and I think it materially harmed the cause of the SQ system and quadraphonic sound in general. Flat-sounding ping-pong decoders didn’t make any friends for quad, and fed the magazine-reviewer stereotype of gimmicky and high-fatigue sound.

In terms of a dynamic matrix, there are many ways of making it less aggressive. For one thing, it can be content-aware ... sharp transients get precise localizations, while diffuse localizations slow down the action, and don’t have to aim for 100% cancellations, since the sound source is diffuse anyway.

This is a subjective call on the part of the decoder designer, and I see no harm in providing the user with a “sharpness” knob that subtly adjusts the dynamic parameters of the decoder. If the user consistently prefers a moderate setting, that’s a great hint the dynamic characteristics of the decoder are irritating and unnatural-sounding ... time constants that are not right, overshoots are present, or the logic sensing isn’t following the music very well.

I found that conventional stereo recordings often had better “encoding” than many SQ-marked records. Could have been poorly encoded 4-2 mixdowns, could have been unwanted mike-to-mike crosstalk, any number of things. The number-one problem to avoid is high listening fatigue, which I think is the most serious problem with quadraphonic and/or surround music. Good decoders make all the difference here. What is desirable is not “separation” or cookie-cutter imaging, but an unstressed sense of natural space that is more realistic and open-sounding than 2-speaker stereo. That was always the aim of the Shadow Vector system, as well as the speakers I later designed.

That was probably the most surprising result of the Shadow Vector project ... listening fatigue was the most serious problem with quad sound, far more than stereo 2-speaker listening. Asymmetries that merely sound odd in stereo are actively distracting in quadraphonic, so measures to control artifacts have to be thorough. I really think this is where more attention should have been paid back in the 1970’s, but then, a lot of the early-transistor sound was pretty aggressive anyway, so it escaped notice at the time. Modern speakers have much more revealing tweeters, so things that were swept under the rug then are more clearly heard now.

 

[Arconada]

How about a "mild logic" decoder configured as follows:

Mild logic delivers most of the time better results. I never go further than 12 dB separation. Increasing the separation results in artifacts. As a matter of fact, the basic field recording setup, one omnidirectional mic combined with two figure-8 mics, results in a natural separation of 6 dB. I never tried what the result is when you diminish the mild logic to that level. Would be interesting, though. Thinking of it, it wouldn't work for SQ because the crosstalk is not to the adjacent channels, but to the opposing channels, front or back.

 

[MidiMagic]

Depends on what you mean by “mild logic”. If there is Sony SQD-2020 gain-riding (for corner enhancement), even a little will be noticeable as a flattening of dimensionality. I frankly think this was a big mistake on CBS’ part ... it’s just not a good way to build a decoder, and I think it materially harmed the cause of the SQ system and quadraphonic sound in general. Flat-sounding ping-pong decoders didn’t make any friends for quad, and fed the magazine-reviewer stereotype of gimmicky and high-fatigue sound.

This is a subjective call on the part of the decoder designer, and I see no harm in providing the user with a “sharpness” knob that subtly adjusts the dynamic parameters of the decoder. If the user consistently prefers a moderate setting, that’s a great hint the dynamic characteristics of the decoder are irritating and unnatural-sounding ... time constants that are not right, overshoots are present, or the logic sensing isn’t following the music very well.

The Front-Back logic in the SQ decoders helped to remove any ambience in the recording. It turned it down. So did all gain-riding systems. It took the Variomatrix to increase separation without destroying the faint nuances of ambience.

I found that conventional stereo recordings often had better “encoding” than many SQ-marked records. Could have been poorly encoded 4-2 mixdowns, could have been unwanted mike-to-mike crosstalk, any number of things. The number-one problem to avoid is high listening fatigue, which I think is the most serious problem with quadraphonic and/or surround music. Good decoders make all the difference here. What is desirable is not “separation” or cookie-cutter imaging, but an unstressed sense of natural space that is more realistic and open-sounding than 2-speaker stereo. That was always the aim of the Shadow Vector system, as well as the speakers I later designed.

This is why I do not like discrete quad. I want a contiguous space, not 4 pongs.

There were some stereo records that decoded into SQ that amazed me, because they were made before SQ existed. They were made with the Haeco CSG system to eliminate the "oversized center soloist" problem when a stereo record was played in mono. It applied a 90 degree phase shift between the left and right channels. to reduce the center-panned sounds by 3 dB.

 

[kfbkfb]

I just made up the phrase "mild logic" to try to distinguish
my idea/request from the old gain riding Front/Back only
logic as implemented in the Sony SQD-1000.

The fixed 12dB is based on both the Involve Audio findings
(corroborated elsewhere) and the SQ 10-40 fixed blend
decoder design for low cost SQ decoders.

Perhaps less than 12dB F/B separation would still result in
a good surround effect while further minimizing the
possible audibility of logic action artifacts.


Kirk Bayne