Not Understanding "Script" Matrix Decoders

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kfbkfb

2K Club - QQ Super Nova
QQ Supporter
Since 2002/2003
Joined
Feb 20, 2003
Messages
2,129
Location
Midwest USA
Been reading some older posts about script (software) based SQ and QS decoders (I haven't used scripts myself, so I'm just inferring what they do).

Some listeners say the script based decoders provide channel separation close to discrete from matrix encoded content.

Without some sort of "logic" assistance, I don't see how basic (3 dB) decoders can sound close to discrete.


Kirk Bayne
 
Adobe Audition 3 has a Centre Channel Extractor function that can be used to remove the out of phase audio but leave in the in phase audio, so the out of phase rear channels info can be removed from the in phase front channels and similarly for the in phase rear channels. The only problem is the artefacts it can cause. If you can hide that it sounds pretty good.
The great thing about the script function is that many commands can be combined and saved for future use so it can then be left to do it's thing on any audio required.
 
Been reading some older posts about script (software) based SQ and QS decoders (I haven't used scripts myself, so I'm just inferring what they do).

Some listeners say the script based decoders provide channel separation close to discrete from matrix encoded content.

Without some sort of "logic" assistance, I don't see how basic (3 dB) decoders can sound close to discrete.


Kirk Bayne

Good concise explanation by @monarchee . I might expand a bit because if you've never done a "script" you might wonder what it is & how do you create one.

A script is at its core a text oriented file that tells the app what to do in an automated way. After making one you can even open it up in something as simple as notepad & see what you've done. In Adobe Audition you create one by using a chunk of a selected song & manually running through the steps you want to do. So, lets say your having some friends over for a BBQ cook out in the back yard (post plague, of course). You want some music & ripped dozens of CD tracks to your HDD. You know you'll want some bass boost because your using those funky Acoustic Research flame speakers you bought at Costco last year. It would also be good to even out the volume level between tracks. So ya open up just one song, select maybe 10 secs of it. Go to the create scripts dialogue box & hit record. Then go back to your opened file. Since you will be boosting bass perhaps first normalize at -6dB. Then go to any one of several EQ options in AA & change what you want. Then it might be good to add some compression so it says more level consistent for this outdoor experience. Finally go back in and normalize to -1dB so the over all peaks are consistent. Go back to the script dialogue box & click save with the label/name you want for this scripts. Maybe backyard.

After that in AA you would go to Batch Processing & add those dozens of wav files you have. Select the script you want to use & click run. For something simple like this it will go fast, really fast, compared to doing it by hand.

As Monarchee said, the "logic" in these scripts decoders is the Center Channel Extractor, re-purposed to eliminate common cross talk after simple decoding. It truly is not "logic" in the common sense as logic uses gain riding to suppress unwanted cross talk. All the cool guy decoders use cancellation techniques & that's what this does.

Consider an RM/QS simple decoded output intended for left front only. The diagonal right back will be silent, no cross talk. Right front & left back will have matching levels of left front cross talk at -3dB. Apply the Center Channel Extractor to right front & left back & this is eliminated. In use this cancels out only the bleed through doing no harm to, as Sansui would say, the sub-dominant signals.

Most of the decoding scripts on QQ have to do specifically with QS or SQ decoding. As a result they use 90 deg phase shifts & work on front left/right and back left/right pairs. In the in the interest of duplicating Sansui Synththesizer decode mode for stereo on the PC, I took a different approach. Firstly no 90 deg phase shifts are used since there is no likley hood that a stereo recording would have that as a constant. Secondly I did it as in mt example above, working on the diagonals creating simple decoded left front and right back tracks, right front/left back tracks. Then run the CCE. Then through the magic of copy & paste I put the tracks back in the proper set of front L/R and back L/R.

The CCE is really pretty nifty. I don't know of any audio editing app that has anything as similar. The cross talk elimination is exactly proportional to what the settings are. Dial -10dB, or -25dB, and that's cross talk elimination you'll get. You could set it higher but past 35 dB it sounds harsh.

1615823520074.png
 
Last edited:
The CCE is really pretty nifty. I don't know of any audio editing app that has anything as similar. The cross talk elimination is exactly proportional to what the settings are. Dial -10dB, or -25dB, and that's cross talk elimination you'll get. You could set it higher but past 35 dB it sounds harsh.

The PhaseBug plugin come in handy for the DAW that haven't a CCE.
 
The PhaseBug plugin come in handy for the DAW that haven't a CCE.
Don't you mean the Phasebug plugin comes in handy for a DAW that doesn't have a phase shift function!!! It was useful when CoolEdit, Audition's predecessor didn't have one. The CCE was quite a unique function of Adobe Audition when it first came out and has nothing to do with phase shifting.
Oxforddickie, late of this parish, always swore by the Phasebug plugin against Audition's inbuilt phase shifter, but then he had golden ears.
 
There is a bootleg dvda called "NEKTAR73-74" (it has Remember the Future & Down to Earth). The decoding is almost too good, I suspect that software/AA was used, but there is no info at all with the file.
Anyone here know what the mystery decoder is on this one?
 
Quick question:

What was the original intent of adding a CCE to AA to begin with? If not to decide Quad, what other purpose would it normally be used for?

Too bad they discontinued it when AA 4 was released, if memory serves.
 
Quick question:

What was the original intent of adding a CCE to AA to begin with? If not to decide Quad, what other purpose would it normally be used for?

Too bad they discontinued it when AA 4 was released, if memory serves.

Actually the CCE can reduce, or boost vocals, drums anything that has inter-channel common mode. If you have a commercial track & want to replace the vocal with another for a demo track, this is how you do it. Maybe a recorded vocal in a stereo mix is a bit too loud. Reduce it. Maybe a vocal is buried in a mix. Enhance it.
Yeh.
 
Actually the CCE can reduce, or boost vocals, drums anything that has inter-channel common mode. If you have a commercial track & want to replace the vocal with another for a demo track, this is how you do it. Maybe a recorded vocal in a stereo mix is a bit too loud. Reduce it. Maybe a vocal is buried in a mix. Enhance it.
Yeh.
And if you take the reduction of vocal to the max. What do you get. A karaoke track!!!!!! 😉
 
Back
Top