Multi-Matrix Quadraphonic Decoder for Plogue Bidule

[S T Riddle]

When I came aboard I mentioned that I had a "Work in progress" computer-based decoder. I will now make it available for your evaluation.

It does SQ (Lynn Olson's "Shadow Vector" - Lynn is a heck of a nice man and was very willing to help me overcome early problems), QS (Vario-Matrix V2), EV-4 (passive), DY (passive with active center removal from rear), Dolby Stereo Surround (old style - not Pro Logic), Uniquad Equal Separation (an idea by Peter Scheiber that I found on MidiMagic's site).

It runs only in Plogue (sorry, haven't found a free VST host that is capable of math computations). There are occasional time-limited trial versions of Plogue but this program is worth the investment and some of the "toys" in the decoder may be useful enough to justify it for many of you (such as a pre-processor inspired by Lynn Olson's "Crosstalk Eliminator" to align phase, balance, and inject crosstalk canceling signals. Also a "poor man's" vectorscope using a free VST.) If you rip vinyl the pre-processor can be used to correct minor set-up errors of your stylus and improve separation.

It decodes wav and flac files in "real time" and properly set up (with an ASIO driver) it can decode "on the fly" straight from vinyl, CD, and other sources.

To get the Plogue Bidule layout for Multi-Matrix go to https://www.surroundbyus.com/
You'll have to register to download (but it's free). The download can be found in the SBU Wiki > General Techniques section.

Like the upmix converter "SPEC" (also found at SBU) Multi-Matrix is Donationware. If you like it consider donating to help me improve it and fix any problems uncovered in the "launch". I am a retired (disability) Electrical Engineering Tech (Associates Degree) so I always run a generation (or two) behind in the software/hardware area but I think I'm on the right track with this decoder. It was years in creating, testing, and improving.

Please give it a try, spread the word, and let me know what you think.

 

[JonUrban]

I will get this sometime this week and give it a spin. I haven't used SPEC in a while, so I probably have a few updates to do.

Well worth the cost of getting PLOGUE and trying this, and SPEC, out.

 

[S T Riddle]

Thanks Jon, I look forward to your review.
The trial version of Plogue is good until September so try it now.
Yes, Plogue ($95) is cheap enough to make this worth it just for the "Shadow Vector" (the main project that started the Muli-Matrix).
To put it in perspective... If I posted that I saw a unit on Ebay that emulated the Audionics "Shadow Vector" for $95 how many of the members here would buy it or try to outbid each other?

 

[fizzywiggs41]

Sounds like it would be a lot of fun S T RIDDLE .

I like that price $95.00, that's a real bargain for all those matrices!!


Any chance of developing an outboard unit ? Sorry I don't have a computer, just my trusty tablet.

 

[georgeshannon]

I'm intrigued. Thanks.

 

[JonUrban]

Thanks Jon, I look forward to your review.
The trial version of Plogue is good until September so try it now.
Yes, Plogue ($95) is cheap enough to make this worth it just for the "Shadow Vector" (the main project that started the Muli-Matrix).
To put it in perspective... If I posted that I saw a unit on Ebay that emulated the Audionics "Shadow Vector" for $95 how many of the members here would buy it or try to outbid each other?

I paid for the Plogue a while back, but I seem to remember that it was some sort of pre-release deal. I haven't messed with it in a year or so, so maybe tonight I will fire it up and see what happens. If I have to buy it again, I suppose that would be a function of there being an "upgrade" for the people who already bought it in its earlier release.

 

[deepsky4565]

Thanks for making this available. Hope I'll have time soon to try it out.

 

[JonUrban]

OK, I'm rusty and old at this.

I checked out by Plogue and downloaded the newer version (I had 9722 and the latest is 9747)

Now, as I recall to run SPEC I needed to use the x32 version, AND start the program with 'P2' instead of Plogue. (I never asked why)

So, to run this decoder plugin, what are the rules and where should I put the decoder file? I would guess in the "Layouts" directory, right?

 

[JonUrban]

One more thing. I just tried running SPEC and now my layouts have red boxes and I think I may have screwed it up. It's been so long since I've done this I can't remember what I did before. I may have to start from scratch.

So bear with me

 

[Colin Dunn]

A question since I'm not familiar with Plogue -

Is it possible to have the multi-matrix VST and Plogue decode real-time into .WAV files? As I am doing conversions I would want to play the LP and get 4 discrete channels of output. Then process the output files (de-click, noise reduction, etc) and merge into a multi-channel .WAV for output to DVD-A.

This is an exciting development - a modernized replacement for the Audition scripts many of us have been using.

 

[S T Riddle]

Sounds like it would be a lot of fun S T RIDDLE .

I like that price $95.00, that's a real bargain for all those matrices!!


Any chance of developing an outboard unit ? Sorry I don't have a computer, just my trusty tablet.

Sorry but no boxes... my days with a soldering iron are over. I'll leave boxes to our Involve boys.

 

[S T Riddle]

Jon,

On the questions about Plogue:
1. If you paid for it once you should have an "Early Bird" license that you can keep updating at no cost.
2. The program runs in X32 (many VSTs aren't 64 bit anyway). As far as starting in P2, I don't (but I don't run SPEC and don't really know what P2 does).
3. The downloaded file is a Layout and goes in that folder.
4. Red boxes mean that the "red' component (or group) is missing something ("right click" red box select group and see if "expand" is an option - sometimes it's a missing piece inside a group) or not properly installed (ex. VST).

 

[S T Riddle]

A question since I'm not familiar with Plogue -

Is it possible to have the multi-matrix VST and Plogue decode real-time into .WAV files? As I am doing conversions I would want to play the LP and get 4 discrete channels of output. Then process the output files (de-click, noise reduction, etc) and merge into a multi-channel .WAV for output to DVD-A.

This is an exciting development - a modernized replacement for the Audition scripts many of us have been using.

"Real Time" decoding from a source depends on whether your computer/soundcard combination will let you use a "Duplex" setup with ASIO. My old computer (Dell running "Vista" with onboard 5.1 card let me but my "newer" system (Dell running Win7 with a Nvidia graphics board with HDMI won't let me.

Worst case: you have to rip the source to wav or flac the run it.

Output is channeled to a file recorder in Plogue that will save as individual wavs (4 or 5 if you want a Center, I don't do LFE that's always a problem waiting to happen). Another option is interleaved(?) in one wav file.

A word of warning: SQ and QS are CPU intensive. I have no problem running them in 44k or 48k but when I try to run at 96k it can bog down sometimes (I don't trust the -90 degree shift in Plogue or Phasebug at that rate). At low frequences most phase shifters just fail either in maintaining levels or waveform. The only one I found that seems good enough is Phase Tool by Flying Pigeon by it's a CPU hog too so 96k is tricky. Plogue will let you process "offline" which runs much quicker but you still need to listen to get it set-up first (hint: set-up while in 44K or 48K then stop and switch to Offline at 96K).

 

[kronos77]

splendid layout the QS vario matrix mode is the one i love more when listen santana song

 

[S T Riddle]

splendid layout the QS vario matrix mode is the one i love more when listen santana song

Thanks Kronos,
I thought all quad Santana albums were SQ... so you must be using the QS mode as a quad synth. Anyone know if Santana has material on any other quads system (perhaps QS or CD-4 in Japan)?

By the way, if you didn't catch it from my description the "stereo expand" mode takes the stereo field and stretches it (Left channel gets transposed to LB, Right goes to RB, and the stereo field is stretched into a semi-circle in front of you). This can be effective on a recording with widely panned elements.

 

[JonUrban]

Well I screwed up my Plogue so badly that I decided to "go back in time". A few months ago I got a new SSD for the C: drive on my big computer, and in doing so I cloned the original drive to the new drive and away I went. So today when I had some time I went back and put the old SSD C: drive back in the PC (making sure to unplug the NIC card so nothing went out to update itself or check for license changes) and ran my Plogue as it was back the last time I used SPEC. Luckily, everything was still good, so I loaded the new SQ decoder layout and gave it a try.

Took clearing a few cobwebs from my Plogue brain, but I got it all running. To see the default little layout created by the download and modified a little bit by me, see below.




The SQ encoded file that I used was "Say Goodbye to Hollywood" from Turnstiles. I had that album ripped to my HD, unfortunately only in 16/44, but straight from the SQ LP. I ran the decoder using all default settings, then opened the file in Sound Forge. Pretty good!

To compare, I pulled out a 4CH file I created using Adobe Audition and one of the older SQ scripts (Pre-OD Final) and looked at and listened to them both. The results are very similar. To see for yourself, the graphic below shows the AA SQ decode at the top, and the new decode at the bottom.



You can see from the wav files they are damn close. Disregard the audio levels, as I did not try to match them as this was just a test. But you can see a very clear definition of the percussion in the rear channels in the AA SQ decode, and they are even clearer in the newer decode.

Interesting. More when I get a chance. Maybe I will post some sample wav files of the rear channel comparision in a few days.

 

[S T Riddle]

Jon,
Thanks for giving it a try and a positive review (at least at first listen). The OD SQ AA file does look very similar and I'll take that as positive also.
Did you process in "offline" mode or while listening? Not familiar with the source file you used, at least by name... if I heard it I might recognize. Was it an easy or hard file in regard to difficulty to resolve positions? (example: DSOTM "Money" is generally hard to resolve on the back channel sequential echoes since they diminish to low levels compared to other signals)

Someone on the SBU forum has stated that the SQ and EV-4 didn't sound quite right (compared to his other decoders - no brands mentioned) so I'm working a re-wire in case I made an error last time I changed it and to also make it more CPU friendly. Will upload when I finish testing it.
Meanwhile, enjoy. And keep me informed of perceived anomalies or shortcomings, it may not beat a Tate but I may be getting closer.

Also, Shadow Vector SQ as described by the inventor, Lynn Olson, on this very forum (do a search for his full comments) was never intended to "force" it's way to maximum separation it was engineered to provide 30-40 dB separation (on a 1000Hz test signal?) while still allowing the natural reverberation in the field to remain where it belonged. I'm not quite that close to his specs in the full circular field with real musical sources but am close at the "cardinal points" (a objection he has to testing decoders vs. any point in the field) when using 1000Hz test.

 

[S T Riddle]

Separation Stats for a quick test of the Shadow Vector SQ decoder as supplied in the layout:

1000Hz test signal encoded to "proper" position encode according to Benjamin Bauer diagram.
Run at 44K processing decoded using layout as furnished in the downlaod.
Test signals were produced using a Phase VST (Phase Tool by Flying Pigeon) which offers better phase coherance and more angle choicces than Plogue by itself. (A word of caution: Phase Tool appears, to me, to have it's output 189 (EDIT: 180 )degrees out and must be inverted).

L to R= 40dB (L or R to either Back= inf; and better than 35dB at any front position)
CF to CB= inf (CF to either Back= inf)
LB to RB= 34dB (LB or RB to either Front= 38dB)

Special cases are used for LSide or RSide as well as a CB encodes according to Mr Bauer:
LSide is .924Lt at -45deg and .383Rt at -135 deg.
RSide is .383Lt at -45deg and .924Rt ar +45deg.
CB is .707Lt at -45deg and .707Rt at +135deg.
These are apparently done this way to keep such signals as isolated from the 4 corner signals as far as posible when recording and/or cutting a vinyl record.

LSide to RSide= 34dB or better.
CB to either Front= inf

As stated (Lynn Olson objections aside) this is only a "static" test.

General disclaimer: "your gas mileage may differ" when used with musical sources utilizing many fast changing levels and frequencies.

 

[user 8639]

It runs only in Plogue...

Obviously a tough work !
Have you tried to work with the guys at VLC ?
They usually implement any codec they can in their universal software...
It would make a very well spread matrix decoder engine, for all

 

[atrocity]

If you rip vinyl the pre-processor can be used to correct minor set-up errors of your stylus and improve separation.

Intriguing...so if I digitized some SQ vinyl, the pre-processor would allow the stereo file to decode better via the Surround Master as well as your decoder?

Would there be an advantage to using the pre-processor even for standard stereo captures?

Thanks!