[fsimms]

I joined several other Salk owners when I bought George’s(zybar’s) Tact 2.2X room correction preamp last week.  I have Salk HT1’s so I needed this Tact version for the sub output.  The difference is great.   The Salks sound much smoother and realistic. The bass is much better.  The Salks are much more forgiving of recordings.  I never dreamed that there could be so much difference.  A friend who is very familiar with my speakers said that they sound much sweeter and are very relaxing.

As a caveat, the Tact requires a lot of understanding.  I am just beginning to learn how to use it.  As one person put it, “it is a journey and not a destination”.  However, if I had to listen to just what I have done, so far, forever it would be a bargain.

[zybar]

Great news Bob!!

I am glad to hear that right off the bat you are reaping the improvements that room correction can deliver when properly implemented.

George

[95bcwh]

Bob,
 Wait till you try my new DAC card for the Tact... aa

Rgds
barry

[DMurphy]

So what kind of correction is being applied?  Is it just to the bass region below, say, 100 Hz, or is it contouring into the highs?  Does it give you a visual image of the corrections?

[ctviggen]

It performs full spectrum correction, with graphs.  See:

http://www.tactlab.com/Products/RCS22XP/RCS22XPAutomaticRC.html

I thought it used MLS, but I cannot find exactly how they are taking the spectrum measurement.  They create a filter that's applied to incoming data in the time domain.

[fsimms]

So what kind of correction is being applied?  Is it just to the bass region below, say, 100 Hz, or is it contouring into the highs?  Does it give you a visual image of the corrections?

It corrects the sub and my Salks.  It also does the crossover to the sub.

It also has a target curve that it can match the final response too from 20Hz to 20kHz.  You can edit the target curve to any response that you want in a visual editor. 

You can see the crossover, correction, target and measured response curves in the visual editor.

This is one of zybar's target curves that he used for his HT3's.

[DMurphy]

Hmmmmmm   That's a pretty strange target.  Most rooms will boost the bass irregularly, and tilt down the highs above 10 kHz.  The plot looks more like a pre-correction smoothed measurement than a target. 

[fsimms]

Hmmmmmm   That's a pretty strange target.  Most rooms will boost the bass irregularly, and tilt down the highs above 10 kHz.  The plot looks more like a pre-correction smoothed measurement than a target.

This is a chart of what the in room response would be that the microphone would hear.  Since you put the microphone at the same place that your ear is, that would be the same thing that you would hear.  If, without correction, you say that in a room the base would be boosted and the highs would be tilted down then this is almost the same response that you say that you should hear. Just without the bumps.  That is exactly the reason that the owner of Tact, Peter Lyngdorf, says that they came of with a curve near this.  Peter says that this mimics the response that you would hear in a large room from a flat speaker.

[csero]

It performs full spectrum correction, with graphs.  See:

http://www.tactlab.com/Products/RCS22XP/RCS22XPAutomaticRC.html

I thought it used MLS, but I cannot find exactly how they are taking the spectrum measurement.  They create a filter that's applied to incoming data in the time domain.

I spoke wit Radomir Bozovich, the designer of TACT. He is not using MLS. He is using 3 independent impulses for 3 bandwidth limited measurement and assemble the full spectrum measurements from the result. In the HF the pulse is only 1 sample wide and he is using very short gating (~1 ms), in the bass the pulse is 128 samples wide and the gating is 2 sec, the midrange imp is in between. The whole cycle is repeated lots of times to increase the S/N.
You can argue a lot about his methods (validity, S/N ratio etc.) but other methods are also have their own problems measuring response farfiled in a reflective environment.

[ctviggen]

I wondered how he got the low frequency response without the deleterious affects using of MLS in the low frequency band.  His techniques seem as good as any in order to measure small rooms.  It's a very complex subject.

[ctviggen]

Woops!  That should be "effects" not "affects".  Dang grammar!

[zybar]

Hmmmmmm   That's a pretty strange target.  Most rooms will boost the bass irregularly, and tilt down the highs above 10 kHz.  The plot looks more like a pre-correction smoothed measurement than a target. 

It is a pre-measured target, not a measurement.

TacT calls this a "target curve".

George

[zybar]

Hmmmmmm   That's a pretty strange target.  Most rooms will boost the bass irregularly, and tilt down the highs above 10 kHz.  The plot looks more like a pre-correction smoothed measurement than a target.

This is a chart of what the in room response would be that the microphone would hear.  Since you put the microphone at the same place that your ear is, that would be the same thing that you would hear.  If, without correction, you say that in a room the base would be boosted and the highs would be tilted down then this is almost the same response that you say that you should hear. Just without the bumps.  That is exactly the reason that the owner of Tact, Peter Lyngdorf, says that they came of with a curve near this.  Peter says that this mimics the response that you would hear in a large room from a flat speaker.

The above is not what the microphone hears, but rather what I was striving for with this particular "target curve".

George

[DMurphy]

The above is not what the microphone hears, but rather what I was striving for with this particular "target curve".

Now you're losing me.  If you achieve the target, why won't that be what the microphone hears?  Unless that's not a target, but rather a mapping of the corrections being applied. 

[95bcwh]

The Tact microphone is only used to measure room response without any digital correction, we call this "measured response".

We can display the "measured response" using the Tact software, unfortunately it wasn't shown in the picture above. Then in the Tact software, we define a "Targe curve" or "Target response", this is the curve displayed in the picture above. The Tact software will then generate a digital filter (we call this "correction filter") by looking at the difference between the "measure response" and the "Target response". When we play the music, before the digital signal gets to the DAC card, it first get "filtered" by the the "correction filter".

I don't know if there's a way to use the same microphone to measure the "target response" after you've applied the correction filter.

I'll posted a picture of my room response and the target curve when I get home tonight.

[zybar]

The Tact microphone is only used to measure room response without any digital correction, we call this "measured response".

We can display the "measured response" using the Tact software, unfortunately it wasn't shown in the picture above. Then in the Tact software, we define a "Targe curve" or "Target response", this is the curve displayed in the picture above. The Tact software will then generate a digital filter (we call this "correction filter") by looking at the difference between the "measure response" and the "Target response". When we play the music, before the digital signal gets to the DAC card, it first get "filtered" by the the "correction filter".

I don't know if there's a way to use the same microphone to measure the "target response" after you've applied the correction filter.

I'll posted a picture of my room response and the target curve when I get home tonight.

Well said Barry.

There is not a way to to measure the sound after applying a correction filter unless you have a second TacT unit.

George

[brj]

It seems odd that you wouldn't be able to measure the corrected response, but is it just a matter of software?

Could you not run a signal generation and measurement package other than the TacT specific software on your laptop, feed a sweep to the TacT - which it will correct according to your target - and then measure it with the same alternative software?  (You could use something free like Room EQ Wizard or something commercial like ETF, etc..)

[DMurphy]

I understand that.  You confused me by drawing a distinction between the target curve and the final measured curve.  One is a goal, the other will be the actual result.  That's fine.  But they should be quite similar if the unit is working.  That said, I'm still not clear on why you would want to keep the room response boosted in the bass and attenuated in the upper treble just because that's how it would sound in a large room.  I take it that "large room" is really large.  What kind of assumptions are made about the wall, floor, and ceiling surfaces?  I assume the room isn't infinitely large.  I realize this is a very complicated area and there isn't any single answer.  But if seems like the result may still sound heavy, even if it's not as heavy as the uncorrected speaker.  And then there's the issue of what the original mics in the recording were hearing.  If large auditoriums suck up the highs and boost the highs, then there will be a double effect in the room.  But I'm getting a headache.  Thanks!

[95bcwh]

See above example, the top two curves are the left/right microphone measurement of my room response. The big dip is room mode and cannot be tamed.

The bottom curve is my target curve.

[fsimms]

I understand that.  You confused me by drawing a distinction between the target curve and the final measured curve.  One is a goal, the other will be the actual result.  That's fine.  But they should be quite similar if the unit is working.

Yes, I think the target and the final resultant curve, which the Tact can't show, are supposed to be almost exactly the same over most of their range.  This final resultant curve would be what you would hear at the microphone or ear. 

That said, I'm still not clear on why you would want to keep the room response boosted in the bass and attenuated in the upper treble just because that's how it would sound in a large room.  I take it that "large room" is really large.  What kind of assumptions are made about the wall, floor, and ceiling surfaces?  I assume the room isn't infinitely large.  I realize this is a very complicated area and there isn't any single answer.  But if seems like the result may still sound heavy, even if it's not as heavy as the uncorrected speaker.  And then there's the issue of what the original mics in the recording were hearing.  If large auditoriums suck up the highs and boost the highs, then there will be a double effect in the room.  But I'm getting a headache.  Thanks!

When I look at the raw measured curves from my Salk HT1'S, that I get before any processing, I get a curve.  This curve is measured at my chair.  This curve has roughly the same shape as the previously posted target curve.  My room is just about 500 square feet with a 9 foot ceiling.  This curve has a lot of wiggles.  The biggest wiggles are below 100 Hz.  Buy running the Tact, with the proper calculated corrections, the response smooths out at the ear.  At least it subjectively smooths out.

But I'm getting a headache.  Thanks!

I am leaving out as many details as I can so your headache doesn't get worse.