[ekovalsky]

Here are my final TacT results with the VMPS RM/X and Alon EGR II.  The only variable is the speakers -- room treatments and listening room are identical.  Room treatment basically follows VMPS guidelines with extensive absorption (120sq ft of 2" fiberglass panels) on the front wall and front half of the side walls.  There is no treatment on the ceiling. so this is a partial LEDE design.  Floor is carpet over concrete slab.  Room dimensions are 14.5' x 17' with long wall speaker placement.  Ceiling is 10'.

Both speakers were initially set up using the manufacturer's recommendations.  For the VMPS, speakers were about 16" from the front wall and a few feet from the side walls with the side woofers firing outwards.  Toe in was added until the speakers cross-fired about 1.5' in front of the listening position.  I did not have subs with the RM/X as they are supposed to be full range.  The main output of the RCS 2.2X drove the midrange and tweeter, while the sub output drove the woofers.  The internal 1st order low pass filter to the woofer and 2nd order high pass filter to the planars were bypassed, and a DSP crossover was implemented by the TacT using symmetric 4th order slopes at 250hz.  Bass cutoff (-6dB) of the VMPS speakers in my room seems to be 35hz, thus I did not have any usable output in the bottom octave.  YMMV.

The Alons are a two piece per side system.  The subs were placed in the corners angled slightly toward the listening position.  The mains are dipole above 400hz and I had the baffle about 4.5' out from the front wall.  Speakers were about 3.5' out from the side walls.  No toe in was used initially.  The main output of the RCS drove the mains, which are not designed to be full range by themselves with their acoustic 8" woofers, via the external passive crossover.  The subs are driven from the sub output of the RCS 2.2X rather than using the supplied DQ-LP1 crossover, which is practically an antique!  I am using symmetrical 8th order DSP filters at 60hz.  I'd like to cross even lower but there is a pesky room dip at 54hz which I have to work around.  The bass cutoff of the subs (-6dB) is about 20hz with usable output to below 16hz.

I spent quite a bit of time repositioning both the speakers and the listening position (microphone) to learn how they interacted with my room and subsequently find the ideal locations in it.  After about six hours with the Alons, they got a little bit of toe in and also moved slightly closer to the front wall.  

Much more time with the VMPS, well over 40 hours in total, as their relationship with the room was significantly more complex.   They could not be moved further out into the room because the bass needed as much wall reinforcement as possible to keep up the planar/tweeter section (with the L-pads dialed down to 11-12:00).  Ultimately some improvements were made by fine tuning the distance to the side walls, and best results were obtained with the factory placement suggestions.

Here are screencaps of the CORRECTION and DUAL DOMAIN screens for each set of speakers.  The X & Y axes are scaled identically so the curves are directly comparable.  The VMPS screencaps were done a few months ago on my wife's computer which has a 1280x1024 display.  The Alon screencaps are more recent, done on my laptop which has a 1920x1200 display.  Thus the Alon screencaps are larger and have a widescreen 16:10 aspect ratio instead of the standard format 5:4.

For both speakers I created target curves that tried to follow the output of the speaker above 200hz and incorporate some mild EQ to my taste.


VMPS measurement


VMPS correction filters


Alon measurement


Alon correction filters



It is worth pointing out that these speakers sell at vastly different price points.  There is no question that both benefit from the TacT.  I don't believe there is a speaker system made that is too good for the TacT, nor do I know of a better way to integrate mains and subs.  

I do think the Alons will benefit from a larger room, hopefully the extra space will expand my placement options so that the 54 and 100hz room dips are eliminated or minimized.  Doing so will allow me to lower the main-sub crossover frequency and raise the target curve level, respectively.  Also I need to reassess the room treatments, as the Alons have different needs than the VMPS.  I'll likely move some of the absorption to the listening end of the room and add diffusors on the side walls and behind the speakers.

[zybar]

Thanks for taking the time to post the data.

I think as more people get to hear systems that have room and/or speaker correction they will begin to understand how powerful and essential these tools are to maximizing their audio enjoyment and investment.

George

[DVV]

Thanks for taking the time to post the data.

I think as more people get to hear systems that have room and/or speaker correction they will begin to understand how powerful and essential these tools are to maximizing their audio enjoyment and investment.

George

I agree, George.

Those of us who had even simpler equlizars 30 years ago, and took the time and trouble to learn how to use them rather than misuse them, and again took the time and trouble to use them to best correct for room deficiencies, have known this for 30+ years.

Cheers,
DVV

[Danny Richie]

Hey Eric,

Maybe I am not following this completely. Can you explain your measurements to me?

Are the top curves before correction and the lower ones after correction?

Are we seeing left speaker and right speaker measurments on each chart?

On the top charts are each line (horizontal dotted line) representing 9db?

Then on the lower chart each dotted line repreents 5db?

[hometheaterdoc]

Hey Eric,

Maybe I am not following this completely. Can you explain your measurements to me?

Are the top curves before correction and the lower ones after correction?

Are we seeing left speaker and right speaker measurments on each chart?

On the top charts are each line (horizontal dotted line) representing 9db?

Then on the lower chart each dotted line repreents 5db?

Both speaker's measurements are being shown on the same graph as well as the proposed correction curve of what the response will be after correction has been applied.  That is for the first screen cap for each speaker

For example, take the first screen cap of the VMPS speaker.  Look at the "legend" on the screen cap next to the word View.  The orange line is the left speaker's mid and highs measured response.  The yellow line is the right speakers mids and highs measured response.  The light blue line is the left woofer's response.  The dark purple line is the right woofer's response.  The pinkish/purple lines represent the digital crossover that you can employ using the Tact.  The lime green line with the red dots on it is what's called the "target curve".  This is the response you want the complete system to have (included room interacations) once you apply the room correction filters.

The second screen cap for each speaker is an alternate view that the Tact software provides.  To quote the manual:  

The Dual domain screen is designed to provide basic time and frequency analyzation of the measured data and correction filters. This software package does require basic understanding of signal theory. It is useful to use as a verification tool, to make sure that room correction is performed correctly. The screen is divided into two parts. The upper part is used to view the signal in the time domain, and the lower part is used to view the signal in the frequency domain.

I hope that helps...

[ekovalsky]

Hey Eric,

Maybe I am not following this completely. Can you explain your measurements to me?

Are the top curves before correction and the lower ones after correction?

Are we seeing left speaker and right speaker measurments on each chart?

On the top charts are each line (horizontal dotted line) representing 9db?

Then on the lower chart each dotted line repreents 5db?

Hi Danny

There are two graphs for each speaker.  

The first one (measurement) shows the measured frequency response in-room at the listening position without correction.  The blue and purple traces are for the subs, or woofers in the VMPS.  The yellow and orange traces are for the main channels, or mid and tweeter in the VMPS.  The purple lines display the DSP crossover selected, but not yet implemented thus not reflected in the measurement curves.  Each horizontal divison is 9dB.  

This graph also shows the selected target curve in green, with orange construction points.  Adjusting the target curve gives you complete control over EQ, if any is desired.  I have found my personal preference to be a gradual rise below 300hz, reaching just over 2dB at 125hz.  Then a small rise around 800hz for some added vocal presence and a modest dip around 2.5hz for a slightly less forward presentation.  This EQ basically followed the natural response of the Alons which is probably why I liked them so much when I heard them uncorrected.

I forgot to mention that on the VMPS measurements, I was giving en extra 6dB boost to the woofers as the S2150 driving them was set at +18dB gain and the mid/treble was set at +12dB.  If both amps were set to the same output level, the woofer output would be decreased by 6dB beyond what is shown.  Adjusting the L-pads would move the entire 6-20khz or 200-6khz bands up or down but did not change the response pattern within those bands.  

The second graph is the Dual Domain screen, with I labelled as "Correction Filters". There are actually two separate graphs, the upper one showing the time/impulse response and the lower one the frequency response.  The correction filters are reflected in the lower frequency response graph after being calculated from the previous measurements, the target curve, and the selected DSP crossovers.  The pink and red traces reflect the sub (or woofer in the VMPS) filters, the blue and green traces the main (or mid/tweeter in the VMPS) filters.  Each division is 5dB.  Since this is a DSP correction, the signal is changed by attenuation and there is no boost -- you really cannot go beyond 0dB without clipping.  For the VMPS up to 20dB attenuation was used across much of the midband to reach the target, to correct the rising output measured in this region. For the Alons just under 10dB cut is used almost uniformly above 200hz to correct for room dips in the bass region.  The uniformity or non-uniformity of these curves above 200hz reflect how much correction is actually taking place -- in other words, the straighter the line the better.  In every room the filters will be active below 200hz because that is where bass nodes and dips will exist, those being the primary targets for DSP correction.

Basically the more correction is getting done the more the sound will change.  When bypassing the correction on the Alons, I get room-induced bass boom at about 40hz and 70hz but otherwise the sound is about the same.  In fact, the sound in bypass mode is great with just two bands of parametric EQ used in lieu of correction to tame these nodes.  

With the VMPS, the entire character of the speaker changes when correction is enabled because of the large amount of signal manipulation being employed throughout the audio band.  To me, the VMPS speakers sounded much better with correction than in bypass mode.

Hope that clarifies it a bit

Unfortunately the TacT RCS 2.2X.  does not offer a direct way to recheck measurements once the correction filters are in use.  However, with the TacT amps you can do this by entering the correction filter directly into the amp itself, then remeasuring.  I did do this and the resulting measurements follow the initial target curve exactly -- indicating that the filters work as intended.

Let me know if there are more questions.

[hometheaterdoc]

yeah.... what he said  He be much better writer than I be........

[ekovalsky]

yeah.... what he said  He be much better writer than I be........

Hometheaterdoc,

Did you ever get any TacT measurements on the Usher or von Schweikert speakers ?

[ekovalsky]

There is an interesting system where you can see the TACs software -- which enables the RCS correction filters on the amps -- here.  It shows what the TacT is doing better than I can explain it.  The graphs are near the bottom of the page in his most recent update.  

Awesome setup this guy has!  Don't miss his pre-TacT setup (update 2/9/2002) which feature a museum-like collection of Accuphase and Gryphon equipment.  He probably could by another chateau with the proceeds from the sale of that ultra expensive gear!

[JoshK]

Eric,

Are you currently, or planning, to do the xo's in the amps?  I think this is a VERY good idea and I only wish I had the funds to buy 3 TACT amps to try it for myself.

[Danny Richie]

Thanks for the clarification guys. It all makes good sense now.

Wow, that first one sure needed a lot of digital correction. Looks like a +/-13.5db response from 1kHz and up. Eeeeeek

It looks like there was a broad peaked area from about 800Hz to about 4kHz but in the corrected response it looks like a pretty broad dipped area through that range.

I am just wondering why it is that it still looks so rough even after correction.

The Alon, by contrast, looks great even without correction.

I have always believed that digital correction in ranges above 200Hz were just band aiding a problem. Allow me to clarify that statement too.

If there is a peak at 2kHz for instance that is caused a side wall reflection. The digital correction system can attenuate that area to balance the output level in that range. However, it may have a balanced level but some of it still consists of a stronger than normal off axis wall reflection. This is actually a delayed time arrival that can be perceived as a smear.

A better solution would be to use some sort of room treatment to attenuate the peak and the direction it comes from, and at the same time do away with the smear created by the delay time.

However, with a speaker having an on axis response that is this far from accurate I guess the advantages of the digital correction in the upper ranges far out weigh the disadvantages. Room correction will not correct for peaks and dips of that magnitude.

In the lower ranges though, below 200Hz, I think this type of system is really killer. Interesting reading. Thanks for posting the reponses.

[ekovalsky]

Eric,

Are you currently, or planning, to do the xo's in the amps?  I think this is a VERY good idea and I only wish I had the funds to buy 3 TACT amps to try it for myself.

Yes.  I expect to do this later in the year.  I may purchase another S2150 used, or I may sell the two I have and upgrade to a six channel BOZ amp or Blade amp (supposedly available this fall).   If a black S2150 shows up at a good used price I'll go that route for the time being.

Everyone I've talked to who has substituted the active DSP multi-amping for passive crossovers has been thrilled with the results.  So even though I have no complaints about the Alon crossover, I want to give it a try.  With the TACs software it should be easy to have the speakers produce a phase and time aligned, flat frequency response which can then be manipulated by the RCS for room correction and EQ.  Much like the French guy did  

[ekovalsky]

I am just wondering why it is that it still looks so rough even after correction.

The Alon, by contrast, looks great even without correction.

I hav ...

Danny, that curve is not post-correction -- rather it is the filter itself, showing the attenuation needed by frequency to form the target curve.  If you take the measurements and sum it with this curve, the result should be the desired target.

Flat in-room response was a goal for Carl M. in desiging this speaker, and  he seems to have done it successfully.  I don't know how it would measure in the nearfield or in an anechoic chamber, ao that is not how I listen I don't really care  

[Danny Richie]

Danny, that curve is not post-correction -- rather it is the filter itself, showing the attenuation needed by frequency to form the target curve. If you take the measurements and sum it with this curve, the result should be the desired target.

Ah, I see. It makes much better sense now.

Flat in-room response was a goal for Carl M. in desiging this speaker, and he seems to have done it successfully. I don't know how it would measure in the near field or in an anechoic chamber, ao that is not how I listen I don't really care

With a line source you really can't measure it in the near field and get any useful information. Even in an anechoic chamber you have to get really far away, farther away than most chambers will allow anyway.

From the response measurements it looks like he did a nice job.

[JoshK]

I'll be sure to let you know how my experimentation with DSP active amping goes, just as soon as I finish up my amps.

[ekovalsky]

I'll be sure to let you know how my experimentation with DSP active amping goes, just as soon as I finish up my amps.

Please keep me posted on your progress Josh.