diffusion combined with absorption, end effect?

I have a false back wall in my room, filled with 14 inchs of 10 kg/m3 fluffy, perfroms really well! Really gets stuck into 60-100 hz and up. Anyway, next step is to mount some diffusers I have made on the back wall in front of the absorption. Here is some info on the particular type, note it is not the fractal version of the device: http://arqen.com/acoustics/acoustic-diffusers/ My concern is , what effect would an array of five of these units (5 modules 1200mm long per module, five modules is 2 meters wide) have on the absorbtion factor Im getting at the moment from the wall of insulation? The diffusers only diffuse from about 800hz up, to about 4khz approx. Would the low end simply wrap around the units, or defract around them if you will, and mostly continue on to be absorbed by the insulation? I of course am going to do before and after tests using REW, but until I can get a chance to get them up I would like to get some theories on what happens in this type of scenario. It would be a bitch to gain some nice mid band diffusion but get a bunch of low end thrown back at me that otherwise would have been nicely absorbed! Also dont want want to see any major changes in the low end waterfall from as it stands right now. To finish, I will be about 11 ft from the diffusers in my listening position. The width of the back wall is 3.2 meters. Thanks!

I guess the first question is: Why do you need diffusion at all? What design concept are you following for your room, and where does the diffusion fit into that? What is the specific problem with the room right now that makes you think it needs diffusion? Also, note that grouping several identical numeric based diffusers into an array increases the overall diffusion and effectiveness, but also concentrates and enhances the lobes. Have you checked how the lobing will affect your listening position?

note it is not the fractal version of the device:

Then lobing will be increased even more. Take a look at the graph on that web site for the non-fractal unit, especially at 750 Hz, and imagine that multiplied by a few dB. There is heavy lobing going on there. And those measurements were taken at 5m, which is greater than your listening position.

I of course am going to do before and after tests using REW,

If you haven't done the "before" testing yet, then how do you know you need diffusion? - Stuart -

Stuart, Sorry, that polar plot is for the panels unmodulated. That is mounted to the wall at the same depth. Mounting them at different depths, one flat on the wall, one a further 5 cm out, the next 6cm, the next 5cm, and the last flat on the wall again largely takes care of the lobing. I have already tested the wall untreated and also now fully filled. Next is to test with diffusion when I can get to it. The reason for diffusion is an attempt to keep some mid/high mid energy in the room as I am going to end up with a lot of broadband absorption in this room, as we all tend to in our small rooms (19.2 ft x 10.6ft x 7.8ft high).I may even redirect reflections from the first reflection points on the side walls to the rear wall, I haven't decided for sure on that one yet. As I mentioned my main worry is that an octave below the cutoff, approx. 400hz, the array would be just one big reflector of low frequencies back to the listening position. Largely negating the thick absorbtion that lies behind the array. This is what I'm looking for some clarification on.

Here is the result of the room empty and completely untreated. As you can see there is quite a drop off starting at around 5khz, for whatever reason Im not sure why. Im confident in the accuracy Im using a calibrated ECM8000 from Cross Spectrum Labs. Well, it will certainly be quite the drop off once a hell of alot of broadband is added. So the diffuser along with reflective traps in non first reflection points are going to be an attempt to keep as much higher frequency in the room as possible.

The reason for diffusion is an attempt to keep some mid/high mid energy in the room as I am going to end up with a lot of broadband absorption in this room,

The normal method for dealing with excessive absorption from broadband bass traps is by simply facing them with 6mil polyethylene plastic, which reflects highs progressively better the higher up the scale you go, and is effectively transparent for mid-lows and lows. 6 mil starts reflecting decently above about 800 Hz, and reflects somewhat above 650 Hz. (At least, for normally incident sound: reflection increases with reduced incidence). Thicker plastic reflects to lower frequencies, thinner starting higher.

I may even redirect reflections from the first reflection points on the side walls to the rear wall, I haven't decided for sure on that one yet.

Sounds like you are considering an RFZ design? If so, the normal recommendation for that is to have a deeply absorbent rear end, usually with hangers.

As I mentioned my main worry is that an octave below the cutoff, approx. 400hz, the array would be just one big reflector of low frequencies back to the listening position.

Exactly. To a certain extent. Based on the rigidity and dimensions of the diffuser panels. You will also be creating the potential for causing edge diffraction artifacts around the edges.

As you can see there is quite a drop off starting at around 5khz,

What other treatment do you have in that room? That high roll-off looks typical of carpet or something like that. Selective, random high-frequency absorption. Also, what speakers are you using? Have you tested the same speakers, mic, and interface out in the open, away from all surfaces that could affect the results? It would be good to confirm that you don't have an issue with the test equipment itself before you start. But I'm guessing that the roll-off is due to something else in the room that is selectively absorptive at high frequencies, like carpet, drapes, non-acoustic foam, etc. Photos of the room would help. - Stuart -

Thanks Stuart No carpet! So im perplexed. Hmm, don't know what to do about the diffusers now. On one hand I like the look and the effect they have on the Mids, but I hate the thought of a large reflector behind me at low frequencies. Anyway for what it's worth , some photos...

Perplexing indeed! There doesn't seem to be anything in there that could be sucking up the highs like that. I'd suggest first taking the speakers outside, far away from anything, laying them flat on their back on the ground, facing the the sky, and set up your ECM8000 at least 1 meter above the front of the speakers, then measure the response like that. It will give you a good baseline for seeing the actual, real response of the speakers, interface and mic. - Stuart -

What about a theory like this: With the room being untreated , there would be a hell of alot of high end bouncing around causing cancellations and hence the dip in the high end? I'm dreaming that up it has to be said, but it sounds plausible, maybe. Ultimately I'm probably going to have to run a test as you have described.

It could be some form of comb filtering, I guess, but it doesn't look much like it. Can you post the MDAT file somewhere that I could download it, to take a closer look? - Stuart -

Thanks Stuart,Id really appreciate you doing that. You can download it here: http://www.fileswap.com/dl/Gj2HGB2FJy/m ... .mdat.html

You can download it here:

I still can't figure it out for sure. It's as though you had thick pile carpet all over the floor, or something like that, sucking up the highs... But I suspect that it might be partially due to the absorption at the back. I would face that with 6 mil plastic, and see what effect that has. What you CAN see clearly on there, is that the rear wall treatment made a nice difference to the modal issues at 28 Hz, 57 Hz, 63 Hz, 88 Hz, 131 Hz, etc., and smoothed the LF response in general. So you are on the right track there, just not enough yet. I would put superchunks in some of the corners, especially the rear ones and perhaps the horizontal side ones. That should help. But there is a stubborn one at 146 Hz: I suspect that is related to your vertical axis (ceiling / floor mode), so that should get better once you get your ceiling cloud in, and other ceiling treatment. Apart from that, it looks reasonable, overall. The overall RT-60 is still way too long, of course, but that should come down too, as you start getting the rest of the treatment in place. - Stuart -