[JoshK]

Some of the lspCAD and modern audio software can model impulse respose as well as ETC curves.  But ultimately I think this is one of those things you should be easily able to hear or not hear.  

I wanted to state for the record that I learned that BH5 and WM2 differences are more than just a moisture layer, the BH5 has an extra dampening layer too.  I was wrong about that.  Does this extra layer add significantly to the reduction in resonance? Is it worth 3x the price for the little extra?  I'll let you decide.  I am not fretting over it myself, and think the WM2 was well worth the price of admission.

[woodsyi]

I guess I will have to get some BH 5 when I get the CD lens.  I have WM2 but have not had time to do the surgery.  In light of all this discussion, I think I will put the WM2 on the largers and BH5 on the RM40s.

[ctviggen]

This software gives you the ability to select a portion of the impulse response:

http://purebits.com/appnotes.html

This means that you'd get "anechoic" results.  ETF does not allow you to do this (I do not believe), so you're stuck with in-room response, which includes reflections (both first and subsequent).  

But what does "hearing" a benefit mean?  If you can't test it, how do you know you're hearing something?  (The infamous cable argument.)  I'm more of a scientist and would like to know exactly what's happening.  Even if I can truly hear a difference, I have to know why.

[csero]

This software gives you the ability to select a portion of the impulse response:

http://purebits.com/appnotes.html

This means that you'd get "anechoic" results.  ETF does not allow you to do this (I do not believe), so you're stuck with in-room response, which includes reflections (both first and subsequent).

Let's not forget that with windowing you are reduce the resolution/accuracy and usability of the measurement in the LF.
This is also a major problem with any room eq - they either use a windowed mls, and the base data for LF correction useless, because is not accurate and does not contain the room response, or use a very long mls which ignore the directional component of the response, or impulse response obtained with slow sine sweep, which ignores the directionality and time variance of the room response.

[ctviggen]

Frank, by "window" do you mean cut the impulse response short or do you mean the normal windowing applied to a discrete-time signal in order to take an FFT?  I think you mean the former.  I can see that would be true.  If you truncate the impulse response prior to the first reflection, you wouldn't get the room response and only get the response of the speaker.  It's also true that with a long MLS you'd get just the steady state response, which would basically ignore all directionality (do they just add up on the frequency response components and put them into bins?).  This is unlike a time domain analysis of the signal, which could give first and subsequent reflection data.   And either long MLS or sine sweep would ignore the time variance of the room response.  

Interesting points.  I'm not sure then how to determine a test that would tell what BH5 does.  Wouldn't a truncated impulse response show some LF effects?  Maybe not.  I mean, we're talking milliseconds before the first reflection hits and the bass driver is barely going to be able to move in that time.  The amount of LF info would be minimal and then the first reflection would hit, followed by all the other reflections.  The only thing I can think of is to take the speaker outside and just deal with whatever reflections the ground causes.    However, all these people manufacturing speakers make frequency measurements.  How do they do it?

[csero]

Frank, by "window" do you mean cut the impulse response short or do you mean the normal windowing applied to a discrete-time signal in order to take an FFT?  I think you mean the former.  I can see that would be true.  If you truncate the impulse response prior to the first reflection, you wouldn't get the room response and only get the response of the speaker.  It's also true that with a long MLS you'd get just the steady state response, which would basically ignore all directionality (do they just add up  ...

Yes I meant the first.
To measure the BH5 windowing is good enough, because  the reflection it tries to damp is within 2-3 ms ( 1-1,5' deep cabinet) and it won't be effective in the LF anyway. In the HF also the back radiation of dynamic drivers are significantly decreasing above 1k. At least it can show a rough estimate wether it does something. Of course it does not hurt if you can have longer window (closer to real anechoic measurement).

However, all these people manufacturing speakers make frequency measurements. How do they do it?...

Let's not comment on this ...

[BobMajor]

Thanks to everyone for the information.
Brian stated that the 626R shouldn't have too much BH5.
How much is enough?
Should it go directly behind the woofer only?
Anywhere else?
Thank you.
Bob Major

[mjosef]

Here is my experience installing BH5 in the RM1 cabinet. I placed the cabinet on its top, removed the base and p/r, removed the fiberglass(there was a lot of fiberglass in there). I had to remove the upper woofer to be able to reach all the way into the top of the cabinet. I cut the BH5 sheet in half resulting in 2 piece 9'X12" which could fit through the p/r hole. I was able to fit the two halves(1 sheet) on one side in the area between the top of the cab and the window brace between the two woofers which measured around 12X24". I realized that the BH5 needed a condiserable amount of pressure to really stick to the wall surface; I gently pulled at the edges of the BH5 and saw it separating from the surface, so I spent a considerable amount of time applying more pressure using open palms, fingers, knuckles along every inch of the applied BH5. My cab has the soundcoat option. Last thing I wanted was for there to be a loose sheet of BH5 flopping around in there. This made sure that all the pieces was securely adhered to the walls, but it slowed the install considerably. Oh, and I also vacuumed the cab inside to get rid of any fiberglass dust.
With one side done, I applied a piece to the top, then did the other side identical to the previous side (1 sheet in halves), then the back area above the crossovers(a 7X16 piece), and finally two small pieces for the side areas below the window brace (8"X10"). Total BH5 used was 3 sheets. Brian said 2 sheets should do it. Mhnnn...
The inside sure looked good with the BH5 linning the walls, kinda like a padded cell. And the knuckle test had a reassuring knock. I had two 13gal bags of fiberglass to fit back in there and thought to myself no way could I get all that fiberglass back in. Got most of it back in, had a small piece left, and Brain said it was ok.
With woofer, p/r and base reinstalled, it was time to flip the cab back upright...man, it seemed heavier. THe other cab was there to be done but I was tired and eager to hear the result. With no regard to placement I wired it up and fired up some music...uh oh, sounded like the speaker was stifling...did I put too much BH5 in or too much fiberglass? Then I thought maybe it just the positioning. I should mention that I had also vitrified the p/r since I had it it exposed. I pinched a little putty off, and it sounded much clearer. Good. My last task was to replace the fiberglass in the mid panel enclosure, and since I did not have lambswool I used PE acoustastuf. It sounded good with the couple tunes I listenned to before calling it a night, I will leave the  fiberglass in the other midpanel enclosure so I can compare it directly.
I am looking at 2-3hours per speaker, certainly labor intensive. I may just leave the other speaker as it and do a direct comparative listen.  Stay tuned.
PS. BH5 is easy to cut using a sharp utility knife cutting along the adhesive side applying enought pressure to compress the foam enabling a clean cut right through.