BBC Matrix H

[omega4]

Can anyone on the QQ Forum offer any advice on a suitable decoder for the BBc Matrix H quad system. I have uncovered a number of tracks recorded by the BBC with various artists (Steve Gibbons Band, Colisseum II, The Stranglers to name three) and broadcast in 1977. Considering the age of the cassette and the budget tuner that I used they are of remarkably good quality. I am going to give Dolby Pro-Logic II and DTS Neo 6 a try, and also SQ. I suspect though that they will not be accurate in decoding. Would an Ambisonics decoder work? If so where on earth do you find them? Any advice will be very welcome.

 

[quadtrade]

Can anyone on the QQ Forum offer any advice on a suitable decoder for the BBc Matrix H quad system. I have uncovered a number of tracks recorded by the BBC with various artists (Steve Gibbons Band, Colisseum II, The Stranglers to name three) and broadcast in 1977. Considering the age of the cassette and the budget tuner that I used they are of remarkably good quality. I am going to give Dolby Pro-Logic II and DTS Neo 6 a try, and also SQ. I suspect though that they will not be accurate in decoding. Would an Ambisonics decoder work? If so where on earth do you find them? Any advice will be very welcome.

I have a Denon decoder for that material, though it is in Oregon at the moment, and i am not. Lets chat, maybe we can figure this out!

 

[omega4]

I have a Denon decoder for that material, though it is in Oregon at the moment, and i am not. Lets chat, maybe we can figure this out!

Many thanks for your reply. I had a recollection that Denon made a matrix decoder but I didn't realise that it decoded Matrix H. I am definitely interested although I am in the UK, but when you have time maybe you could send me a private message.

Do you know any more about Matrix H or indeed the development by the BBC, Matrix HJ? I must also say that the QQ Forum is fantastic. I spent many years in the quad wilderness here in the UK not knowing anyone else with four speakers (on the same system!) Since the internet came along I find that I am definitely not alone. I see from your profile that you are a founding member - so thanks to you and your fellow founding members.

 

[The Quadfather]

Cantares makes an ambisonic decoder, though I don't know how good it is or whether it is suitable for Matrix H or HJ.

The Quadfather

Many thanks for your reply. I had a recollection that Denon made a matrix decoder but I didn't realise that it decoded Matrix H. I am definitely interested although I am in the UK, but when you have time maybe you could send me a private message.

Do you know any more about Matrix H or indeed the development by the BBC, Matrix HJ? I must also say that the QQ Forum is fantastic. I spent many years in the quad wilderness here in the UK not knowing anyone else with four speakers (on the same system!) Since the internet came along I find that I am definitely not alone. I see from your profile that you are a founding member - so thanks to you and your fellow founding members.

 

[Malcolm2010]

Try using a QS decoder. matrix H was derived from Sansui's QS system, and it should play well.

A Quote from

http://www.transdiffusion.org/rmc/features/width.asp

"This begs the question of whether or not there has ever been any other surround sound broadcasting in the UK, and the answer is yes, there has. At around this time the BBC also began looking at the ‘matrix’ encoding systems which attempted to encode information for 4 speakers into the two available stereo channels and the BBC made a number of broadcasts using a system they called Matrix H, which was a very slightly modified version of Sansui’s QS system. Very few Matrix H decoders were in use but despite that almost an entire season of Promenade Concerts was broadcast in this way. "

An informative article.


Malcolm

 

[ingresman]

Sorry to drag up an old thread.
Have a look at https://www.quadraphonicquad.com/forums/showthread.php?t=6495
If someone can modify the Audition QS script to add the extra 60degree shift you should be able to get a good decode.
The matrix-h I have or had ! is
Pink Floyd WYWH
Steve Gibbons - various
The Motors - various
Jethrow Tull sessions perhaps minstril in the gallary
Queen sessions from News of the world (it's late, get down make love)
Jack Bruce in concert
All on cassette tape and probabbly too phasey to decode, but I might give it a go !

 

[Lucanu]

Hi
you can simply apply a 60 deegre shift to left channel on your samples and save it as front.wav/rear.wav.
Now you can get into the script.

 

[Malcolm2010]

For lots more information on Matrix H see my post here :-

https://www.quadraphonicquad.com/forums/showthread.php?t=3783

and here for the phase shift info here :-

https://www.quadraphonicquad.com/forums/showthread.php?t=3793

Malcolm

 

[tryphon]

The original Integrex decoder designed by Michael Gerzon has a decoder matrix for the BBC HJ system.

 

[grill]

Hello Quad Heroes,
I'm really interested in WYWH BBC Matrix H conversion. Has anybody done it?

 

[oxforddickie]

Hi Grill

As a member of SS i'm sure you have seen my posting regarding tests on improving Matrix H decoding. I have just released the test files to those who replied to my request for people to do some listening tests. Contact me there if your interested in being involved. if possible, it would be good to get good copies of your tape to add to the blogs

Richard (OD)

 

[MidiMagic]

Matrix H is halfway between UMX and QS.

BHJ Ambisonic is based on Matrix H.

RM does a partial job. There is less front to back separation.

You can fake a decode of H with an RM decoder, an SQ decoder, and 6 amps and speakers:
Set up the RM decoder in the usual way. Then add two more speakers, with SQ left back between the front speakers, and SQ right back between the back speakers. Your ears will mix them into H.

 

[oxforddickie]

Matrix H was the first "public" matrix system and was soley a BBC creation, which then became HJ when they started to work with Nikon/Denon (UD4/UMX etc). This then morphed into 45J then Ambisonics which is when the NRDC took over devolpment and Nibmus became involved.

Why fake H when it's possible to decode it. This is what my tests are for, to improve on the somewhat incorrect idea that all you need to do is shift the right channel by 60 degrees (which, in BBC words at the time, was only a "compramise" anyway)

Also, using SQ to decode Matrix H is a non starter, they are poles apart. As you know, it is very close to QS, hence the BBC using modified QS decoders in their tests.

OD

 

[MidiMagic]

Matrix H was the first "public" matrix system and was soley a BBC creation, which then became HJ when they started to work with Nikon/Denon (UD4/UMX etc). This then morphed into 45J then Ambisonics which is when the NRDC took over devolpment and Nibmus became involved.

Why fake H when it's possible to decode it. This is what my tests are for, to improve on the somewhat incorrect idea that all you need to do is shift the right channel by 60 degrees (which, in BBC words at the time, was only a "compramise" anyway)

Also, using SQ to decode Matrix H is a non starter, they are poles apart. As you know, it is very close to QS, hence the BBC using modified QS decoders in their tests.

OD

This was intended for people who have vintage decoders, but no fancy computer audio system.

Matrix H is halfway between QS and UMX:

QS uses inphase for the front and antiphase for the back.
UMX uses the clockwise circular modulation (left leads right) for the front, and the anticlockwise modulation (right leads left) for the back.

H uses an elliptical clockwise modulation for the front, with the inphase being the major axis.
H uses an elliptical anticlockwise modulation for the back with the antiphase being the major axis.

My suggestion uses the SQ decoder to provide the clockwise (SQ left back) and anticlockwise (SQ right back) modulations. I do not use SQ logic when doing this. Placing the SQ speakers at center front (clockwise) and center back (anticlockwise) mixes these modulations with the center front (inphase) and center back (antiphase) modulations from the RM decoder in your ears. You do have to reverse the phase of the clockwise speaker when using a standard SQ decoder.

Using just an RM decoder reduces the front to back separation when playing an H recording.

 

[MidiMagic]

I found an interesting list of the matrices evaluated by the BBC in their online archives:

http://downloads.bbc.co.uk/rd/pubs/reports/1974-29.pdf

Matrix A was QS, with no logic

Matrix B was SQ with no logic.

Matrix C was QS VarioMatrix

Matrix D was SQ Full Logic

Matrix E was a strange tetrahedral encoding proposed by Peter Scheiber. (Not the one that is close to EV Universal)

Matrix F was UMX

Matrix G was a compromise between QS and UMX

Matrix H was a refinement of Matrix G

 

[oxforddickie]

if you are going to try and prove something, at least put the correct information down so others can learn correctly"

UMX should read BMX

and don't tell me it's the same, UMX was the "Umbrela" name and consisted of more than one part, BMX was the name of the Matrix

One does have to remember while reading this document, that the logic decoders used were pretty primative, and were soon improved in nearly every aspect.

There is also something your missing in you way of decoding Matrix H using Qs and SQ decoders. As i said, and it's enforced by that document, SQ is not capable in any way of decoding Matrix H. All you're listening to is a phasy mush, with incorrect instrument placment etc.

You'll be glad to know, my Matrix H tests are over, and i'll be releasing "Instructions For Angels" on my blog for members this week.

I'll also be releasing modified versions of the QS & SQ scripts, and one for Matrix H, on my blog once it's fully running again

OD

 

[MidiMagic]

if you are going to try and prove something, at least put the correct information down so others can learn correctly"

UMX should read BMX

I used UMX because the magazines NEVER referred to BMX. They always said UMX, saying it was short for Uniform MatriX. If I say BMX to most enthusiasts (especially here in the US), they think of a type of bicycle. The only place I have ever seen BMX is in technical documents from Denon.

I also have trouble with the obscentity checker when I type BMX.

And don't tell me it's the same, UMX was the "Umbrela" name and consisted of more than one part, BMX was the name of the Matrix

Technically, BMX is the 2-channel matrix disc, UMX is the 4-channel encoded 4-track tape (and anything else with all 4 encoded channels), and UD-4 is the carrier disc.

There is also something your missing in you way of decoding Matrix H using Qs and SQ decoders. As i said, and it's enforced by that document, SQ is not capable in any way of decoding Matrix H. All you're listening to is a phasy mush, with incorrect instrument placment etc.

Did you catch that I said to move the speakers?

SQ is perfectly capable of decoding BMX in a rhombus speaker configuration, IF you do these:

- Move the speakers: Move LF and RF farther apart. Put LB between them in the front. Put RB in the middle of the back of the room (or have two RB speakers at the LB and RB positions)

- Reverse the phase of the LB speaker (now in front)

- Turn off the 10/40 blend

- If you have front-back logic, turn it off. But the Wavematching Variblend logic can be used.

This exactly matches the BMX modulations for left center, right center, center front, and center back.

Now, H is halfway between BMX and RM. This means the following are true:

- The left center and right center modulations are the same for all three (H, BMX, and RM).

- The H center front modulation combines both the clockwise modulation of BMX center front and the inphase modulation of RM center front.

- The H center back modulation combines both the anticlockwise modulation of BMX center back and the antiphase modulation of RM center back.

So I placed the SQ back speakers in the same places I put them in the BMX decode, but put them in the middle of an RM array of speakers. Once I get the phase set right, the H locations are created acoustically (but not electrically).

RMLF . . . SQLB . . . RMRF


. . . . . . . . me


RMLB . . . SQRB . . . RMRB

If I wanted to, I could put the other SQ speakers to my right and left. Then I can have RM with one set on, BMX with the other set on, and H with both. And if the RM decoder also has SQ, I could have SQ too, using the RM speakers.

It's not exact, but very close. Each speaker is at least as close to the correct modulation (or closer) as the left channel of a stereo set is to a sound panned halfway to the left on the stereo stage.

I am a college instructor. They don't pay me enough that I can gallivant around eBay and buy stuff. So I have to make do with what I already have. There is no way I am buying a BMX decoder for the one record (marked UMX) I have. Likewise, I have only four H recordings.

 

[ingresman]

I had a quick search on this site and couldn't find this article http://www.epanorama.rackhost.net/s.../Equalizers/Multisystem Ambisonic Decoder.pdf
So who fancies getting a soldering iron out ????!!!!
Just had a quick read through and it mentions part 3, so I had a further look and found this at his archive site ( http://www.michaelgerzonphotos.org.uk/ )
Michael Gerzon, "Multi-System Ambisonic Decoder",
Part 1: "Basic Design Philosophy", Wireless World, vol. 83 no. 1499, pp. 43-47 (1977 July)
Part 2: "Main Decoder Circuits", Wireless World, vol. 83 no. 1500, pp. 69-73 (1977 Aug.) later parts never written & published.

 

[mrfloydin]

All,

As a companion to my BBC Radio Show Index originally posted In June, I have just posted a similar index focusing on the mid-'70's BBC quad radio broadcasts done in Matrix H and HJ formats. I have also posted significant updates to my other previous radio show indices from that date.

From some of the feedback I received from my first postings, I got directed to some forum threads and learned more of the nature and extent of these unique broadcasts. From this I initiated some research and received helpful input from others that encouraged me to create and just now post a specific index of these broadcasts.

As with my previous indices, I again honestly state that these are 'rough' versions that I plan to reformat later, though that effort could take me some time. So it may be worth checking back on it in a number of months, to view the improvements.

Should you be interested to check out this new addition, please follow the initial link at: http://www.eclipse67.com/bbchr.htm

If you follow the links in this main page, it will take you to the new index at www.eclipse67.com/bbch.htm, as well as the multi-page sections that are devoted to the King Biscuit Flower Hour and the BBC Transcription Services & BBC Rock Hour programs. These include the cumulative indices for these shows as well. As stated, these are intended to be 'works in process' and are not assumed to be completely accurate or totally comprehensive. There are acknowledged holes in them and my hope is that I will receive clarifying input from viewers, so that I can update and refine these pages over time.

Therefore, I would sincerely appreciate feedback on specific content, clarity of the format, and with any general comments regarding this effort.

Feel free to share this link with any others whom you feel would be interested.

I hope that you find these new pages enjoyable and informative.

mrfloydin

 

[tadploe]

Hi Folks.
Did anyone on this thread manage to download the Matrix H encoded Queen BBC sessions from 1977?
Or perhaps the SQ encoded Queen At Golders Green in 1973?
I might get lucky, miracles do happen! :-D