Vinyl "compensation" - applied only to SQ?

[kfbkfb]

IIRC, the vinyl compensation is better level matching between L(t) and R(t) prior to sending the signals to the SQ decoder.

Would the QS decoder benefit from better Lt/Rt level matching?

Also, I have one prerecorded compact cassette that may be SQ encoded, compact cassettes can sometimes have level differences between L and R too, will the SQ vinyl compensation help improve the accuracy of SQ decoding in that case?


Kirk Bayne

 

[Sonik Wiz]

IIRC, the vinyl compensation is better level matching between L(t) and R(t) prior to sending the signals to the SQ decoder.

Would the QS decoder benefit from better Lt/Rt level matching?

Also, I have one prerecorded compact cassette that may be SQ encoded, compact cassettes can sometimes have level differences between L and R too, will the SQ vinyl compensation help improve the accuracy of SQ decoding in that case?


Kirk Bayne

Good question. Accurate L/R input balance is important for both SQ and QS. Why let one mode be less correct than the other? Is this even possible with the circuit design of the SM?

I don't think I've ever read a good explanation of the difference between the original vs vinyl edition of the decoder. Maybe Chucky could illuminate us?

 

[chucky3042]

Dear All

SQ is a shit sandwich at the best of time with effectively a 3 dB window to store all the directional information. If there is any mis-tracking of the cartridge then the decode goes whoops out the window. So we had to take a few measure to reduce its sensitivity to this issue. QS on the other hand is way more tolerant in terms of its data window and really does not require any fiddling about.

 

[Sonik Wiz]

Checked around a bit & I felt Chucky shared some good info about the SQ Vinyl Version in Post #257.

 

[kap'n krunch]

Dear All

SQ is a shit sandwich at the best of time with effectively a 3 dB window to store all the directional information. If there is any mis-tracking of the cartridge then the decode goes whoops out the window. So we had to take a few measure to reduce its sensitivity to this issue. QS on the other hand is way more tolerant in terms of its data window and really does not require any fiddling about.

Hey, I totally understand your P.O.V. regarding SQ but then...is my hearing that bad? Cause I did some really decent sounding software conversions which look like were much more than 3dB separation between channels (I have said many times that my conversion of the 1st Santana LP came out so goo d that it was very close to the discrete SACD released in Japan..but then I have smoked a lot of weed! ) ...any elaboration would be extremely interesting to hear...

 

[chucky3042]

Hey, I totally understand your P.O.V. regarding SQ but then...is my hearing that bad? Cause I did some really decent sounding software conversions which look like were much more than 3dB separation between channels (I have said many times that my conversion of the 1st Santana LP came out so goo d that it was very close to the discrete SACD released in Japan..but then I have smoked a lot of weed! ) ...any elaboration would be extremely interesting to hear...

Hi
The 3 dB is the detection range in the stereo encode, the decode is still in excess of 30 dB.

Thanks Sonik.........I totally forgot about that SQ post!!

 

[Sonik Wiz]

Hey, I totally understand your P.O.V. regarding SQ but then...is my hearing that bad? Cause I did some really decent sounding software conversions which look like were much more than 3dB separation between channels (I have said many times that my conversion of the 1st Santana LP came out so goo d that it was very close to the discrete SACD released in Japan..but then I have smoked a lot of weed! ) ...any elaboration would be extremely interesting to hear...

I'm curious what software & what channels in specific? I've seen multiple apps/scripts over the years for SQ & most all do not have separation enhancement. In other words basic SQ decoding. In that case there will be high separation between front left/right, and rear left/right. But only 3dB separation between anything front or back.

And thus CBS developed 10/40 blend.

 

[par4ken]

I'm curious what software & what channels in specific? I've seen multiple apps/scripts over the years for SQ & most all do not have separation enhancement. In other words basic SQ decoding. In that case there will be high separation between front left/right, and rear left/right. But only 3dB separation between anything front or back.

And thus CBS developed 10/40 blend.

Only the most basic script has no separation enhancement. The others work by using the center channel extractor built into Adobe Audition. It works by removing common information from the signal and leaving the stereo information behind. By running it repeatedly at different phase angles the inter channel crosstalk can be almost completely removed. The only problem is that if too much filtering is applied it can sound unnatural.

I've listened to one or two of the Kap'n's conversions and can testify that they sound great. OD kept pushing his process/script decoding farther to extract more separation but I don't think that they always sounded better than that of the more basic script. Lucanu rules!

I was doing some conversions myself but since getting a second (working) S&IC and fixing my original unit I've lost most interest in script decoding.

 

[Sonik Wiz]

Only the most basic script has no separation enhancement. The others work by using the center channel extractor built into Adobe Audition. It works by removing common information from the signal and leaving the stereo information behind. By running it repeatedly at different phase angles the inter channel crosstalk can be almost completely removed. The only problem is that if too much filtering is applied it can sound unnatural.

I've listened to one or two of the Kap'n's conversions and can testify that they sound great. OD kept pushing his process/script decoding farther to extract more separation but I don't think that they always sounded better than that of the more basic script. Lucanu rules!

I was doing some conversions myself but since getting a second (working) S&IC and fixing my original unit I've lost most interest in script decoding.

I've had a few versions of QS/SQ scripts but have never done any ambitious LP decoding that way. Mainly I just wanted to test them & play around a bit. I am well aware of the Adobe Audition center channel extractor. I designed my own method for QS decoding that cancels common crosstalk on the diagonal chs, which is quite a bit different than other approaches. My intent wasn't really for QS decoding but to mimic the Sansui Synthesizer decoding mode. Put some out of phase blend from the AA Channel Mixer at the front end of the process & there ya go. Most of my projects were video related... for example all 14 of Chisato Moritaka DVD's in wrap around surround, Basia's Laserdisc, and the Wizard of Speed and Time feature film. The latter had lots of foley & added effects, dialogue panning etc. Done years ago & I just enjoyed it again last night. It holds up well.

Similar to you I've done very little PC oriented surround work in long time but for me it's because of the SM v2. Still, it's fun and a bonus to pre-synth things occasionally in AA3 to get enhanced decoding.

 

[J. PUPSTER]

I've had a few versions of QS/SQ scripts but have never done any ambitious LP decoding that way. Mainly I just wanted to test them & play around a bit. I am well aware of the Adobe Audition center channel extractor. I designed my own method for QS decoding that cancels common crosstalk on the diagonal chs, which is quite a bit different than other approaches. My intent wasn't really for QS decoding but to mimic the Sansui Synthesizer decoding mode. Put some out of phase blend from the AA Channel Mixer at the front end of the process & there ya go. Most of my projects were video related... for example all 14 of Chisato Moritaka DVD's in wrap around surround, Basia's Laserdisc, and the Wizard of Speed and Time feature film. The latter had lots of foley & added effects, dialogue panning etc. Done years ago & I just enjoyed it again last night. It holds up well.

Similar to you I've done very little PC oriented surround work in long time but for me it's because of the SM v2. Still, it's fun and a bonus to pre-synth things occasionally in AA3 to get enhanced decoding.

So are you guys saying that a SM can be improved on for SQ decoding?

I know some prefer the S&IC for that IINM. So after all this time perhaps the most excellent Quad (SQ) Tech heads here should join forces and come to some kind of consensus and build a "New" SQ hardware decoder.

For most SQ LPs, I'm in the Chucky camp that SQ is basically a shit sandwich!

 

[Sonik Wiz]

So are you guys saying that a SM can be improved on for SQ decoding?

Not if you ask Chucky but.... I think a 270 deg wrap around mode for stereo would be a real bonus & can be done in either SQ or Involve/QS mode. Much simpler to implement in the latter.

I know some prefer the S&IC for that IINM. So after all this time perhaps the most excellent Quad (SQ) Tech heads here should join forces and come to some kind of consensus and build a "New" SQ hardware decoder.

I think the quad tech heads in Australia have already done that! Never in history has there been a tri-band SQ decoder and with such clean sound & high separation.

 

[J. PUPSTER]

I think the quad tech heads in Australia have already done that! Never in history has there been a tri-band SQ decoder and with such clean sound & high separation.

Sort of joking about that, because it's tough to get this QQ crowd to have a consensus about what day it is

 

[par4ken]

For most SQ LPs, I'm in the Chucky camp that SQ is basically a shit sandwich!

Had Involve been designed for SQ right from the beginning Chucky might have a different opinion. It's not that simple to take a unit designed for one system and use it on another, there are bound to be compromises made. SQ Involve is very good all the same. The fixed blend used in even the best SQ logic decoders is undesirable to me, you don't get any of that blending with Tate decoding nor with Involve.

It was pointed out that centre front to centre back separation is less with Involve SQ mode compared to QS mode. That is not a fault of the encode system, there is no technical reason that the separation can't be just as high in SQ mode, just a design choice or the limitation of rigging a unit designed for QS to work on SQ?

 

[J. PUPSTER]

Had Involve been designed for SQ right from the beginning Chucky might have a different opinion. It's not that simple to take a unit designed for one system and use it on another, there are bound to be compromises made. SQ Involve is very good all the same. The fixed blend used in even the best SQ logic decoders is undesirable to me, you don't get any of that blending with Tate decoding nor with Involve.

It was pointed out that centre front to centre back separation is less with Involve SQ mode compared to QS mode. That is not a fault of the encode system, there is no technical reason that the separation can't be just as high inSQ mode, just a design choice or the limitation of rigging a unit designed for QS to work on SQ?

So you're saying their isn't two separate decoding systems (QS vs. SQ) in the Surround Master; but that the SQ part of it is derived from the QS somehow?

 

[kfbkfb]

IIRC, a post by Disclord (haven't found it yet) mentioned that SQ decoding suffered when there was a level difference between Lt and Rt and phono cartridges typically have have a spec of +/- 1dB L/R channel balance.


Kirk Bayne

 

[par4ken]

So you're saying their isn't two separate decoding systems (QS vs. SQ) in the Surround Master; but that the SQ part of it is derived from the QS somehow?

There are two different decoding systems in the SM, my point it the SM was designed for QS so the design would have been optimised from the start for QS. SQ was an afterthought, they had to work around that. That would be one of the reasons it's was such a bitch.

 

[Sonik Wiz]

IIRC, a post by Disclord (haven't found it yet) mentioned that SQ decoding suffered when there was a level difference between Lt and Rt and phono cartridges typically have have a spec of +/- 1dB L/R channel balance.


Kirk Bayne

For playing LP's my phono cart plugs into an Adcom GFA-555 preamp. I can easily adjust L/R balance from there. I put on a true monophonic record (haha usually an old Martin Denny from the 50's) and listen to only the rear chs from the Surround Master. Using either the QS/SQ mode seems to work equally well. Just adjust the L/R balance until the rear output is at a minimum. It's really quite a distinct squelching and my balance knob is just barely CCW off the center detent. When playing from Oppo 105 I just hook it direct to the SM as I've tested it & channel balance is excellent.

Many receivers new or old even with analog ins and outs don't let you do this. Another reason using separate components has worked out so well for me over the years.

 

[par4ken]

IIRC, a post by Disclord (haven't found it yet) mentioned that SQ decoding suffered when there was a level difference between Lt and Rt and phono cartridges typically have have a spec of +/- 1dB L/R channel balance.


Kirk Bayne

That specification of 1dB can be found in the Audionics SQ-106A manual. The better the balance the better the decode will be. The S&IC also has "Axial Tilt" which is compensation used to improve channel separation of your phono cartridge. The difference between 20dB and 30dB will not be really be audible in stereo listening but helps to better balance the two quadrature encoded back signals for optimum decoding. I have found that it makes a more noticeable improvement with lesser decoders than with the S&IC.

 

[chucky3042]

Had Involve been designed for SQ right from the beginning Chucky might have a different opinion. It's not that simple to take a unit designed for one system and use it on another, there are bound to be compromises made. SQ Involve is very good all the same. The fixed blend used in even the best SQ logic decoders is undesirable to me, you don't get any of that blending with Tate decoding nor with Involve.

It was pointed out that centre front to centre back separation is less with Involve SQ mode compared to QS mode. That is not a fault of the encode system, there is no technical reason that the separation can't be just as high in SQ mode, just a design choice or the limitation of rigging a unit designed for QS to work on SQ?

Actually the hardware did not constrain us, it really was a major re write of the software

 

[chucky3042]

IIRC, a post by Disclord (haven't found it yet) mentioned that SQ decoding suffered when there was a level difference between Lt and Rt and phono cartridges typically have have a spec of +/- 1dB L/R channel balance.


Kirk Bayne

I fear Disclord died from diabetes, has anyone confirmed that?