Hi All
We have been busy examining this controversy today and as it turns out we/ you may all be correct, we have a confusing picture.
First we used 3 test QS and SQ sources for our tests these can be found in this dropbox:
https://www.dropbox.com/sh/3jpgx0qhaxcfgzy/FiiHs7DykE
Please download these.
The SQ Q4 Vinyl test tones.wav is the CBS Q4 test disc and is as the name suggest recorded on vinyl.
The SQ test tones.wav is a "script" based series of tones and is 100% accurate in terms of magnitude and phase - "by the book"
Its test tone sequence is:
Centre Front - Right Front - Centre Right - Rear Right - Centre Rear - Left Rear - Centre Left - Left Front - Centre Front
The QS test tones.wav is also a "script" based series of tones that are 100% accurate in terms of phase and magnitude- also "by the book".
Its test tone sequence is:
Centre Front - Left Front - Centre Left - Rear Left - Centre Rear - Right Rear -0 Centre Right - Right Front - Centre D
Also included below are the test results we obtained from a Surround Master randomly pulled out of stock.
Please note for our tests today we had a noise floor of -42.2dB we actually needed more dynamic range for comprehensive results but the results are fine for what we are testing.
The results for the QS script are remarkably symmetrical showing a separation of generally better than 35 dB, this reduced to approx 13 dB for central sides. Level discrepancies were less than 0.5 dB. In short it looks like all levels are balanced for SQ and Involve mode.
Now things get more interesting:
For SQ test tones.wav script based encode again the results were very good showing very symmetrical results with 30- 35 dB separation, the worst results were as expected (for SQ) the side center symmetry that showed the center side position could be pulled to the rear by about 10dB. This is one of the week aspects of SQ. It is one of the reasons SQ cannot achieve stable side central images (unlike QS- perfectly linear). In short- no sign of problems with the correct encode format.
For the Q4 Vinyl test tones.wav the results are relatively poor showing a leakage of fronts to rear amounting to only 9 - 12 db separation IN THAT DIRECTION - all other directions the separation was better than 25 dB.
So TAB - YOU ARE CORRECT!!!!!!!!!!!
Now the issue is WHY?
Upon examination of the CBS Q4 test disc we found that the encode was in error. SQ largely picks direction based on Left/ Right inter channel leakage. We found leakage in the L/R channel of say -15db and a phase error of 45 degrees. On the script based encodes these parameters are precise. This CBS disc was recorded from vinyl onto CD and it is most likely this leakage and phase shift is typical of most magnetic cartridges and RIAA equalisation curves.
At the time of developing the SQ Surround Master we were under significant "pressure" from you know who- the great Oxforddickie to ensure we were in fact decoding "BY THE BOOK". Well folks the test results today clearly show if the encode source is correct (as it will be for script based sources supplied by Oxfordickie and Bob Romano) you will get the full 35 db separation in all directions. For Vinyl sources (yes I fully accept this is the majority) the results will not be as good and dependent on how good your magnetic cartridge is in terms of separation and phase shift.
All our development work was with the script based encode sources and not with the Vinyl as we wanted to be "by the book". We note that most users of the SQ SM have reported as good or better results than the Tate yet now we have 2 users who clearly prefer the Tate. I suspect this is for 3 main reasons.
1 Given the non script - vinyl sources used they were only getting say 9 - 12 db separation in the front leak to rear direction.
2 Our previous (and controversial) tests indicate test audiences cannot pick separations above 12 dB and so clearly we fall under this and any other level issues such as amplifier pot mismatch, seating centrality, speaker sensitivity will tip the system over the edge so you will clearly notice the leakage.
3 Some people are more sensitive to separation than others - for example I listen more to things such as clarity, transient response and am less picky on separation.
Also the bulk of the users who state that they prefer our SQ SM to the Tate have reported issues of clarity and transient response. This observation is a direct result of the SQ SM being a full tri band decoder unlike the Tate that is single band.
SO WHAT ARE WE GOING TO DO???
Looks like the Tate (being designed in the late 70's) was designed specifically with vinyl in mind and has been adjusted to allow for a much smaller level sensing range (it pushes harder to the edges) than the SQ SM that has been designed for script based encoding.
I offer the following:
1 Fully money back upon return of the SQ SM if unsatisfied.
2 We will schedule some research time in say 2 months to review the internal settings of the SQ SM to have it optomised for vinyl not script - for those who want it.
3 We will send an updated memory chip in the mail to all those who want the revision - so it can be installed by a qualified technician. Probably $20
4 We offer a free update if you pay for the return transport to our factory.
WHAT NOW?
We will continue to offer the existing SQ SM for sale as I really think it is the most accurate for script based encodes but hopefully we will offer two setting versions in a few months time. My recommendation is patience but remember the version set for vinyl will not be as linear as the current version and may be more prone to pumping (as several Tate users have reported). SQ is a VERY fiddly format and is not forgiving on encode source errors- unlike QS, unfortunately there is no "one size fits all" solution.
I know this is an imperfect solution but on the bright side at least our decode is "by the book".
Regards
Chucky