SQ Decoding question....

[Bob Romano]

So... let's say I got a few SQ or QS records I want to turn into nice sounding DVD-A's.

I record the records into the computer at 96/24. So as to not ruin the SQ or QS encoding...what would you do for the next step(s)?

a) remove any ticks/pops?
b) do a "restoration" of the vinyl (light as possible - i.e. surface noise reduction)
c) decode the SQ or QS... then do the cleaning
d) something I hadn't thought of....

I have the beautiful scripts for Audition that do a killer decode of the SQ and QS, I won some groovy records on Ebay, now I want to see what I can do.

I won:

Cat Stevens - Foreigner SQ
Barbra Streisand Live SQ
Leonard Bernstein's Mass SQ
Deodato - Prelude SQ
Loggins & Messina - Native Sons SQ

So any suggestions would be most appreciated....

 

[Lucanu]

I would work in two passages,
1) if your records are in a good/perfect condition, then I would go to your c)

2) for anything else do b)

At the end do a click/pop remove from your decoded front/rear wave files.

...maybe you should do a test (1 minute of sample would be fine) with both options to see if your results are different.

 

[Cai Campbell]

I always decode prior to clean-up. It's more effort to clean up two or four wave files instead of one, but I'm always afraid that I will introduce decoding anomolies if I clean up a matrix file before decoding.

 

[Dylan Berichon]

Decoding should always be the last step of the process, whether you use a hardware or software decoder.

You should always clean the matrixed source audio first. The better your matrix source sounds prior to decoding, the better your decoded audio will sound.

Don't forget that any noise/clicks end up getting decoded too. Clicks and pops especially can make Adobe Audition's Center Channel Extractor do some really strange things. And forget about trying to remove clicks and pops after you've run your source through the center channel extractor. You'll have to do all declicking manually, which can be a real pain, and even then it won't sound quite right.

I know that a lot of declick filters probably won't declick a decoded file properly, especially rear channels.

Another thing to consider is that if you run a decoder over a noisy source, you've just multiplied the amount of processing you're going to need to apply to your material, and your noise reduction tools won't work as well.

Ultimately, make sure your recordings are of the best possible quality they can be from the get go, and then clean up the audio. Decoding is always last.

 

[Dylan Berichon]

I always decode prior to clean-up. It's more effort to clean up two or four wave files instead of one, but I'm always afraid that I will introduce decoding anomolies if I clean up a matrix file before decoding.

The noise that is (typically) contained in the matrixed source will likely introduce much more destructive anomalies then any repairs would.

In fact, repairs to the audio, if done right, will not introduce any problems. They should actually make it decode that much better.

With SQ, for example, phase shifting a click 90 degrees (which is what happens with SQ decoding) is going to make declicking it much more difficult.

The idea is that you want to get your matrix source as close to the original master tapes as you can, and then decode it.

 

[Cai Campbell]

With SQ, for example, phase shifting a click 90 degrees (which is what happens with SQ decoding) is going to make declicking it much more difficult.

This has not been my experience. I've never had problems with cleaning up an audio file after decoding. Perhaps it doesn't really matter which is done first (decoding or cleaning).

Of course, I only start with superb source, so there isn't going to be much noise to begin with. I won't even consider a matrix LP for conversion if it isn't in pristine playback condition to begin with. But I still have trouble with the concept of manipulating an audio signal prior to decoding.

Maybe for a marginal record, with lots of noise, clean-up before decoding might be a better approach, but then again, I can't help but think that the more "heavy-handed" the clean-up process, the greater the likelihood of introducing decoding anomolies. I mean, my theory is that the more pristine the original source signal (noisy or not) the better chance the decoding process has of doing its job correctly. Any audio clean-up tactics will alter that signal and will potentially make the decoding of that signal sub-optimal.

Yes, any noise is going to propagate to the decoded channels, but then you can clean those up individually, and again, it has been my experience that matrix-decoded channels present no greater clean-up challenge than regular ol' stereo/matrix channels.

 

[Dylan Berichon]

This has not been my experience. I've never had problems with cleaning up an audio file after decoding. Perhaps it doesn't really matter which is done first (decoding or cleaning).

Of course, I only start with superb source, so there isn't going to be much noise to begin with. I won't even consider a matrix LP for conversion if it isn't in pristine playback condition to begin with. But I still have trouble with the concept of manipulating an audio signal prior to decoding.

Maybe for a marginal record, with lots of noise, clean-up before decoding might be a better approach, but then again, I can't help but think that the more "heavy-handed" the clean-up process, the greater the likelihood of introducing decoding anomolies. I mean, my theory is that the more pristine the original source signal (noisy or not) the better chance the decoding process has of doing its job correctly. Any audio clean-up tactics will alter that signal and will potentially make the decoding of that signal sub-optimal.

Yes, any noise is going to propagate to the decoded channels, but then you can clean those up individually, and again, it has been my experience that matrix-decoded channels present no greater clean-up challenge than regular ol' stereo/matrix channels.

Like you said, it really does depend on the amount of "damage" that's been done to the audio in the first place. Since you are only using high quality sources to begin with prior to doing conversions, you are right that it most likely won't make too much of a difference which order it comes in.

I would think that with some hardware logic decoders, if the noise was bad enough it could screw with the "steering" of the sounds. As you said though, this is likely only going to be a problem with a pretty marginal LP.

I'm at a (rather early) stage in Quad Collecting that I'm happy just to even have a particular title in my possession, regardless of it's condition! (Okay, that's not entirely true, but I have settled for a few titles that weren't in the best shape just so I could have a copy of them.) For these titles that aren't so pristine, I still maintain it's best to apply noise reduction first.

It may be that for some noises it would be better to remove them before decoding, while for others it's better to remove them after words. It's tough to say for sure, but from what I understand about the matrix systems (which admittedly isn't as much as some of the other great minds that participate on this forum) it just seems like it's best to clean up the audio before decoding.

Ultimately I'm sure we all agree that the biggest factor in quality when it comes to decoding these albums (after the condition of the physical source it self) is how good the "capturing device" (in most cases the record player) is. It's surprising how much of a difference a better stylus can make when it comes to this!

I'm going to have to try to do this both ways to see which one sounds better to my ears. I'll report back my findings when I do get around to doing this.

 

[Bob Romano]

thanks for the help... can't wait to get my records!!!

 

[Bob Squires Jr]

Decoding should always be the last step of the process, whether you use a hardware or software decoder.

You should always clean the matrixed source audio first. The better your matrix source sounds prior to decoding, the better your decoded audio will sound.

Don't forget that any noise/clicks end up getting decoded too. Clicks and pops especially can make Adobe Audition's Center Channel Extractor do some really strange things. And forget about trying to remove clicks and pops after you've run your source through the center channel extractor. You'll have to do all declicking manually, which can be a real pain, and even then it won't sound quite right.

I know that a lot of declick filters probably won't declick a decoded file properly, especially rear channels.

Another thing to consider is that if you run a decoder over a noisy source, you've just multiplied the amount of processing you're going to need to apply to your material, and your noise reduction tools won't work as well.

Ultimately, make sure your recordings are of the best possible quality they can be from the get go, and then clean up the audio. Decoding is always last.

I have to agree there....
Considering I try to start off with the best possible source I can find...
I do a manual large click and pop removal by going over the source file in
say... sound forge or adobe and isolating any major pops I hear during
transfer (which I drop a marker on) and doing away with them (carefully of course).
I then run a "light" click filter over it (I have had good luck with wavepurity)
to eliminate any little annoying clicks.

I do believe Kempfand, Neil Wilkes and Bob Romano can attest to some of
my decodes coming out rather nicely using this method.

I have tried to clean after decode but decoded pops and clicks X 4
can be a nightmare at times.

Besides, wouldn't a pop filter only work on pops and have little if anything to do with matrix phasing?
If so, light click removal on the 2 channel master shouldn't affect the outcome of the decode, no?

Of course if you are decoding an LP that looks like the cat had a go at it,
it wouldn't be worth shite to decode anyway so no worries there on heavy filtering!
I've never tried on one and wouldn't bother anyhoo since scratched vinyl makes me cringe!

-Bob

 

[Bob Romano]

Welcome there Bobby!!! It's about time you joined us here. I can attest to Bob's SQ decodes... Very nice indeed!!!