[JoshK]

OK, before you go thinking this is a flame thread waiting to happen, I intend for this to be a FAQ thread, or atleast contribute to one.  Let's try to highlight the key features, differences, etc between the two and then we can add in others as we go (Behringer, et al).

DEQX:
- powerful DSP engine
- freq based room correction (time based RC promised in near future)
- speaker correction
- 3 way stereo digital crossover
- currently only accepts 16/44 (24/96 beta software ETA: end of month)
- uses FIR filters, allowing for group delay correction of steep slope xo's.
- includes ADC standard, options for digital outputs and/or balanced analog outputs.
- integral measuring mic & preamp

TACT:
- powerful DSP engine
- freq & time based room correction
- no speaker correction
- accepts up to 24/192 signal
- higher end models (RCS 2.2x) do 2 way stereo xo for sub integration
- uses FIR filters, allowing for group delay correction of steep slope xo's.
- options for ADC and DAC modules (standard passes digital out)
- works integrally with TACT digital amplification keeping signal in digital domain till the final stage
- integral measuring mic & preamp
- various aftermarket software & hardware enhancements available

Note: this isn't meant to be which is better, but rather to inform.

Feedback?  Any part of this incorrect?  Additional points?  I admit my understanding of the TACT features is somewhat limited.

[ss397]

this is a great thread, hope some users will share experiences. i think i am leaning towards DEQX, because my goal is DIY 3 way speakers. i want to try a clone version of the montana speakers when time and budget will allow.

[csero]

What is "time based room correction"?

[JohnR]

I was wondering that too.

[JoshK]

by time based RC, I mean time domain room correction.  The technical aspects of which I have no clue, I am repeating what I have read on the DEQX beta forum.   Could this be possibly related to decay time?

[csero]

I can only imagine that they will do a full or partitioned convolution with the impulse response in the time domain instead of the more common FFT and correction in the freq domain.

[JoshK]

I can only imagine that they will do a full or partitioned convolution with the impulse response in the time domain instead of the more common FFT and correction in the freq domain.

No idea, but I would love to read up on this stuff.  I need to find a good primer on the subject.  Too bad their software wasn't open source.

[ekovalsky]

Time based correction aligns the impulse responses of the driver groups in multi-way loudspeakers and/or the main loudspeakers and subs.  I should point out that the TacT does indeed have the capability to do this, but the measuring tool needed is obscured in the basically undocumented "dual domain" module and it is generally not an automatic process, at least if you want it performed properly.

I thought the DEQX had the ability to do this too, but maybe it doesn't ? Maybe John Ashman, mac, or other DEQX owners can comment.

You can time correct main speakers and subs with the RCS 2.2X pretty easily, as it will measure the physical distance and any phase discrepancies between the mains and subs on its own and correct them.  What it won't do automatically is calculate the additional delay needed by the mains when a DSP crossover is implemented.  DSP crossovers are more precise but slower than passive crossovers, and the output of the low pass can lag the high pass function by 1-30 msec depending on both frequency and slope! See the graph in my gallery to determine this delay then add it manually to the main delay figured by the TacT and you should achieve main-sub time alignment.  The graph only works for symmetrical filters (i.e. same frequency and slope on high and low pass).

Time aligning a multi-way speaker is more difficult, because you have to bypass the factory crossover and multi-amp.  The RCS with TacT digital amps, in conjunction with some powerful but free add-on software called TACS, does time and frequency based correction very well.  Based on measurements of each driver group and listening impressions, you can choose the ideal crossover frequencies and slopes and then adjust the phase/delay of each driver group so their impulses align perfectly.

Very, very few speakers are time aligned stock from the factory.  A few that claim and may actually be are time aligned include the top models from Vandersteen, GMA, Ascendo, and Wilson.   The vast majority of speakers that claim to be time aligned fail because of simplified baffle design and the use of 2nd or higher order filters, which require one or more drivers wired out of phase to achieve reasonable frequency response.  

Although the only way to build a truly time aligned speaker without DSP, first order crossovers severely stress the drivers as the are forced to operate cleanly over large frequency ranges.  DSP gives the benefits of first order networks (actually even better because of the zero group delay -- the derivative of crossover-induced phase change -- of FIR crossovers) to higher slopes that will either improve a driver's performance and allow a larger selection of drivers to be used.  

Here is an interesting system using a big collection of TacT equipment (with TACS) for frequency and time based correction.  The TacT gear replaced about $1,000,000 worth of Accuphase, Gryphon and Cary gear  

 http://www.aca.gr/pop_jyk.htm

[John Ashman]

Eric,
     Time-based room correction is notably different from simply time-aligning the drivers or phase correcting the signal.  It is an alternative and more sophisticated method of fixing room modes.  

I'm not quite sure I fully understand TBRC except that you could theoretically measure the reflected bass energy in the room and actively send out a cancelation signal based on a delayed, inverted and somewhat EQ'd signal.  In theory, it would also cancel dips, which FBRC can't do.  And you'd lose the sound of the room, not reduce the music in order to compensate.  It would reduce, even virtually eliminate, the "bbbbbllllluuuuurrrrr" in the room, which would notably tighten up the bass, making it sound more like what it would sound like outside with no room reflections.  But, to be honest, I'm not *exactly* sure if that's what it does.  It's just the only way I can imagine it working unless there's some sort of complex phase thing you can do.  

Anyway, I know NHT didn't bother to implement the current DEQX FBRC, choosing to wait for TBRC currently under development.  I guess they must see it as a major upgrade.  DEQX has TBRC in action, but they're refining it at this point for the PDC.  It would be a lot harder to develop and "perfect" than simple FBRC because of all the variables you could address.

[Occam]

http://www.meridian-audio.com/w_paper/Room_Correction_scr.pdf
The modes boss, the modes!
This is the approach used by Meridian as well as Rives.
FWIW