[ekovalsky]

Sorry I've been scarce around here of late, my life has been very busy and I've been devoting more time to family as well as other hobbies.  But I have been immensely enjoying audio too, and I thought I'd give an update on my system.

I recently received a customized TacT S2150 to replace the Crown K2 on my sub towers.  In addition to the extensive modifications my other three S2150 have, including upgraded MOSFETs, this amp has new inductors in the output stage. 

All room/speaker correction, delays, and level adjustment is now done in the amps.  The RCS is used in bypass mode, so basically it is just a glorified digital volume control and digital signal router.  The recently released, latest version of TACS software added several key features including correction limits (essential for dealing with nulls!) and target curve manipulation.  Also the measurement module is now much more powerful, far surpassing what the RCS 2.2X software can do.

I am also using a AudioControl mic that has essentially flat response, eliminating the need for a calibration file.  It is a great alternative to the Earthworks mic which I have almost bought a few times, just to shy away because of the cost.  The TacT mic is far from flat in the upper octaves, and the provided calibration data is usually very bad.  If you're using the TacT mic with its calibration file, you probably are not getting as good a result as you should. 

For each channel I initially measure with no crossover, except on the tweeter where I use a 60dB/octave high pass filter @ 2khz for safety.  Then TACS uses these measurements to create the output filters, which are calculated based on the specified crossover, correction, target curve, and EQ.  You can specify the range over which correction is active (I cut off the tweeter correction at 8100hz), specify how to handle transitions, and also limit correction levels.  The latter is a new feature and very useful in dealing with room nulls, like the main one I have around 55hz. 

Here are my filters, all based on 24dB/octave Linkwitz Riley type crossovers at 50, 400, and 2700 hz.






Then program the filters into the amps, adjust each amps volume control (no loss of bits!) to match driver output, then determine time delays.  Then perform a global measurement and ... here is the combined time, frequency, and phase plot from the full range pulse:



The measurement is pretty but more importantly, how does it sound ?  OBSCENE !!!  I am amazed every time I hear it particularly with live recordings -- I close my eyes and I am there!  I started on the all-digital route about two years ago and finally have reached an end point that I'm totally content with. 

By the way the CD I'm currently listening to -- Muso Ko by Habib Koite & Bamada -- is some amazing world guitar music.

[warnerwh]

It appears you've got as close to perfect as the real world will allow. I'm wondering if after you're done using the equipment you need to do any tweaking by ear? 

[zybar]

Very cool stuff Eric.

Glad to hear that you are enjoying the system and that all that hard work has paid off.

George

[ekovalsky]

Most of the "tweaking by ear" was done in the past with the RCS when I would play with different target curves, and also via adjusting crossover points and slopes to best suit the drivers. 

After a lot of experimentation, I ultimately chose crossover frequencies very close to what the manufacturer had designed with his passive boxes, but the first and quasi-second order slopes are replaced by fourth order slopes, while maintaing ideal phase behavior.  I was able to lower the mid-tweeter transition from 3500hz to 2700hz which reduced comb filtering artefacts, resulting in a smoother vertical axis frequency response.  Steeper slopes are available but didn't sound as good, probably because of more prominent ringing artefacts.   

Basically at this point I've optimized individual driver as well as global frequency response and correctly time aligned each group, and minimized the effect of the room.  Depending on the program material, the system can sound like a sensational pair of headphones or a live venue.  I feel like I've taken this big system about as far as it can go -- now I'm just enjoying the ride. 

I am rather amazed at how the system is usually setup.  Drivers are driven beyond their comfort zones because of the low slope crossovers.  This increases distortion and limits power handling.  Time alignment is basically ignored.  There are complex room interactions between the vented subwoofers, sealed woofers, and dipole midranges.  Integration of the subs (driven by an electronic line level low pass crossover) and main channels (driven full range) is totally random.   And there are some comb filtering issues between the midrange and tweeter lines, because of the fairly high crossover frequency demanded by the Raven R1 when used with a low slope high pass.  And still HP seems to like them 

[warnerwh]

Well you've done what I've only dreamed of   This does seem to be the most intelligent way to optimize a sound system. If it weren't for the cost I'd do it myself. Just a fully treated room and some digital equalization helps substantially. Getting the time, phase and comb filtering effects minimized I can imagine would also have a very positive effect.

The reason I ask about the ear tuning is actually twofold. One because a single microphone is not two ears separated by several inches with hopefully a brain in the center . Also because even though a target curve has been input it may not be quite what you want or expect. With the vagaries of recordings it may be hard to hit a bullseye with any certain room curve. This no doubt could take a significant amount of experience from experimenting.

The Tact 2.0S is getting into my affordable zone. At this point I'm very pleased being able to enjoy music the way I can now. The attributes a Tact unit has is at this point mystical to me. I'm eager to get one. Even this little Behringer unit however has been very educational and helpful. I just have to figure out a way to squeeze out a couple of extra grand somewhere for a used Tact as I couldn't afford the latest model.

No doubt dsp is the future. All speakers and rooms can benefit from dsp. Brian will be implementing the DEQX into the Vmps speakers soon. With even cheap little ht receivers having some room correction capabilities built in I'd hope the high end could keep pace.

Thanks for the info. It's been very interesting.

[Val]

I am also using a AudioControl mic that has essentially flat response

ekovalsky, is this the microphone you're using? Do AudioControl dealers sell it?

[ZooDog]

Very impressive looking, and I'm sure sounding, system.  I have NHT Xd's so I can appreciate the benefits that DSP crossovers and EQ can bring to a speaker, albeit on a much smaller scale.  I applaud you for embracing the kind of technology that it seems many audiophiles with your means seem to shy away from.

[zybar]

I am also using a AudioControl mic that has essentially flat response

ekovalsky, is this the microphone you're using? Do AudioControl dealers sell it?

I know the question was directed at Eric, but the answer is yes.

I ordered the mic and will be receiving mine later this week. 

Hopefully it will prove to be an improvment over the stock TacT mic + calibration file.

George

[warnerwh]

What's wrong with the mike that Tact supplies?

[zybar]

What's wrong with the mike that Tact supplies?

It is a relatively inexpensive mic and like anything, can be improved upon.

George