How do I create a matrix-encoded CD?

[neil wilkes]

How do I go about making a matrix CD?
What format would I need to encode the final mix in - would creating a standard surround mix and encoding to an Lt/Rt with something like SRS Circle Surround do the job, or am I better off with LRCS matrixing?

 

[Malcolm2010]

How do I go about making a matrix CD?
What format would I need to encode the final mix in - would creating a standard surround mix and encoding to an Lt/Rt with something like SRS Circle Surround do the job, or am I better off with LRCS matrixing?

Neil,
Depends what you want to emulate, SQ or QS, and the target system. Either way if you produce a Dolby Surround mix, it will decode satisfactorily on a DPL II system or SQ/QS decoders, but not as you originally mixed it. In fact a lot of the guys here on QQ say that DPL II decodes exactly as QS, so maybe that is the one to go for.

The question to ask, for someone who is well into DVD-A etc, is Why?

Cheers

Malcolm

 

[Malcolm2010]

Neil,
Here is a link to an excellent article showing the various flavours of quad, you should be able to generate them using you editing package.

http://www.mtsu.edu/~dsmitche/rim456/Quad/Quad_Formats.html

Malcolm

 

[neil wilkes]

The "Why" is actually quite easy to answer.
I have a project on the go - waiting for final go ahead from the client - where the release is to be a full DVD-A title, but also a CD version too, as they are basically a bunch of old West Coast Hippies who want to consider their entire audience and not just Audiophiles.
They are also into technology and sneaky surprises.
So, one of the ideas on offer right now is for the CD version to be Surround encoded but playable in a bog standard stereo CD player. The users will find out about the decode thing one of 2 ways:
1/. Visiting the bands website, where the info will be there, or
2/. Finding out by accident.

The DVD-A part is no problem at all, neither is the Surround mix.
Given the encoders I have available to me, what I would like to know is what one gives the best results when encoded as Lt/Rt matrix.
I could always experiment, and probably should, but I was simply trying to be lazy I guess. Discrete - no problem.
Matrixed - not used it much as I prefer discrete. Time to get practising.
The encoders I own are a straight LRCS one and the SRS Circle Surround, and I am about to order a PL II Encoder as well, but that will be a standalone one.

Ah well.
I guess I should stop trying to get others to do my research for me.

 

[winopener]

Given the encoders I have available to me, what I would like to know is what one gives the best results when encoded as Lt/Rt matrix.

Neil,
it's easy to find today on the market a DPL2 decoder build into some hardware: even many cheap dvd player has it, notably the ones based on Mediatek 1389xE chipset. SQ and QS are just out of timeframe, CS while good is hard to find. DPL2 can be decoded also on PC software.
If you need to use a matrix IMHO dpl2 is the best choice today.

 

[kempfand]

If the source is good (24 bits), I'd vote for UHJ (which means you 1st have to pan your mix into B-format) :banana:

Kind regards,
Andreas

 

[quadtrade]

The "Why" is actually quite easy to answer.
I have a project on the go - waiting for final go ahead from the client - where the release is to be a full DVD-A title, but also a CD version too, as they are basically a bunch of old West Coast Hippies who want to consider their entire audience and not just Audiophiles.
They are also into technology and sneaky surprises.
So, one of the ideas on offer right now is for the CD version to be Surround encoded but playable in a bog standard stereo CD player. The users will find out about the decode thing one of 2 ways:
1/. Visiting the bands website, where the info will be there, or
2/. Finding out by accident.

The DVD-A part is no problem at all, neither is the Surround mix.
Given the encoders I have available to me, what I would like to know is what one gives the best results when encoded as Lt/Rt matrix.
I could always experiment, and probably should, but I was simply trying to be lazy I guess. Discrete - no problem.
Matrixed - not used it much as I prefer discrete. Time to get practising.
The encoders I own are a straight LRCS one and the SRS Circle Surround, and I am about to order a PL II Encoder as well, but that will be a standalone one.

Ah well.
I guess I should stop trying to get others to do my research for me.

Neil i have a circle surround encoder, a qs encoder, and a sq encoder.. .. i think circle surround decodes well in pl and pl2, excellent in circle of course. QS would be a good choice too. I vote CS or PL2 if ya get it.

 

[neil wilkes]

Thanks for all the suggestions!
I particularly like the link to the various matrix math. That could come in very useful indeed.
Encoders.
Got standard LRCS, SRS Circle, and am shopping for a PLII encoder too. SurCode is looking likely. PLII also has other uses, so would be worth buying.
Kempfand - the B-Format is a thought, but PLII is the most likely. I'm just not very good with Ambisonics, and not at all sure what I am doing there. Very, Very green in that department.

 

[Scottmoose]

The only other point I'd raise is that the Tate II for example, and variomatrix often produced results that were just as good as matrix encoded material from a decent, if agressive stereo mix with plenty of phase shifts. DPLII probably isn't that disimilar in this respect I imagine. If you're not sold / non too confident with ambisonics, I'd go for DPLII or SQ: the former if you want relative convenience, the latter if you want to take a trip on the wild side and discover why all those engineers in the 1970s though it was such a gigantic pain in the backside! The ambisonic software plugins I've used in deriving surround from stereo seem to work well though, if a little clumsily. Best to practice on something not required for a client first I suspect...
Cheers Neil, and good luck with the mix -I shall look forward to listening to both (no chance of letting us know who it is is there?).
Regards
Scott

 

[neil wilkes]

Scott. See PM.
However, this is not for an upmix. The original multitracks are available to me. The Artist wants the main release for DVD-A, but also a CD version as well, and it is the CD version I am looking to Matrix encode.
Circle Surround or PL II still looks likely.
I need it to play in stereo on a stereo setup, or Surround on a Surround setup hence the matrix requirement.

Cannot name the Artist publicly as yet....

 

[QUADradial]

I am able to encode DPLII from 4 to 6 discrete wav files.

Tab

 

[neil wilkes]

I've gone with the SurCode encoder - the price was too good to turn down.
SRS is appalling - serious gain boost on the rears, it seems to add 3dB to the Ls/Rs on decode. Not good.
DPL II is admittedly very limited for music use, but it was not designed for music.

 

[phydo]

Neil,
Depends what you want to emulate, SQ or QS, and the target system. Either way if you produce a Dolby Surround mix, it will decode satisfactorily on a DPL II system or SQ/QS decoders, but not as you originally mixed it. In fact a lot of the guys here on QQ say that DPL II decodes exactly as QS, so maybe that is the one to go for.

The question to ask, for someone who is well into DVD-A etc, is Why?

Cheers

Malcolm

Malcolm:

I am not really into the technical side of quad/surround sound, but I thought I had read years ago that a form of SQ, rather than QS, was what the motion picture industry used when they developed the original movie surround sound. Of course motion picture surround sound has become so much more sophisticated with competing formats that maybe an advanced form of QS is in use. I am not trying to pick nits - just trying to learn from the best.

 

[neil wilkes]

The Why is easy.
I am currently mixing a surround project for DVD-A/V release, but the band also want a stereo version on CD.
I thought it would be a blast to have a stereo version that is also matrix encoded for surround too - that way, it's like a surprise. The info will be on the band website telling those who go there what they have.
That's it really.
I just don't want to do a straight stereo - it sticks in my throat somewhat.

AFAIK, SQ encoding mutated into the original Dolby Surround/Pro Logic, with the encoding going to LRCS rather than LRLsRs.
When it got changed to DPL II, despite the incremental name, it was actually a totally different system.
DPL II is easy - as soon as the new SurCode VST DPL II encoder is released I will have a copy and r8un it straight off the Nuendo timeline, as I already do with DD/DTS.

 

[Disclord]

AFAIK, SQ encoding mutated into the original Dolby Surround/Pro Logic, with the encoding going to LRCS rather than LRLsRs.
When it got changed to DPL II, despite the incremental name, it was actually a totally different system.
DPL II is easy - as soon as the new SurCode VST DPL II encoder is released I will have a copy and r8un it straight off the Nuendo timeline, as I already do with DD/DTS.

Dolby MP is actually an outgrowth of Sansui's QS. Dolby Pro*Logic-The Sequel is a 'great circle' encoding/decoding system and the Ls-Rs channels are encoded with a 180 shift and +-3db imbalance. So, a 180 phase shift and sine/cosine panning will encode the back channels. Circle Surround and Logic-7 encode the same way. Don't ever think about using 2-channel UHJ (BHJ) - it's unlistenable with a 45 degree phase shift for CF sources. Of the various multichannel matrix formats, SQ had the best stereo compatibility. It also had the most directional information for the logic system - and it wasn't just limited to "Quadrifrontal" mixing. SQ became a Kernel system (like UHJ) when the Ghent Microphone was used. The Universal SQ (USQ) system even offered compatible 4-2-4, 4-3-4 and 4-4-4 discrete performance. But, SQ is dead now... sadly.

Cheers!
Ty C.

 

[phydo]

Dolby MP is actually an outgrowth of Sansui's QS. Dolby Pro*Logic-The Sequel is a 'great circle' encoding/decoding system and the Ls-Rs channels are encoded with a 180 shift and +-3db imbalance. So, a 180 phase shift and sine/cosine panning will encode the back channels. Circle Surround and Logic-7 encode the same way. Don't ever think about using 2-channel UHJ (BHJ) - it's unlistenable with a 45 degree phase shift for CF sources. Of the various multichannel matrix formats, SQ had the best stereo compatibility. It also had the most directional information for the logic system - and it wasn't just limited to "Quadrifrontal" mixing. SQ became a Kernel system (like UHJ) when the Ghent Microphone was used. The Universal SQ (USQ) system even offered compatible 4-2-4, 4-3-4 and 4-4-4 discrete performance. But, SQ is dead now... sadly.

Cheers!
Ty C.

Thanks for the info.

Justin

 

[Old Quad Guy]

While reading this thread I found the link below no longer works. Did anyone archive it and if so could it be posted somewhere, perhaps here?

I'm interested in learning more about the mathematical theories of SQ and QS matrix encoding/decoding.

Neil,
Here is a link to an excellent article showing the various flavours of quad, you should be able to generate them using you editing package.

http://www.mtsu.edu/~dsmitche/rim456/Quad/Quad_Formats.html

Malcolm

 

[sspsandy]

It is my understanding based on writings by Jim Fosgate that PL II was designed for music as well as for movie soundtracks. Fosgate has always pioneered the idea of surround in music with the emphasis on matrix encoded material as opposed to digital formats, as well as processors that can derive surround from any stereo source. I have found that PL II encoded stuff decodes super well on a Tate, and that SQ stuff decodes quite well on a true PL II processor, such as the RFQ 5000 or the FAP T1, all the others were dumbed down by the manufacturers who felt the customers didn't want such active surround channels.

 

[jdmack]

While reading this thread I found the link below no longer works. Did anyone archive it and if so could it be posted somewhere, perhaps here?

I'm interested in learning more about the mathematical theories of SQ and QS matrix encoding/decoding.

There is a book called "Four Channel Sound" by Leonard Feldman that gives detailed explanations about all the various matrixed quad systems. As of today, there are two used copies of this book available at half.com and seven used copies at amazon.com. But here is the short version of quad encoding equations:

QS System

Left = (left front) + (.414 right front) + (left rear, phased +90 degrees) + (.414 right rear, phased +90 degrees)
Right = (Right front) + (.414 left front) + (Right rear, phased -90 degrees) + (-.414 left rear, phased -90 degrees)

SQ system

Left = (left front) + (.7 left rear, phased -90 degrees) + (.7 right rear)
Right = (Right front) - (.7 L rear) + (.7 right rear, phased +90 degrees)

J. D.