Questions about poly diffusers

[Matt C.]

Hi, I'm hoping to get some information about the construction of poly diffusers. I'm planning a new studio space for myself and thought these could be a useful room treatment tool (if needed). I'm mostly thinking of putting them up in a live room, not a control room.

I've already done some research and come across many different ways to build them, so I'm hoping the people here can help me sort out the pros and cons/acoustical properties of certain construction details. For example:

- Fill the area behind the diffuser with absorptive material, or leave it empty?
- If there is absorption behind the panel, should it touch the back of the panel to provide damping, or not?
- Should the cavity behind the curved panel be open or sealed?
- Should I build curved bulkheads/ribs to sit behind and support the panel (as suggested in Master Handbook of Acoustics)? Or should the panel only be fixed at the two edges?

Any info would be great, thanks!

[Soundman2020]

I'm planning a new studio space for myself and thought these could be a useful room treatment tool (if needed). I'm mostly thinking of putting them up in a live room, not a control room.

That's the more common use for them, yes.

I've already done some research and come across many different ways to build them,

Did you find how to get the correct shape? Many people think that they are based on a circle, but that's not correct. The name says it all: "POLY-cylindrical". Thus not cylindrical. In other words, the radius is not constant; it changes. The strictly correct shape is technically called a "catenary curve", which is the shape you get when a length of chain hangs freely, but pretty much any curve where the radius changes constantly across the surface is fine. One way to get this is to take a piece of paper, put it flat on the table, out your hands flat on either edge, and gently push them together, until the middle part of the paper rises up off the table. That's a good curve too. Not catenary, but still useful, acoustically. You can get the same effect on a larger scale with a think sheet of plywood laid flat on the floor, and two people pushing from opposite edges. It's a mistake to make a "poly" using a constant radius cylindrical shape.

- Fill the area behind the diffuser with absorptive material, or leave it empty?

It doesn't hurt to fill it! It can help with bass absorption, and will also damp any resonances that might happen to occur inside.

- If there is absorption behind the panel, should it touch the back of the panel to provide damping, or not?

Yes, it would be better to do that. The poly is not meant to be a resonant device, so anything you can do to damp resonance would be good.

- Should the cavity behind the curved panel be open or sealed?

It doesn't really matter. It would only matter if you wanted to make it into one of my favorite toys: a slotted poly! I have devised several ways to use those in various studios I have designed. It combines a slot resonator with a poly diffuser, all in one unit. In this case, the backing panel must be very rigid, and the cavity MUST be sealed, except for the slots. And the insulation MUST be thin and placed immediately behind the front curved face, fully touching it.

Here's one thread from a studio I designed that is under construction at present, and the owner is in the process of building his slotted polys right now: http://www.johnlsayers.com/phpBB2/viewt ... =2&t=21368 Those are designed for deep bass absorption, as well as dealing with SBIR and also keeping the room live.

And here's a couple of photos of a slotted poly I designed for another studio, under construction and completed:

BUILD-30th-construction--slotted-poly-diffuser-under-construction--2014-07-23-photo1-SML-ENH.JPG

BUILD-30th-construction--slotted-poly-diffuser-completed--2014-07-23-photo2-SML-ENH.JPG

In this case, the front is covered with fabric in addition to having thin insulation behind the panel. So you can't see the slots, but they are there. Here's an image taken from the actual design stage, showing the slots:

Rod-MO-USA--Complete_Studio-V5--S376--SLOTTED-POLY-DETAIL-01-ENH-SML.png

The spacing and width of the slots were carefully calculated to provide the tuning I needed for that studio. You can see the final results here: http://www.johnlsayers.com/phpBB2/viewt ... =2&t=20471 including acoustic response testing of the room.

- Should I build curved bulkheads/ribs to sit behind and support the panel

See photos above... In other words: it depends on what you are doing! In the top case (4x4 slotted poly), the curve is created by the careful placement of the 4x4's on a poly-cylindrical grid, with careful angling, and since the 4x4s are extremely robust, there's no need for any backing ribs: just the top and bottom panels hold it all together. But for the other case above, which is Studio Three, there are ribs to keep the shape of the front panel, which is plywood in this case.

Basically, if the front panel is thin and would not keep the shape properly by itself, then use ribs. If the front panel is rigid enough to keep the shape, then you don't need them. Also, if you do decide to do my "slotted poly" concept, then you might consider that the ribs can be used to divide the cavity into sections, as in the case above for Studio Three: The slots at the top are different widths compared to their siblings at the bottom, so I could get better tuning to a different range for each segment of the cavity.

So I guess the best answer I can give you to all of our questions is: It depends on what you want the device to do! There are many options, and each of them requires that you build it in a different way.


- Stuart -

[Matt C.]

Thanks so much for the information and ideas, that's very helpful. The space itself has yet to be built and measured so I don't know exactly what it'll need yet. But my general idea was to build units that would absorb a broad range of bass/low-mid frequencies while diffusing frequencies above that to avoid deadening the room too much. So my thought was to essentially build a "typical" simple broadband absorber panel with 4" of rigid fiberglass, then bend a sheet of 1/8" plywood across the face of that. Mainly I was concerned about whether it's bass trapping capabilities would be compromised by obscuring its front face (although the rear/top/bottom would still be open, not completely sealed up). Also in my hunt for information, some people talked about a poly's bass trapping being based on the curve's membrane-type action, in which case I would expect sealing/not sealing the cavity would make a bigger difference. But it sounds like from what you've said that my simple and cheap idea should do what I intend it to do?