Picture of the BOZ Digital Amplfication system - Page 3

I also do know about Fourier Transforms and I guess we have quite similar background (although I left the field a while ago).

Please don't misunderstand me, I am a TacT fan, I am using their units for about 6 years, I had/have the 2.2x, TCS 1, TCS 2, five S2150 and three S2150X. When you know the limitations of the product there are often ways to work around them.

I have been in regular contact over the years with many of the Yahoo forum members, yes that's true, on the forum there was a lot of controversy but the energy that some of users put in understanding the TacT units is impressive (some will even talk about reverse engineering). Please have a look to the TACS software (which was by the way sponsored by the official TacT distributor in europe), it is truly a nice piece of code and Boz himself was surprised the day he first saw it (I was there that day).

You can also wonder why these softwares (TACS, GV, Acourate) were developped... some users tempted by the adventure of active multiamping had the feeling they were too limited by the original TacT software.

However, these developments were a double edge sword, on the one hand they helped opening new possibilities (you will notice that TACS contains a 2150 firmware not available anywhere else, this firmware was actually developped by Boz to help them with this development of multiamping) but on the other hand this knowledge also revealed some issues here and there. This created heated debates to know whether these were bugs or features :-) At the end, when the new generation of hardware was around the corner, TacT took the decision to move all the computation back inside the units (as in the 2.2xp) which most likely close the door for future third party developments. Now we have no other way than trusting what is inside the box. BTW Lyngdorf took the same approach for their RoomPerfect system.

Michel

GGA---I am certain that TacT has been doing phase response correction from day one. (Whether they were doing it right is another matter.) Simultaneous magnitude and phase/delay correction is exactly why time domain correction is done in the first place. Otherwise you might as well just equalize.

Hi

Could the TacT/Boz owners on this forum share their impression of the system in term of sonics... Not the improvement brought by the Room Correction Software, rather how do they sound to you...

As an aside is SigTech and the Z-? currently in production?

Hello Michael,

There have been some lively threads at Tact Yahoo about time domain correction. Rereading some of these shows that you are correct by one definition:

http://tech.groups.yahoo.com/group/T.../message/24362

First off realise by reading the above link that it's NOT possible to
completely correct for time as well as freq. It would only be possible
over a very very small area and you would need your head clamped in
order to benefit.

The standard implementation of only correcting frequency and letting
phase sort itself out is hard enough to do correctly that today there
are only a very small number of commercial products available
(Tact/Lyngdorf/DEQX/Sigtech) which can do this at all and most hifi
firms will claim that the technology is impossible.

To go to the next level and to carefully introduce some distortion in
one area to reduce phase distortion in another is a very, very subtle
science and there are a) no commercial implementations of such a product
and b) even in research areas you have only Denis's DRC program and
Uli's filter generator that even attempt such a thing. I had a chat
with the new chief engineer at Lyngdorf audio and he does not even
believe that it is possible to make these time based corrections...

So basically be happy that you have frequency response correction at
all. It's not trivial to do even that!

So if designing a good RCS system is hard, then designing one which can adjust phase is much, much harder and few people even believe that it is *possible* much less understand how to do it.

AND ALSO:

http://tech.groups.yahoo.com/group/T...3693?var=1&l=1

From theory a room pulse response can be split in two parts. One part is
minimum phase and the other part is excess phase. TacT does minimum phase correction only [my bolding]. Thus the excess phase is not corrected. The excess phase can be seen as an allpass filter where all amplitudes are transferred 1:1 but the phases are changed.

From theory the only correction here can applied by introduction of an
additional delay and the right phase inversion. The general opinion is that
excess phase correction is not possible (see e.g. description of Meridian
correction system). Anyway it introduces problems of dividing by zeros and
thus it can cause unstabilities.

The internal structure of the RCS as a multirate filter does not allow an
excess phase correction. So your goal is not achievable with the RCS.

Despite of negative opinions about excess phase correction IT IS POSSIBLE.
It requires some knowledge and a careful design. AFAIK there is no commercial system out there.

--------

The minimum phase describes a response with the shortest possible delays
necessary to reach a given amplitude response. So if you extract the minimum
phase part of a room response (e.g. by a homomorphic deconvolution = math
method) you have a signal that fully describes the amplitude response and
that has the shortest delays.

The inverse of such a minphase signal used as a correction filter corrects
also these shortest delays.

The excess phase is of course also measured. And the information is all
there for a possible phase correction. As said it is necessary to introduce
a bigger playbck delay with an excess phase correction. And so one reason
for not applying excess phase correction may be that you may loose lipsync
in combination with video. Another reason is tht the selected filter
structure makes it difficult. And finally a reason may be that it is not
well known how to handle it.
========================================

So it sounds like a matter of definitions. Tact does correct minimum phase but not excess phase. There's never a simple answer to anything.

Hello Frantz,

I use the 2.2X with the tape loop of my Placette active preamp. Setting the 2.2X in Bypass allows one to hear its sonic signature. In Bypass the signal goes through the complete Tact circuit but no corrections are applied but you can engage crossover, balance, and delay. There is a loss of transparency but it is not the end of the world.

However, once I correctly set up room correction the improvement in my room was so great that the loss of transparency is (for the moment anyway) of small concern. The goal would be to get that some frequency response but that would be quite a challenge.

I use Lavry Blue external DACs and ADC to replace the Tact's. One can chose any DAC or ADC to use with the Tact.

Sigtech is no longer being made. Z-Systems is alive and well and information can be found on their website (don't remember if they still make their eq). Lots of good information on the Robert E Green (REG) website, the TAS writer (and mathematician).

GGA, interesting posts. I'm frankly skeptical of their contents. One claim made is that full and exact phase correction is often not possible. That is indeed true. On the other hand, those posts claim that the RCS cannot do any excess phase correction due to their internal architecture; and it is that point on which I am particularly skeptical. It is possible to correct for nonminimum-phase behavior if you allow the RCS to introduce a strict time delay into the system (above and beyond its processing delay).

But at least it is clear that at least some phase correction is performed.

EDIT: Further reading of that thread makes it clear that the poster is not claiming that excess phase correction is never possible; indeed, he apparently has some software to compute the necessary corrections. He's only claiming that TacT doesn't do it. I'm still surprised by that, but as I don't own a unit I defer to those who do.

Michel,

Actually, signal processing isn't my field of expertise at all. The expert on
signals here is really Michael Grant. I have a rudimentary familiarity of
signal processing from unfolding the instrumentation response from
experimental data.

My background is actually nuclear physics, and computational physics.

Michael,

I think we share similar concerns. Although I don't have a detailed familiarity of
the actual algorithms implemented by the TacT RCS; I, like you; am skeptical of
some of the claims.

It would seem rather straight-forward to solve the delay problem by the
introduction of a time delay into the system. The RCS is after all meant for
2-channel applications; so it doesn't have to worry about syncing with a video
stream, for example.

The RCS can take all the time it needs - the listener is never going to know.

I'm skeptical that a device that's correcting the in-band temporal behavior of its
two pass-bands; would then turn around and introduce a large shift in the
timing between the two pass-bands.

The in-band time correction would be "guilding a lilly" in that case; and, as you
point out; there's no need for this inter-band timing difference when a delay
is all that's needed to accomodate the processing needs.

The all-digital TCS/Boz system brings some new issues that are not encountered in conventional analog systems.

Mixing speakers with different efficiency will bring out an issue that most people probably would not think about: Since the system is all digital, the dynamics of the whole system will be limited to the dynamics of the least efficient speaker! When the system is calibrated, the maximum level will be set such that 100% volume is 100% power to the least efficient speaker. (Big problem with the Magnepan CC3 which is ~9dB less efficient than a biamped MG20.1. I had to remove the CC3 from my system).

When purchasing TacT/Boz you'll need a lot of patience. Four of my five Boz amps had cosmetic and functional issues out of the box. (Yes, I purchased them from an authorized dealer). Now 3 months later, I'm still waiting for the last two amps to be fixed... That just sucks for a $30k purchase...

Peter

I would be understanding if it was shipping damange. Not so. FYI, the problem is ironically with the light 2200s (20lbs). The heavy 216 (80lbs) and TCS were flawless from day one!!!

Two of the 2200s had cosmetic defects from the factory. (Blemishes in front panel, one of the unit had a nut GLUED with Epoxy resin on the bottom!!!)

Two of the 2200s had functional problems. One unit had a dead left channel, the other one reports timeout problems, messing up the whole system.

When I sent in the unit with the dead channel for service (it was cosmetically flawless), it came back with THE FRONT PANEL SCRATCHED. The original problem was fixed, but instead it had a new problem: It would only power on every second time... 1 functional bad => 1 cometic bad + 1 functional bad

With all this waiting, it really feels very unprofessional of TacT...

Our home theatre is "down" since July when I sold my 9 Parasound JC-1s. It was supposed to go live in October when I received the TCS/Boz. Now it's still down... No movie in the Nielsen house since early fall... Waiting to get the Boz units fixed...

Did I say it sucks?!

Once I've solved all the problems, I may consider selling everything (due to bad after taste with all the problems I've gone through). If somebody wants to avoid problems, let me know. I will sell you the TCS,216, and 5 2200s for the same amont I paid. You will save all the agony an pain of getting brand new units fixed/replaced...

I feel that this is the worst purchase / most costly mistake I've ever made in the home theatre area...

Now, let me point out that I'm coming from Parasound, where 14 of 15 units arrive PERFECT out of the box, or 8 perfect JC1s out of 9. It is quite a chock to get 4 bad Boz2200 out of 5. (Quiz: The JC1 is $3500 and weighs 80lbs. The Boz 2200 is $3000 and weighs 20lbs. Which one should arrive cosmetically perfect?)

Peter