Old time four channel greetings

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Status
Not open for further replies.


Found this a few months ago, it has fairly good directionality listening with my Hafler/DynaQuad speaker matrix based surround sound system (2 Polk Monitor 4 front speakers, Polk computer speakers for the surround channel).

(Dolby Pro-Logic 2 music mode decoding doesn't sound as good with this Matrix H encoded content)


Kirk Bayne
 
The mods that I've seen didn't look too bad. Articles in Wireless World May 1977 and HiFi News & Record Review May 1977. I even built the phase shift circuit just incase I ever came across any Matrix-H, but I never did!

Were there more comprehensive mods to get better H decoding? Was an actual proper Matrix-H decoder ever built?
 

Attachments

  • Matrix H adaptor.pdf
    338.4 KB · Views: 0
Last edited by a moderator:
Currently I am searching for Dolby Atmos content that sounds as exciting as the old Q8 or CD4 did. Or maybe I am remembering through rose tinted ears.
Welcome to QuadraphoncQuad!
What is it about Atmos your unsatisfied with?
There are ton's of Atmos recordings that are extremely exciting.
Tell us about you system setup
 
Around 1972 everyone was selling "Hafler" type quad adaptor boxes. I bought mine from "The Record Club of Canada". Radio Shack and almost everyone else were selling them as well. The lower cost stereo systems of that time almost all boasted "quad" sound with "Hafler" type speaker connections for the back speakers. I suspect that marketing ploy confused people into thinking that they had a real quad system.

Actually, they did have real quad. I have that Vanguard record which was encoded for either the Hafler diamond or for the DynaQuad. And by moving the speakers, they can get the same images (to the ear) the "real" decoders make for EV-4, QS, and Scheiber (if you can find a recording). And I made recordings myself in it.

What the system(s) did was to extract the random out of phase from the stereo source and played it via the single or double speakers placed in the rear.

Often, a low value wire wound resistor, or wire wound variable control, was placed between the negative lead of both speakers and earth, thereby introducing a small amount of the audio from the fronts to create a small degree of separation on the rears.
This has nothing to do with RM or anything else. All of the matrix systems were not so easily decoded, you would obviously obtain some sound out of the rears, but to say it is decoded is going too far.

Not true. That resistor gives a valid decoding angle. Adjusting the resistor changes the decoding angle.

uq-1-v.png

I built this passive decoder that decodes any RM style matrix.
- Adjusting the WIDTH control adjusts the front decoding angles.
- Adjusting the DEPTH control adjusts the back decoding angles.
- Adjusting the AUTOVARY control provides separation enhancement.

I have two of these operating in my house in two rooms (in addition to the main system). It can be set to decode accurately Scheiber, QS, EV-4, DQ, and DS. It can also partially decode SQ, EV-U, H, and UHJ.

Playing my DQ, EV-4, QS, and DS records produces the same images the real decoders produce.

I listen to the radio in the kitchen with one of these. Many of the ads and some CDs are obviously in Dolby Surround.
 
Matrix H was undecodable by any other means than by a matrix h decoder, which never really existed for the consumer. There was a kit available but as it was just based on qs its performance was sub standard.

I can use phono stylus vector maps and the Poincare sphere to show the intercompatibility of these systems:

I will use the Hafler diamond and DS to explain the system, because they have the same diagrams.

qim-ds.gif
qsp-ds.gif


The left diagram shows the phono stylus motion vectors for both Dolby Surround and the Hafler diamond.

Left (cyan) and right (red) are the 45-45 stylus vectors of the Westrex system.
Front (olive) is the mono record vector, center in stereo, and front in DS and Hafler.
Back (violet) is the back channel in both DS and Hafler. It is L and R out of phase.
Ignore the brown and black for now.

On the Poincare sphere, each possible 2-channel modulation is on the sphere surface.
Left is on the far side of the sphere (large dot). Right is on the near side (small dot).
Horizontal modulation is on the right limb. Vertical modulation is on the left limb.

Scheiber, RM, and QS are interleaved between the DS and Hafler modulations

qim-rm.gif
qsp-rm.gif


LF is green, RF is yellow, LB is blue, and RB is magenta.

EV-4 has wider front encode and narrower back encode separations:

qim-ev.gif
qsp-ev.gif


Also, the EV decoder has different matrix parameters from the encode ones:

qim-evd.gif
qsp-evd.gif



The DQ encode parameters are only slightly different from the EV ones.
The DQ decode parameters are quite different:


qim-dqd.gif
qsp-dqd.gif


Since the record groove can contain any of these, playing one back through a different decoder will place the sound image found in the encode diagram in the same location in the diagrams in the decode diagram, with the3 image shifted as the decode diagram places it.

My passive decoder can decode any of these.

Continued next post.
 
Continued from previous post.

The top limb of the Poincare sphere shows clockwise stylus rotation.
The bottom limb of the Poincare sphere shows anticlockwise stylus rotation.

I have already shown the QS/RM diagrams. I repeat them to compare them:

qim-rm.gif
qsp-rm.gif


Now I add the BMX matrix (and the baseband for UD-4):

qim-um.gif
qsp-um.gif


Notice that the BMX locus on the Poincare sphere is rotated 90 degrees from QS.

Notice how the left-right separation is the same, but the front back info is quite incompatible. Each plays center front and center back material from the other system from all four speakers equally.

Now I put in Matrix H. Matrix H is half-compatible with QS and half-compatible with BMX. Both will have reduced separation in the front-back direction when playing H. Matrix H will also play either QS or BMX with reduced front-back separation.

qim-hm.gif
qsp-hm.gif


Notice that the H locus on the Poincare sphere is rotated 45 degrees from QS.
Notice that the H locus on the Poincare sphere is also rotated 45 degrees from BMX.

There were two H encoders. One put the L and R on the L and R channels. the other put them at the other L and R dots on the sphere. The stylus diagram shows the one with the other dots.

Now for UHJ:

qim-uj.gif
qsp-uj.gif


UHJ is partially two-way compatible with QS, BMX, and H.

These diagrams were derived from the actual matrix equations.

Now you can see the differences.

Note that BMX, H, and UHJ all have mono compatibility for center back.

Last, I include SQ to show how incompatible it is with the other systems:

qim-sq.gif
qsp-sq.gif


However, SQ can play BMX by moving the speakers to a diamond pattern:
LF moved to LC. RF moved to RC. LB moved to CF. RB moved to CB.
Also reverse the phase of the LB speaker.
 
Last edited:
Matrix H was undecodable by any other means than by a matrix h decoder, which never really existed for the consumer. There was a kit available but as it was just based on qs its performance was sub standard.

Then what is this?

20150811_191440.jpg


There is an H position.
 
In my case, I'm just happy to have a "free lunch" of surround sound by just using my old computer speakers, a cannibalized mini headphone jack Y adapter connected to L+ and R+ and a mini headphone extension cable.

Trying out my new (black friday special) 32" Philips HDTV w/Madonna music videos DVDs (soundtrack in stereo or 5.1), the 5.1 downmixed to Dolby Surround by my Sony player and "decoded" w/Hafler\DynaQuad sounds really good.


(maybe most posts in this thread should be moved to a new thread in the Matrix section entitled something like:
"Hafler/DynaQuad decoding of other Matrix Systems")


Kirk Bayne
 
I may be wrong, but wouldn't you need to be able to broadcast Atmos in a discrete platform? SQ, and QS, were matrixed, and could be broadcast in the same manner as stereo. Nothing special needed to air those records in quad.

There is a way to encode sounds coming from anywhere in a sphere surrounding the listener with matrix techniques. The Atmos material could be fed to such an encoder.
 
https://www.radioshackcatalogs.com/flipbook/1972_radioshack_catalog_ver1.html(page 26 - The Quatravox device + speakers ~$49)

https://www.radioshackcatalogs.com/flipbook/1975_radioshack_catalog.html(page 22 - Quartette device $6, + speakers ~$25 [sold through 1978])
aside:
(page 18 - Radio Shack quad receiver w/CD-4, I didn't know Radio Shack ever sold CD-4)


I visited many Radio Shack stores during the Quad era, I never saw any of these speaker matrix devices hooked up/demonstrated (only their Stereo-4 clone decoder).


Kirk Bayne
 
I think the point that Midi is trying to make (and he brings it up all the time) is that moving your chair, let's say toward the back a bit will position you back into the centre of the mix while listening to (lets say) EV-4 decoded via QS. Not too practical nor necessary in my view but I do understand what he is saying.

Taking it a step farther moving your speakers around the room to compensate for decoding via the wrong decoder is a novel idea but not something that I would care to do!

I agree that decoding of UHJ will not be correct without an Ambisonic decoder.
 
Here in the USA, it was mostly the SQ matrix (QS struggled without major record labels supporting it), matrix H was unknown.

Nimbus records (CDs really) used UHJ to encode a large number of their releases, my first contact with the UHJ matrix was one of their classical singles I bought in the late 1980s, I have only ever heard (matrix systemwise) no logic SQ and QS, Dolby Pro-Logic (1 & 2) and Hafler/DynaQuad (also, several other matrix system appeared in the USA in the 1980s &1990s - Circle Surround, Shure StereoSurround, (unnamed) Telarc Surround sound).

Hafler/DynaQuad can do a fair job of extracting some sort of surround sound from all matrix system.


Kirk Bayne
 
Scooby Don't - I'm interested in the reasoning behind developing Matrix H (the CBS SQ matrix claimed excellent mono and stereo compatibility too).

I know that SQ (evaluated with the CBS gain riding decoder) didn't do well in the 1974 BBC localization tests, but during the time that Matrix H was developed, there were variable matrix SQ decoders that localized sounds much better.


Kirk Bayne
 
Status
Not open for further replies.
Back
Top