PC Based SQ DECODING - ALMOST DONE!

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I'm still having problems with the rears being too loud, +3 db is not enough for the fronts, +6 db seems more like it. Is this correct? After all I did normalize down by -6 db.

In the normalize step I go:
Edit > Group Waveform Normalize and click tab 3 Normalize. I check "Normalize to a level of" and use -6 db as per the instructions. I have also left the factory presets of AA 1.5 alone and therefore "Use Equal Loudness Contour" is selected as well as "Use Limiting". It works well enough but is this correct or even matter?

Also, should I clean up my audio files before I convert or is it better to clean up the 4 mono signals after I convert so the SQ encoding doesn't get screwed up?
I have MAGIX audio cleaning lab 10 or is there something in AA 1.5 that's better?

Thanks in advance for any help here. I've read this entire thread many, many times already. It prints out to 40 pages of tiny print in case you we're wondering.
Jim
 
Last edited:
The front/rear balance also depends upon the original mix, it could be that the rears were mixed higher than 'normal'. I recently switched to using AA 2.0 and had to play with the levels a few times before I got them right. I just use my ears. I don't like changing the settings on my receiver, they are set so almost all of my DVD-As and SACDs sound balanced front to rear. So, I just played with the levels untill I got what I wanted. Usually I raise the fronts 3db more than the rears, but if 6db sounds right to you than I wouldn't worry that you did anything wrong, that's just the way that conversion sounds best. My last conversion sounded best with the front raised 4db more than the rears - at 3db I found myself leaning forward while listening.

As far as cleaning up the .wav files, I do it before the conversion. Some people do it the other way, but I see two main reasons to do it first 1) it's easier - 4 mono files would take 4 times as long a 1 stereo file; and 2) it seems to me that the clicks and pops could throw off the decoding. I remember one instance that I had a nasty patch of noise that I fixed after the decoding because it was easier, but otherwise I do everything first. Enjoy! I know that this can drive you insane, but it's supposed to be fun :banana:
 
Thanks BananaSlug for the advice. I'm having a blast with this! I've only had 3 hours sleep in the past 24 hours, but it's extremely rewarding preserving these Quad mixes.

After the normalize step the waveform looks much more expanded than before. After running the script the rear waveform still looks expanded, but the front.wav looks normal on the 2 conversions I've done so far.
Jim
 
Last edited:
Also, should I clean up my audio files before I convert or is it better to clean up the 4 mono signals after I convert so the SQ encoding doesn't get screwed up?
I have MAGIX audio cleaning lab 10 or is there something in AA 1.5 that's better?

I always declick before converting. I use the Click Fix for Adobe Audition plugin. I use the normal 33 1/3 setting for the whole album, and then a stronger setting for quiet sections, such as the final chord of a song. If I can avoid it, I don't use any noise reduction at all, before or after decoding. That's because the SQ processing already adds a bit of burbling artifacts to the sound, and the noise reduction process adds even more artifacts. However, I am using the noise reduction filter that comes with AA 1.5. I haven't tried MAGIX.

J. D.
 
Thanks J. D. MAGIX only costs about $10 and did wonders on a record that was as scratchy as sandpaper. But as you mentioned before, noise reduction can change a recording in a negative way.
 
Is there a way to scientifically automate the Front / Rear volume so that both are exact? Or is that impossible due to variants in each record being decoded?
 
Last edited:
Would anyone be able to create a definitive (locked even?) thread which just has how to do the SQ and QS conversions step by step with no other chatter e.g. how it was done etc?, this thread is going a bit off topic with noise reduction etc. I think a specific thread on noise reducion would be good. I raised the question a month or two ago about using magix and I think I came to the conclusin do as little as possible to the SQ source before the decode.
 
Would anyone be able to create a definitive (locked even?) thread which just has how to do the SQ and QS conversions step by step with no other chatter e.g. how it was done etc?

I agree with ingresman. We need a locked step by step definitive thread on how to do SQ and QS conversions (with links to diagrams, zip files and MP3s) with commentary about when and how to use noise reduction and volume levels. People new to doing conversions have to sort through 8 pages of thread that prints out to 40 pages of tiny print. Thanks!
Jim
 
Last edited:
We need a locked step by step definitive thread on how to do SQ and QS conversions (with links to diagrams, zip files and MP3s) with commentary about when and how to use noise reduction and volume levels.

How about a separate html-file on QQ?

-Kristian
 
Thanks so very much first and foremost to everyone that has actually worked on the AA (Adobe Audition) SQ and QS Decoding projects, and second to Ingresman for asking for a final conclusions type post and Old Quad Guy for taking the time to make that type of post happen!

OQG's thread is the most helpful overall post for newbies I've seen so far in this discussion.

I've been wading through the SQ and QS AA Decoding threads for several hours just reading all the discussions, not totally grasping all the mathematical theories behind each process but enjoying the discussion anyway, and it's easy to get lost in the extra banter that gets randomly injected by readers looking for quick and simple results, wanting an argument, or making needless and oversimplified comments that serve as no real contribution (almost like this post, admittedly) rather than leaving the thread of conversation by the experts flowing smoothly to help them get from start to finish and make sure they've ended up with one or two scripts, hopefully for both AA1.5 and 2.0 that they all agree on.

I was trying to read the entire thread looking for the purest resulting script(s) rather than the fanciest sounding/enhancing ones and I'm still not certain which is which yet but I've at least got a much better idea where to begin.
 
If you PM me with your email address I can email the scripts to you for both AA1.5 and AA2,0.

Not sure if either of them will actually work with CEP 2.0 but you can give it a shot.

For the record, both the 1.5 and 2.0 scripts provide excellent results but the AA2.0 scripts are much faster with the same results as 1.5 IMO.

That goes for everyone here....
 
Open that file in the edit view. Then go to effects/filters/graphic phase shifter and apply a -90 degree shift with the Channel set to "both."

When I apply a -90 degree shift, the back-left and back-right channels are swapped in the result. When I apply a +90 degree shift, it's correct. Did I overlook something? I am using the Quadrafile disc set (from Hi-Fi News & RR Magazin) as a source. Regards
 
When I apply a -90 degree shift, the back-left and back-right channels are swapped in the result. When I apply a +90 degree shift, it's correct.

Ever since Cool Edit Pro, I've been ``encoding'' as it were from 4-track half inch stage tapes using -90-degrees phase shift for the left rear and +90-degree phase shift for the right rear and they play great on the DPL in the car, or at the movie theatre I used to work at where we jerry rigged it so we could play the records used for between shows through the surrounds (something usually not done). I read eons ago on the web that the +/- 90/90 was necessary to maintain matrix balance, and so the rear channels would be 180-degrees opposite EACH OTHER, but still be out of phase enough from the front channels to decode properly.

Just curious.


Now, regarding separation vs. fidelity issues.

Has anybody tried what the movie studios do?

On titles such as White Christmas with Bing Crosby, where the separate vocal and effects tracks have been destroyed,
they'll take the monaural low-fidelity optical mono composite sound master with the dialogue music and effects, and mix that in with the monaural but high fidelity music-only master that they have, and then mix THAT in with the 4-track stereo mid-fidelity sound elements recorded back off an old Vista Vision magnetic release print with its' narrow tracks and splotchy tracking. So you end up with mono low fidelity D, M&E, so you have the dialogue and effects, high fidelity monaural M, so you have a good copy of the music and then a mid-fidelity 4-track separation of everything all blended together at once.

So I was wondering if anybody tried that with an SQ and a Q-8 , SQ and QR or SQ and CD-4 of the same title, retimed back to itself and laid on top? I realize the retiming work would be extensive, but it would be interesting to see if one could be laid over the other to achieve both fidelity (SQ) and separation (CD-4, Q-8, Q-R) or vice versa or verse vicea.
 
Last edited:
In your opinion "fidelity" belongs to SQ and not to QR?
 
Jesus - this is getting technical - but also VERY exciting! Recently Phonogram Germany released some cd's, SZENE STAR stating SURROUND SOUND on the cover... surely there is surround in it, but they don't state which would be the correct way of decoding the material... I have e.g. Vicky Leandros' greatest hits with this technology... any ideas?

all the best from
Copenhagen, Z
 
In your opinion "fidelity" belongs to SQ and not to QR?

Surely not!
@ ndiamone - QR fidelity is IMHO far superior compared to any vinyl format at all - sure, tape sheds & loses it's oxides if not carefully looked after (although I must admit I have never heard of consumer reels shedding - only "master" grade tape, ironically enough) but vinyl degrades every time it gets played.

Can I get it straight in my mind what you are referring to here as well? the post reads almost as if you are thinking about taking discrete titles & creating matrixed versions. That would mean using DPL II most likely otherwise mass compatibility just went right out of the window.
Whilst this can indeed sound like an attractive idea - indeed I have considered not using a dedicated stereo mix but an Lt/Rt matrix on projects myself, but rejected the idea after testing showed that not only does the surround mix lose far too much of it's discrete character and separation, but it also heavily compromises the stereo as well.

Or am I simply getting this all wrong?
 
Back
Top