PC Based SQ DECODING - ALMOST DONE!

[winopener]

Also, I am wondering if pushing the separation past the cartridge rating on vinyl played SQ is causing some artifacts. I think if you try to get more separation than the stereo cartridge could provide might be a mistake. My cartridge is rated at around 25db left/right. Pushing the software past that makes it sound watery during the most quiet passages...

Tab,
really possible. I was thinking something alike, especially on trying to improve the stereo separation of the magnetic carts. Unfortunately this is a process that is cart-dependent so it can't be automated at all, but it is possible to raise L/R separation by 9-10dB using a 180deg. sum of the other channel properly reduced. The best way to set a correct value of attenuation is to preview the result using channel mixer and monitoring carefully only *one* channel at the time.
Having a 35dB separation to start with for the SQ decoding instead of a 25dB will do a lot better.

This week i will be very busy so i will not have any time for testing (work calls!), after easter i can try something.

 

[Malcolm2010]

Tab,
really possible. I was thinking something alike, especially on trying to improve the stereo separation of the magnetic carts. Unfortunately this is a process that is cart-dependent so it can't be automated at all, but it is possible to raise L/R separation by 9-10dB using a 180deg. sum of the other channel properly reduced. The best way to set a correct value of attenuation is to preview the result using channel mixer and monitoring carefully only *one* channel at the time.
Having a 35dB separation to start with for the SQ decoding instead of a 25dB will do a lot better.

This week i will be very busy so i will not have any time for testing (work calls!), after easter i can try something.

The Audionics S&IC has an "Axial Tilt Control" that tries to compensate for slight mis-alignment of the cartridge. It operates in a similar way to what you describe, cancelling crosstalk left to right and right to left to maximise separation. Personally I would be happy with 25dB of smooth separation without artifacts than 30 dB with them. It is also dependant on the alignment of the original cutter head, and the quality of the stamping, both can introduce left/right shifts even on a perfectly aligned cartridge.

Malcolm

 

[Scottmoose]

the separation of phono cartriges goes down at higher frequencies (mine has >25db at 1KHz but "only" 15db at 15KHz, and it is a good MC cartirge...

Tell that to Decca. The London cheerfully reaches around 40khz, give or take with little if any reduction in separation. I'm not sure I'd want to use one for quad though as it was originally designed for mono... :banana:

 

[berninahusq]

I have automated a bit the processing; here's how i do it:
1) place the wave file sq-encoded on c:\sq
2) open it in AA, convert (if not already) to 32-bit res. and normalize at -6dB
3) save it as front.wav
4) save it again as rear.wav
5) run the script below; after the script complete, raise front.wav by 3dB

***begin of script***
Collection: Winopener
Title: SQdecoding
...

Hi Winopener,

I was so excited about everyone's experiments with this that I bought and installed Audition 1.5 last Saturday. I am not getting the results that everyone is getting with this script, however. Yet I have not tried the steps manually with Andreas' instructions but plan to tonight or tomorrow night.

I copied this script between the begin and end of script and created and saved a file according to Andreas' instructions

Neil,

Just copy the entire script from here into Notepad, and
-> save it as "SQdecoding.scp" (w/o the .txt)
-> into C:\Program Files\Adobe\Audition 1.5\Scripts

I've done it that way and it works like a charm !

Regards,
Andreas

Then I followed your instructions to create c:\sq\front.wav and c:\sq\rear.wav files converting it to 32-bit and normalizing it to -6db (do I normalize it by selecting amplitude at 94%?). Then I close all files and run SQdecoding.scp. After executing the script for about 8 minutes or so for one song (I have a 3.2 Ghz Pentium 4 with 1 GB PC 3500 RAM) and Audition indicates that I finished running the script, I had varying results.

On an SQ LP-source wave file and a track from the Lotus CD, I could tell little difference between the front and the rear files. On the SQ LP-source track, I know what it sounds like through a TATE II and the Audition files were no where near in separation, more like synthesized matrix.

On an Annie CD http://www.cduniverse.com/productinfo.asp?pid=1088017&cart=236836293&style=music&bab=A track recording wave file (single inventory SQ recording), the WHOLE rear.wav was completely silent when the script was all done. Interesting why Audition would take all the music out of the Annie rear wave file.

Can you please tell me what I am doing wrong?

Thanks!!

Ed

 

[imploder]

Hi Winopener,

I was so excited about everyone's experiments with this that I bought and installed Audition 1.5 last Saturday. I am not getting the results that everyone is getting with this script, however. Yet I have not tried the steps manually with Andreas' instructions but plan to tonight or tomorrow night.

I copied this script between the begin and end of script and created and saved a file according to Andreas' instructions



Then I followed your instructions to create c:\sq\front.wav and c:\sq\rear.wav files converting it to 32-bit and normalizing it to -6db (do I normalize it by selecting amplitude at 94%?). Then I close all files and run SQdecoding.scp. After executing the script for about 8 minutes or so for one song (I have a 3.2 Ghz Pentium 4 with 1 GB PC 3500 RAM) and Audition indicates that I finished running the script, I had varying results.

On an SQ LP-source wave file and a track from the Lotus CD, I could tell little difference between the front and the rear files. On the SQ LP-source track, I know what it sounds like through a TATE II and the Audition files were no where near in separation, more like synthesized matrix.

On an Annie CD http://www.cduniverse.com/productinfo.asp?pid=1088017&cart=236836293&style=music&bab=A track recording wave file (single inventory SQ recording), the WHOLE rear.wav was completely silent when the script was all done. Interesting why Audition would take all the music out of the Annie rear wave file.

Can you please tell me what I am doing wrong?

Thanks!!

Ed

I still do it manually...the rear mixdown could be run by script, but doing it manually remembers me always to the basic principle.
Do not trust the scripts, do it by hand, then you will see the step where it is failing.
Please do also use the one "rear basic deconding method" that produces in phase results, like my handdrawn scheme.

One hint: As I decoded only material(exept Tab's test files) I recorded myself directly from LP to PC, I am sure that my decoding method is correct. Be sure that all source files are in right phase and channel order to get good results.

Regards

Andy

 

[Malcolm2010]

For Experimenters, here are a set of SQ encoded test tones. These are digitally produced with correct phase and amplitude. The file contains 5 seconds each of the following, centre Front, left front, centre left, left rear, rear centre, right rear, centre right, and right front.

They are here :- www.grizwald.plus.com/quad/SQTones.mp3

Malcolm

 

[imploder]

Unfortunately the link does not work.

 

[Malcolm2010]

Unfortunately the link does not work.

Strange, it does for me, ok go to

www.grizwald.plus.com/quad

Right click on SQ Test Tones, and save target. I will investigate the link.

Malcolm

 

[Malcolm2010]

Strange, it does for me, ok go to

www.grizwald.plus.com/quad

Right click on SQ Test Tones, and save target. I will investigate the link.

Malcolm

Problem solved, it is a *nix based server, I missed the capital letters, link now should work

Malcolm

 

[imploder]

Thanks, now also the old link works, strange.

I also have finally found out a method for QS and RM decoding that separates the channels completely with almost no crosstalk, it seems so that the QS matrix is not as sensible as SQ.
I still work on reducing the necessary steps in AA1.5, but basically it works perfectly.

Is somebody interested in that or are there only SQ freaks outside?

I only can say that the test material I have sounds very, very free and "untreated" after decoding. Perhaps QS was the better system???

Greetings and Happy Easter to everyone

Andy

 

[Malcolm2010]

Thanks, now also the old link works, strange.

I also have finally found out a method for QS and RM decoding that separates the channels completely with almost no crosstalk, it seems so that the QS matrix is not as sensible as SQ.
I still work on reducing the necessary steps in AA1.5, but basically it works perfectly.

Is somebody interested in that or are there only SQ freaks outside?

I only can say that the test material I have sounds very, very free and "untreated" after decoding. Perhaps QS was the better system???

Greetings and Happy Easter to everyone

Andy

Andy,

I have the same set of tones encoded in QS if you want them. I can upload to my web site for you

Malcolm

 

[imploder]

Andy,

I have the same set of tones encoded in QS if you want them. I can upload to my web site for you

Malcolm

That would be nice! Thanks in advance!
I am sure the process works, I took the scheme from formerly mentioned Wendy Carlos' website and treated Tab's test material as it should be treatedb theoretically. Out came perfectly separated material.

Andy

 

[bizmopeen]

I also have finally found out a method for QS and RM decoding that separates the channels completely with almost no crosstalk, it seems so that the QS matrix is not as sensible as SQ.
I still work on reducing the necessary steps in AA1.5, but basically it works perfectly.

Is somebody interested in that or are there only SQ freaks outside?

Bring it on. I'm a multifaceted freak...

Thanks,

Allan

 

[Malcolm2010]

For Experimenters, here are a set of QS encoded test tones. These are digitally produced with correct phase and amplitude. The file contains 5 seconds each of the following, centre Front, left front, centre left, left rear, rear centre, right rear, centre right, and right front.

They are here :- www.grizwald.plus.com/quad/QSTones.mp3

Malcolm

 

[imploder]

So...
is everybody lost in decoding out there?
Can somebody write something about the results and personal impressions? I would be interested!

After easter holiday I will try to explain a simple way to decode QS, it works with the same tools, but is a bit more complicated.

Andy

 

[QUADradial]

I am still decoding with it. I will say that the FL/Ls and the FR/Rs side image problem had me stumped. There is a definate drop in separation. But...running the same file through the Tate yielded THE SAME FLAW. A smooth 360 degree pan around the room is not going to happen with SQ. If a sound is panned somewhere between Front Left and Surround Left, the Surround Right starts picking up a horrific bleed. Same thing happens if you try it between the Front Right and Surround Right...the Surround Left starts gaining amplitude. This problem is in both the software decode and the Tate decoder. So...that being said...the software decoder probably provides 50 to 75% more separation without drift.
What is your process for QS? Any luck in getting more separation out of the left/right?

 

[proufo]

So...that being said...the software decoder probably provides 50 to 75% more separation without drift.

Hello Tab.

Is that conclusive then? Can we expect better results in all areas from software decoding than from a Tate II?

What can be done to improve on the Tate using software, in order to be able to make smooth pans for example?

 

[proufo]

If a sound is panned somewhere between Front Left and Surround Left, the Surround Right starts picking up a horrific bleed. Same thing happens if you try it between the Front Right and Surround Right...the Surround Left starts gaining amplitude.

This is with test tones, right?

Are there any real albums that try to do that, place a sound midway between the same side speakers or do a smooth pan between same side speakers?

Maybe that fault was known by producers and they avoided it.

Also, can any of the experts explain why is this so, with vectors and angles?
Is it a consequence of the basic engineering or is it the result of real-world effects?

 

[Cai Campbell]

Maybe that fault was known by producers and they avoided it.

Yes, of course. The limitations of SQ are well documented as is the fact that engineers had to limit their surround-mixing options to circumvent them.

 

[QUADradial]

This test was done with music. Good ol' Enoch Light, in fact. You know his ping-pong spin-you-around technique. I used the opening from "The Night They Drove Old Dixie Down" as the test. It spikes the opposite channels when a side pan is performed.

I don't think this technique will help a Tate, but...you can use a non-logic decoder now and process those files!