S2S from QS via PC, OK? Part 2

Help Support QuadraphonicQuad:

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
May 30, 2005
Kansas City
Edit Note 6/25/05: Thanks to Kempfand's sharp eye a typo error has been fixed. The Channel Mixer Preset #3 has been corrected to read as intended.

Thank you, indeed, to those paying such close attention to my note. And my apologies to anyone wasting valuable time trying to make this work with an incorrect preset value.

Greetings again to everyone with more than two ears!

As promised I’ll detail the actions to get high quality, high separation surround sound from stereo via the PC. This will be split into four (HA!) sections: setup, decoding, pre/post decoding enhancements and variations on this process.


Make it easy on yourself. Make some presets in AA 1.5 so they’ll be ready to go. Start in Channel Mixer.
Preset 1:
Left = 1.0
Right = - 41

Left = - 41
Right 1.0
Add and label this QS PRE SYNTH

Preset 2:
Left = 92
Right = 38

Left = - 38
Right = 92
Add and label this preset QS LF/Rs

Preset #3:

Left = 92
Right = -38

Left = 38
Right = 92
Add and label this preset QS RF/Ls

Some of the other AA presets already existing will be used too.

Now, go to Center Channel Extractor & make two presets.

Preset #1:
Get audio = center
Center Channel level -10 dB
Xover = 80

Phase disc = 6 Amp Disc = 1 dB
Spectral = 0 Amp Bandwidth = 5 dB
FFT = 12000 Overlay = 10 Interval = 27 ms

Add and label this as QS Crosstalk Reduce

Preset #2:
As above but set Center Channel Level to + 10dB and Xover to 77.

Add and label as QS Center Front.

Also the files you are working on will go through several steps so I suggest making individual folders to keep it straight. It can be very confusing and some of the files will actually be mislabeled after exporting so do your best to keep things organized in a way that will make sense to you. If I’m working from a CD the files go into CD RIP folder. After creating multiple copies of the ripped wav files they go into the (WHATEVER NAME) WORK FILES folder. After decoding they are exported and go to (WHATEVER NAME) EXPORT folder. After DTS encoding they are saved in the (WHATEVER NAME) DTS folder. Working from Laserdisc capture to go on to DVD requires some additional steps and ripping from DVD needs other folders to hold things. Use labels and names to suit yourself. If possible when using multiple HD’s try to arrange it so you’re bouncing from one drive to another. Reading from one HD & writing to another goes much faster. Yup, that’s why I’ve got four internal HD’s in my workstation.

In AA editor mode, open up one of the files to start working on.
Select all & copy.
Save copy as (NAME) LF Rs & put it in the WORK FILE folder.
Save another copy as (NAME) RF Ls & put it in the WORK FILE folder
Save another copy as( NAME) CF & put it in the WORK FILE folder
Save a fourth copy as as (NAME) LFE & put it in the WORK FILE folder

Obviously if you are working with only a four channel end product then you can forget about center front & the Sub woof copies.

Close all files.
Open AA multitrack.
Insert audio files top to bottom as (NAME):

Open up LF Rs into the Edit window. This is the point you would do any pre synthesis to get wrap around effect or whatever is desired. Right now lets focus on just decoding & I’ll come back to the pre synth stuff.

Select entire wave and go to Channel Mixer.
Select QS LF Rs preset & click OK. You have now just done simple QS decoding for the Left Front & Right surround channel.
Select right channel only and click on invert. Go back and select both channels again and get into the Center Channel Extractor.
Select the QS Crosstalk Reduce preset and apply. You have now reduced common mode signals between Left Front & Right Surround to very low levels. As was stated in the SQ note thread, it doesn’t take a lot of dB reduction to make a big change. Setting the Center Channel level to minus 10dB seems to be the sweet spot. As others experiment with this I really look forward to any further optimization on these settings. Two last things to do here. First go back to right channel only & click again on invert to put the polarity back where it was. Then go back to both channels and save. Saving frequently lets the temp files be freed up & keeps things working smooth. It also means you can’t undo after saving so make sure it’s done right & the way you want it before saving.

A note on the channel inversion before & after reducing crosstalk. This is important and can not be overlooked. Also it might seem that simply choosing “surround” for CCE to work with instead of “center” would do the same thing. It doesn’t, I tried it.

At this point you should be able to play back & check by listening to what you’ve done. The left channel should be louder than the right, with more direct sounds & the vocals in the left while there’s more ambience & reverb in the right.

So far so Good? Than go back to multitrack view and open up RF Ls in the edit window.
Again, at this point you would do any pre synth effects. Forget about that for a moment. Select entire wave and go to Channel Mixer. This time select the preset RF Ls and apply. Invert left channel only this time. Apply the same preset in CCE to reduce the crosstalk. Invert the left channel again & save after double checking it so far.

Go back to multrack view. If you are making a four channel mix then you’re essentially done and ready to export. If you’re making a 5.1 mix there’s still a bit to do. The LFE is made very simply, just channel mix 50/50 and low pass FFT filter at 80Hz or whatever works for you. I usually normalize the LFE to -10dB. This puts it in a range that can be easily adjusted with the sub woof level during playback to suit you.

The Center Front is always trickier than it seems it should be. It might seem at first that all you need to do is mix the channels equally, maybe do some high pass filtering & be done with it. Of course this also puts the hard panned left only & right only sounds center front where you don’t want it. So I suggest this. Open the CF file in the editor window and apply the preset labeled QS CF to the file. This will change the ratio of true center front to side sounds & enhance the CF by 10dB. Going beyond 10dB will cause noticeable artifacts. If anyone can experiment & get a nice clean Pro Logic like center front that that would be great. Let me know .At any rate, the phantom center front imaging is very good & unless I’m using the CF as a strong accent sound, then by the time I’m done I have it so low in level I just barely hear it, or notice it’s gone if I mute it.

Now all you need to do is export the files. This is done conventionally in the AA Multi Channel Encoder. Save them in the EXPORT file folder. HOWEVER! Because the four corner channels were created diagonally rather than as front & rear pairs, they will not be exported with the correct labeling. The export file labeled Right Front will actually be Right Surround. The export file labeled Right Surround will actually be Right Front! All the other channels will be named correctly. You can go back & simply rename the files or just think backwards & switch the files if you are creating DTS. My wife says I’m very good at backwards thinking so that’s what I do.

That wraps it up for actual decoding! I’ll be back soon for the final part.

"It's okay, were safe now. I sold my peoperty to pay property taxes", M. Jittlov '89
Last edited:

Sonik Wiz

2K Club - QQ Super Nova
QQ Supporter
May 30, 2005
Kansas City
As noted at the top of my preceeding post, I have edited & made a correction. Evidently this doesn't get the note bumpmed up so I'm doing it now to get get potential users attention.