GIK Acoustics 244 Sound Panel Review

[RLA]

This is one of the best discussions we have had here in a long time.
Can you believe you are all getting this kind of education for free?
I feel like a student with multiple professors prepping me for the big exam.
Thanks to all for contributing.

 

[Ethan Winer]

Gerry,

Thanks for your comments, and I'll address as much as I have time for. I also look forward to meeting you at HES 2006 when we'll both be on the same acoustics panel. Maybe this discussion can serve as a warm-up for that.

> I was always haunted by high-end two channel systems I installed without side wall treatment that sounded pretty darn good. <

Sure, I've heard systems that sounded "pretty darn good" even in rooms with no treatment at all. A room that's wide enough can get away without side-wall first reflection treatment, and a room that's large has a fewer bass problems than a room that's small.

> Mirror point absorption improves the sound of speakers with poor off-axis response but reduces sound quality and envelopment (soundstage width detrimentally) for speakers with good off-axis characteristics. <

You need to define what is "good" off-axis response. To me, this is a speaker that radiates over a wide horizontal range, with a frequency response that remains relatively consistent over as much of that width as possible. Is this what you mean, or are you talking about speakers that have an intentionally narrow radiation and thus don't "splash" much sound toward the walls in the first place?

> Reflections also, most assuredly reduce the audibility of comb filtering due to time alignment errors. <

Click HERE to see the response I measured in my living room with and without absorption at the side-wall first reflection points. Then tell me how the "without" response could possibly be considered better.

> Listening to two speakers then moving in and out of the sweet spot is all the evidence you need of the audibility of this comb filtering. <

Agreed, and I show people this all the time in my living room HT. Stand behind the couch, out of the Reflection Free Zone (RFZ), and the music is plainly heard coming from the left and right speakers. Then lean forward over the couch into the RFZ and all of a sudden everything opens up, becomes much clearer, and much larger both vertically and horizontally.

Likewise, even though my center speaker is just below the screen of my RPTV, sitting on the couch in the RFZ you can't even tell where the sound is coming from. It seems to be coming right from the actor's mouth in the middle of the screen. Note that I have RFZ panels on the ceiling as well as on the side walls, and I also have a reasonably absorbent carpet. So this is a total RFZ, not just half-done with side-wall panels only.

> This is not up for debate. I haven't read a reputable scientific discussion of room acoustics in some time that does not acknowledge that a properly tuned parametric EQ reduces the resonance (ringing) of modes in a small room. <

Of course it's up for debate! I have argued with "EQ reduces ringing" folks many times, and not once has anyone ever shown me hard evidence that EQ reduces ringing in practice. Again, I've seen Floyd Toole's oscilloscope display showing reduced ringing so I know it's possible in theory. But without knowing what happens a few inches away, how long it took to tweak the EQ, and how stable that tweak remains over time, it remains theory. As soon as someone shows me that EQ can reduces ringing a meaningful amount over an area of at least one square foot, I will jump ship to your side.

> Ringing causes skewed response. <

Well, almost. These are separate issues that almost always go hand in hand. Usually, whatever causes the skewed response (a room mode, an EQ boost) is also what causes the ringing.

> As to bass, I prefer a tuned trap (Helmholtz, diaphragmatic absorber etc.)which focuses on specific frequencies more than a broad band absorber that lowers bass level indiscriminately. <

A tuned trap is fine when you have a specific problem mode and it's too low in frequency to target effectively with porous absorption. But using tuned traps exclusively is a mistake because it ignores that all rooms have problems at all low frequencies. Not just those related to the room dimensions.

> The perception of increased bass caused by absorption is because the treatment lowers the level of midrange energy making the bass (we are less sensitive to) appear louder.

No!

The perception of more bass is usually due to raising half a dozen nulls that had been 20 or more dB down.

> Bass traps are a very good thing but they shouldn't be sold as the only thing. <

I mostly agree with that. I have 38 traps in my living room HT, and when I replaced my Carver Sunfire sub with an SVS earlier this year I used the SVS's one-band cut-only parametric EQ to tame the room's second-order axial length mode. The EQ did help, but it was already truly fabulous before. Many of my friends are professional recording engineers and professional musicians. They all tell me my living room system is the best thing they ever heard. (Not to rely only on "arguement from authority.")

> a truly flat response would not be to most folks liking. <

I disagree with that strongly. Most people (including me) do not like a system that has been equalized in order to make the response measure flat with test gear. That's a far cry from not enjoying a flat response per se.

--Ethan

 

[Ethan Winer]

Jeff,

Thanks as always for contributing great info from your vast knowledge. Every time I discuss anything with you, I walk away knowing more than I did before.

> the impedence "tube" <

Good point. Does anyone have an impedance tube large enough to measure a 2 by 4 foot panel?

> testing an absorber with ETF in a bedroom will only reveal what that one device does in that situation. <

Well, sure, but if you measure three devices in the same session a test like this will clearly show how well each absorbs at the room's own resonant frequencies. So if a room has modes at 70 and 103 Hz, you won't be able to measure at 85 Hz. The room I've used for this type of testing has modes spaced well enough to be quite useful. And I don't have to pay IBM $1,000 every time I want to test something!

> What varies by 50%? <

See ASTM C423, page 7, describing the variations they measured in round robin tests. At 125 Hz absorption varied as much as 50 percent. You can be sure that at lower frequencies it varies even more. Also, the MiniTrap tests at Riverbank you did while at Auralex (which you have said you stand behind) show three times more absorption at 80 Hz than what we measured at IBM. You measured 22.44 Sabins at 80 Hz, and IBM measured 7.44. Now, this was for corner placement, not a standard C423 test.

> that much variation is not out of the question when the same samples are measured in different labs <

Yes, that's all I meant.

> I'm sure Ethan and Glen can show that 5000 bass traps also virtually eliminates modal variation! <

ROF,L. Click HERE to see what I was able to achieve with 38 traps in my living. And yes, this is with a few dB of cut-only EQ at ~45 Hz using the EQ built into my SVS subwoofer.

--Ethan

 

[Glenn Kuras]

>See ASTM C423, page 7, describing the variations they measured in round robin tests. At 125 Hz absorption varied as much as 50 percent. You can be sure that at lower frequencies it varies even more. Also, the MiniTrap tests at Riverbank you did while at Auralex (which you have said you stand behind) show three times more absorption at 80 Hz than what we measured at IBM. You measured 22.44 Sabins at 80 Hz, and IBM measured 7.44. Now, this was for corner placement, not a standard C423 test.<

Ethan, as I call you a friend you are, but you know (because you told me) that the test that was done at Riverbank with 2 miniTraps was BS. In fact if you would like for me to post you talking about it on other forums I would be more then happy to post that.
I do understand that GIK got a great review (which we are very proud of), but please sir (I am saying please) don't take this time to spread something that you call BS yourself. I really think it is in poor taste.

Thank you,
Glenn

 

[Vaughan Odendaa]

I saw your response Ethan. It looks great. Only another 15-20 more traps to get a flat response to within a 3 dB window.

BTW, Glenn, are there any pictures of your traps ? I've been waiting a long time and I haven't seen too many pics of your white traps. Please post some pics if you can (I've seen all the pics on your website).

Thanks.

--Sincerely,

 

[Glenn Kuras]

I saw your response Ethan. It looks great. Only another 15-20 more traps to get a flat response to within a 3 dB window.

BTW, Glenn, are there any pictures of your traps ? I've been waiting a long time and I haven't seen too many pics of your white traps. Please post some pics if you can (I've seen all the pics on your website).

Thanks.

--Sincerely,

We finally got a web person that can DO THE JOB!! (anyone else have web guy problems? ) So there are a bunch of reviews I have been sitting on that we need to post with pictures.

Glenn

 

[Ethan Winer]

Glenn,

> you know (because you told me) that the test that was done at Riverbank with 2 miniTraps was BS. <

Yes, good point, and I agree that data is BS, or at least suspect. In this one case I figured since it's Jeff's own data and he has told me he "stands behind" it, it was fair game to use it this one time in a reply to Jeff. But you're correct the data is at least suspect because only one fourth the required material was tested. Not that I want to open up that tired old discussion again here! And it wasn't even necessary for me to bring that up because "off by 50 percent" is already explained well enough in the ASTM document.

Glenn, thanks for keeping me honest.

--Ethan

 

[Glenn Kuras]

I will do some more searching myself and email them to you. I am sure I have loads of them in my email.

Glenn

 

[Savant]

> I was always haunted by high-end two channel systems I installed without side wall treatment that sounded pretty darn good. <

Sure, I've heard systems that sounded "pretty darn good" even in rooms with no treatment at all. A room that's wide enough can get away without side-wall first reflection treatment, and a room that's large has a fewer bass problems than a room that's small.

> Mirror point absorption improves the sound of speakers with poor off-axis response but reduces sound quality and envelopment (soundstage width detrimentally) for speakers with good off-axis characteristics. <

You need to define what is "good" off-axis response. To me, this is a speaker that radiates over a wide horizontal range, with a frequency response that remains relatively consistent over as much of that width as possible. Is this what you mean, or are you talking about speakers that have an intentionally narrow radiation and thus don't "splash" much sound toward the walls in the first place?

> Reflections also, most assuredly reduce the audibility of comb filtering due to time alignment errors. <

Click HERE to see the response I measured in my living room with and without absorption at the side-wall first reflection points. Then tell me how the "without" response could possibly be considered better.

(Italics above aren't mine - I can't figure out where they're coming from...)

I figure I might as well jump into this discussion, too...

I feel care must be taken here. Much of the "bad press" that early side-wall reflections have received over the years has been in the context of critical listening environments - recording studio control rooms, audiophile listening rooms, mastering control rooms, etc. In that context, a neutral acoustical environment is typically the goal so that recorded content can be evaluated and/or appreciated in its truest form. When we delve into the world of home theaters, 6+ channel audio, sound fields, surround, etc., we are no longer as concerned with reproduction for the sake of creation or artistic appreciation. We are now into the realm of reproduction for the sake of entertainment. This paradigm shift is important. The ultimate goal of many a home theater is to bring the movie-going experience into the home. Therefore, the rules change. Early reflections may serve a purpose, particularly since the source material is often skewed heavily towards dialog. In that context, early reflections from surfaces within 4'-6' of the listener can often be deemed "pleasing," when evaluated subjectively. I.e., that's the sound that is expected.

Back when I was at Auralex, we performed a rough experiment once where we took a small portion of a movie soundtrack - one of the LOTR movies, I believe - and processed it using two different impulse response files measured in the same room before and after treatment. When the audio (only) was played back to a member of the Auralex staff not as well-versed in "what good acoustics sound like," the listener's preference was for the "before" track. I.e., they preferred the sound of the movie played back in the untreated room. This was a telling example of the aural expectations people have when they wish to be entertained and aren't listening "critically."

Therefore, what a line on a graph is saying about comb-filtering becomes much less important than what the end-result is expected to sound like. Go to any movie theater and there is ample comb-filtering in most seats. This is the most common frame of reference people have for "what a movie is supposed to sound like." This is the experience people wish to bring into their home. If it sounds objectively "better," that may not translate subjectively "better" in the contexts of home theaters. Most people want clear dialog and earth-shattering bass effects. Floyd's 5000 subs will definitely give the latter. For the former, the fact that a majority of the dialog is being funnelled through a single loudspeaker actually should make things easier. A good center channel with a wide dispersion will give the illusion of better sound around the whole room. Too much absorption on the side walls could actually be detrimental in this case.

A happy medium, IMO, is a mix of absorption and reflection/diffusion. Personally, diffusion can get muddy if the walls are too close. And the absorption tends to sound better if it is of the flat-panel (fiber or foam), or stretched fabric wall variety. This type of absorption is non-linear with respect to angle-of-incidence/reflection. Absorption decreases as you move off-axis, which is a typical configuration for side-wall treatments. Therefore, the sound tends to be better - subjectively - than side walls treated with an "equal" amount of sculpted material, like "wedge" foam.

There are other ways to skin this cat, as well. If Glen doesn't mind giving up a little of the thunder in this thread, RPG, Auralex, and MSR (Tony Grimani's company) all have good offerings in this vein.

> This is not up for debate. I haven't read a reputable scientific discussion of room acoustics in some time that does not acknowledge that a properly tuned parametric EQ reduces the resonance (ringing) of modes in a small room. <

Of course it's up for debate! I have argued with "EQ reduces ringing" folks many times, and not once has anyone ever shown me hard evidence that EQ reduces ringing in practice. Again, I've seen Floyd Toole's oscilloscope display showing reduced ringing so I know it's possible in theory. But without knowing what happens a few inches away, how long it took to tweak the EQ, and how stable that tweak remains over time, it remains theory. As soon as someone shows me that EQ can reduces ringing a meaningful amount over an area of at least one square foot, I will jump ship to your side.

I believe you've been referred to papers elsewhere that demonstrate this. My own personal recommendation to you would be to read The Loudspeaker-Room Interface — Controlling Excitation of Room Modes by Rhonda J. Wilson, Michael D. Capp, and J. Robert Stuart, from the AES 23rd International Conference, Copenhagen, Denmark, 2003 May 23-25.

> As to bass, I prefer a tuned trap (Helmholtz, diaphragmatic absorber etc.)which focuses on specific frequencies more than a broad band absorber that lowers bass level indiscriminately. <

A tuned trap is fine when you have a specific problem mode and it's too low in frequency to target effectively with porous absorption. But using tuned traps exclusively is a mistake because it ignores that all rooms have problems at all low frequencies. Not just those related to the room dimensions.

Tuned absorbers are a great solution for problems below 80 Hz, which are the specific topic of discussion. No other problems are being "ignored." As Gerry pointed out, the true value of bass traps comes in above this 80 Hz range. Below it, the modal problems are, for typical HT-sized rooms, a relatively small handful. Narrowing it down to the modes that are actually problematic usually reveals one or two frequencies that need special attention. For which both narrowband "traps" and "EQ" are good solutions (in that order, of course ).

> The perception of increased bass caused by absorption is because the treatment lowers the level of midrange energy making the bass (we are less sensitive to) appear louder.

No!

The perception of more bass is usually due to raising half a dozen nulls that had been 20 or more dB down.

I think you're both right.

> a truly flat response would not be to most folks liking. <

I disagree with that strongly. Most people (including me) do not like a system that has been equalized in order to make the response measure flat with test gear. That's a far cry from not enjoying a flat response per se.

Study after study of subjective listening tests has shown that "flat" is not pleasing. The science on this is solid. We have the A-weighting curve, equal-loudness contours, recommended RT curves, STC curves, and so on and so forth, all based on the non-flat human response to sound. Pleasing is usually a little bit heavy in the bass, and a little bright on the high-end. Of course, that's not the case with everyone. If I may be so bold, if you think flat sounds enjoyable, then I would seriously doubt you're listening to flat!

 

[Savant]

> the impedence "tube" <

Good point. Does anyone have an impedance tube large enough to measure a 2 by 4 foot panel?

Yes - RPG has one. But the test method will actually work in a chamber. The wavelengths are quite large.

> testing an absorber with ETF in a bedroom will only reveal what that one device does in that situation. <

Well, sure, but if you measure three devices in the same session a test like this will clearly show how well each absorbs at the room's own resonant frequencies. So if a room has modes at 70 and 103 Hz, you won't be able to measure at 85 Hz. The room I've used for this type of testing has modes spaced well enough to be quite useful. And I don't have to pay IBM $1,000 every time I want to test something!

My point is that this sort of technique cannot be used to establish any sort of process by which different manufacturers can compare products universally. Just because product A outperformed product B in your bedroom or living room does not mean that that will always be the case. A good lab technique qualifies performance in a manner that makes the results fairly (as in - in a manner that is fair to all) comparable.

> What varies by 50%? <

See ASTM C423, page 7, describing the variations they measured in round robin tests. At 125 Hz absorption varied as much as 50 percent. You can be sure that at lower frequencies it varies even more.

If you are referring to Section 13.1.2 and the accompanying Table 3, I can see where you are coming from. However, I would point out that the 50% you are calculating is a statistical parameter. Strictly interpreted, this section (on Reproducibility of test results between labs, for those not privy) there is a 95% probability that the absolute difference between two test results will be below the R value from Table 3. Assuming a standard distribution of results, most values for the absolute difference would be well below the R value listed. So while this does state that (for A-mounting) a 50% difference is possible, it also implies that it's not probable.

Also, the MiniTrap tests at Riverbank you did while at Auralex (which you have said you stand behind) show three times more absorption at 80 Hz than what we measured at IBM. You measured 22.44 Sabins at 80 Hz, and IBM measured 7.44. Now, this was for corner placement, not a standard C423 test.

If you really want to go there...

This is completely irrelevant. As I have said many, many times before, I stand behind the data because they represent the results of a closed comparison of the performance of different absorbers. The test results should NEVER be compared with the results you obtained from IBM. This actually brings up one of the reasons I removed all the data I had control over from the web - it was being misinterpreted, just like it is being misinterpreted by you right now. Oh, the irony.

************

One more side thought on the 50% difference thing: If the same sample of absorber A is measured in different labs and significantly different results are obtained, it is pretty safe to assume the same sort of difference (up or down) would result when absorber B is measured in the same two labs. I.e., if absorber A measures 8.0 sabins in lab X, 8.8 sabins in lab Y, and absorber B measures 10.0 sabins in lab X (all in the same frequency band), then it is probably safe to assume that the result for absorber B in lab Y will be around 11.0 sabins. Of course, I cannot stress enough that these numbers should not be used as absolute values. The trends are important, not the numbers themselves.

 

[Vaughan Odendaa]

Glenn, for some reason my email isn't working. I can't access email and I'm at MSN. Please email to snuggles_1@alumnidirector.com. God, now I'm really embarrassed.

--Sincerely,

 

[Ethan Winer]

Jeff,

> a neutral acoustical environment is typically the goal so that recorded content can be evaluated and/or appreciated in its truest form. <

Agreed, but that's supposed to be the goal of audiophiles too, no? It's certainly my goal!

> When the audio (only) was played back to a member of the Auralex staff not as well-versed in "what good acoustics sound like," the listener's preference was for the "before" track. I.e., they preferred the sound of the movie played back in the untreated room. This was a telling example of the aural expectations people have when they wish to be entertained and aren't listening "critically." <

That's a great point, but I'd say the ultimate goal is for reproduction to please those who are more sophisticated than the typical neophyte. Before I learned about the importance of treating early reflections I was one of the "unwashed" too. Now, I can't imagine listening any other way, and I can spot the damaging effects of early reflections in someone else's system immediately. We all learn to appreciate better sound and more detail. I've been playing music since the fourth grade, and professionally since I was in my early 20s, but when I started playing the cello 14 years ago at the age of 42 my sense of pitch quickly became even more acute. So I can't agree that we should dumb down the goals of high fidelity to appeal to uneducated and inexperienced listeners.

> Too much absorption on the side walls could actually be detrimental in this case. <

Agreed. I have a lot of panels in my living room, but they are all at very specific places. I have only two on the side walls at those precise places between me and the front loudspeakers, plus two more for the surrounds.

> I believe you've been referred to papers elsewhere that demonstrate this. My own personal recommendation ... <

I don't have those papers, and it's a waste for me to get them because most have math that's way over my head. Math and highly technical papers are not needed to show what is a simple concept. Either EQ can be shown to reduce ringing or it cannot. In fact, I'd argue that math doesn't prove anything (in this context) because what matters is empirical results. I've seen math used to "prove" that non-modal frequencies cannot exist in a room. Well, we all know that's not true! So if one of you guys has a paper or some other evidence proving that EQ can reduce ringing in practice, and over a practical phsyical distance, please post an excerpt.

> both narrowband "traps" and "EQ" are good solutions (in that order, of course ) <

I agree. I am not dissing EQ. I've acknowledged many times that it can help at the lowest frequencies, and that I now use the (minimal) EQ built into my subwoofer. My only objection is to what I see as unrealistic claims for EQ, or innaccurate criticism of bass traps. Like claims that bass traps are ineffective below 80 Hz. Glenn and I both have lots of ETF data proving otherwise. Empirical evidence trumps theory every time.

> Study after study of subjective listening tests has shown that "flat" is not pleasing. <

I'm not talking about flat with regard to energy content. I'm talking about flat reproduction. If you wouldn't accept a preamp or CD player that's down 10 dB at 15 KHz, then you shouldn't accept that in your speakers or listening room either. I have Mackie 624s in my living room HT. They are spec'd flat by Mackie, and have been independantly measured (David French) as being very flat. If a record producer makes a mix using flat speakers in a flat room, I want the same in my room so I can hear it exactly as he or she intended. See your earlier comment I quoted at the very top of this post!

--Ethan

 

[Ethan Winer]

Jeff,

> Just because product A outperformed product B in your bedroom or living room does not mean that that will always be the case. <

Within the confines of the available test frequencies - the room's modes - why won't an absorber that measures better at, say, 60 Hz in my test room also be better in every other room at that same frequency?

> A good lab technique qualifies performance in a manner that makes the results fairly (as in - in a manner that is fair to all) comparable. <

Of course.

> the 50% you are calculating is a statistical parameter. <

If it can be shown that even one lab reported absorption 50 percent higher than another at 125 Hz, then I don't care about probability or how the variances are distributed.

> So while this does state that (for A-mounting) a 50% difference is possible, it also implies that it's not probable. <

As I read that, what's being reported are the actual results of a real "round robin" test. Not a theoretical deviation, but what was actually measured. Again, empirical evidence trumps theory - and statistics! - every time.

> If you really want to go there... <

Hell no, and I already said that.

> The test results should NEVER be compared with the results you obtained from IBM. This actually brings up one of the reasons I removed all the data I had control over from the web - it was being misinterpreted, just like it is being misinterpreted by you right now. Oh, the irony. <

The irony indeed. Back atcha buddy!

> I cannot stress enough that these numbers should not be used as absolute values. The trends are important, not the numbers themselves. <

Bingo.

Thanks Jeff. You are da man.

--Ethan

 

[Savant]

Within the confines of the available test frequencies - the room's modes - why won't an absorber that measures better at, say, 60 Hz in my test room also be better in every other room at that same frequency?

Because if the same product is placed in a different manner in the same room, a different result is likely. Because there's no control for the test. Because there cannot be a calculation of reproducibility of the test because no one else has the same room with the same setup. A conclusion about the behavior can certainly be made. But that conclusion will only hold up until a purer test is conducted. At that time, the initial conclusion will either be debunked, proven, or it will become clear(er) that more work is needed to formulate a more solid conclusion.

If it can be shown that even one lab reported absorption 50 percent higher than another at 125 Hz, then I don't care about probability or how the variances are distributed.

I believe the round robin results are available from ASTM. And if you don't care about probability, why did you bother with the calculation in the first place? The reproducibility value, R, from the aforementioned table is a probabilistic value.

> So while this does state that (for A-mounting) a 50% difference is possible, it also implies that it's not probable. <

As I read that, what's being reported are the actual results of a real "round robin" test. Not a theoretical deviation, but what was actually measured. Again, empirical evidence trumps theory - and statistics! - every time.

The tables show mean values of absorption coefficients and estimates of repeatability and reproducibility. And the text clearly states the latter as being (estimates of) statistical values.

Now, I don't know what the exact results of the round robin were (see above). It is likely that some of the 14 labs were way different from others. The fact that the stated values are probabilistic would make me think most of the 14 labs were not at the extreme 50% end of the scale at 125 Hz.

 

[Buckeye_Nut]

Wow....

This thread has surely taken a left turn.

 

[Ethan Winer]

Jeff,

> Because if the same product is placed in a different manner in the same room, a different result is likely. <

Well that's why I put all the things I'm testing in the same place!

> It is likely that some of the 14 labs were way different from others. <

Yes, exactly my point.

--Ethan

 

[Savant]

Taken in its full context:

Now, I don't know what the exact results of the round robin were (see above). It is likely that some of the 14 labs were way different from others. The fact that the stated values are probabilistic would make me think most of the 14 labs were not at the extreme 50% end of the scale at 125 Hz.

 

[Buckeye_Nut]

Hey everyone,

I finally have pictures up of the GIK panels in place if anyone wants to see how I have them positioned. My sig links to the photos as well as comments regarding the impact they've made.

 

[Glenn Kuras]

Hey everyone,

I finally have pictures up of the GIK panels in place if anyone wants to see how I have them positioned. My sig links to the photos as well as comments regarding the impact they've made.

wow the room looks totally different from before. Great job sir!! Could you email me the pictures also?

Glenn

 

[Ethan Winer]

Glenn,

> Could you email me the pictures also? <

This works on all web sites:

Put your mouse over a picture
Right-click
Select Save Picture As...
Tell your browser where to save it to

--Ethan