divider

Studio Ceiling Diffusers/Reflectors

Documenting the ceiling reflector/diffuser panels I build for my music studio.

The Problem

After the renovation of my music studio, the existing ceiling was removed, a sound-absorbent layer was laid against the underside of the floor above, and the support beams were left exposed. In conjunction with the carpet flooring, this left the studio very dead sounding. I wanted to liven up the room, without having the parallel walls render it too fluttery. Ceiling diffuser panels would "cover" the absorbent lining, and "break up" the echoes from the walls (rather than simply absorbing them).

Philosophy

The mathematics behind QRD and PRD diffusers, while stunningly elegant, suffer from one massive "elephant in the room"-type flaw: the beautifully even diffusion profile only happens at a single target frequency. (Fractal panels are one attempt at extending this - but they can only manage three or at most four target frequencies.) At lower frequencies, the panels appear as a smoothly curved reflector, at higher frequencies a random series of flat reflectors. So for the majority of the frequency spectrum, you get random, uneven diffusion. This is not a bad thing - that's what all good sounding rooms have ever done. My aim then was therefore to focus on random reflections rather than chasing a perfect diffraction wavefront. Bear in mind that the "main" problem was a dead sounding room.

As an aside, this 1990 paper by Robert Walker describes primitive root diffusers in great detail, and includes a design that appears similar to RPG Inc's 1993-filed patent for their "Skyline" diffuser. I am sure a lawyer somewhere can point out the differences.

Design

ceiling panel design 1

ceiling panel design 2

ceiling panel design 3

ceiling panel design 4

ceiling panel key 1

ceiling panel key 2

Construction

ceiling beams with rails 1

ceiling beams with rails 2

The first step was to attach "rails" (40x19mm pine) to the exposed beams, so that the panels would have something to hang from.

The panels could then be inserted into the cavities at an angle, clearing the rails, then straightened up and released.

beading attached

bracing attached

After cutting (and testing!) the MDF panels, I attached short beading strips to stop the panels from slipping sideways.

The bracing was probably unnecessary, it was there because my original design did not have the added strength of the "box frames".

box frames attached

reflectors attached

The box frames were added to improve the aesthetics (after a couple of prototypes).

All the reflectors have been cut and attached. For aesthetic purposes I "bevelled" the bits that extended beyond the box frames.

closeup #1

closeup #2

Closeup of a lengthwise reflector panel.

Closeup of a crosswise reflector panel.

after painting

mild adjustment

All painted and ready for installation.

During installation some panels required adjustments. In most cases, the adjustment required was pretty mild.

mounted belt sander

wild adjustment

The mild adjustment was done with a DIY mounted belt sander (mounted with dowels).

Owing to a momentary lapse of concentration, two panels required more significant "adjustment".

After Installation

ceiling panels installed 1

ceiling panels installed 2

The installed panels look as good as I'd hoped. Notice the balance of randomness and regularity.

I actually tossed a coin to determine reflector size, placement and spacing.

Sound Samples

In this sample, you can hear some drumstick clicks and handclaps, recorded from across the room. I was standing beside a drum kit, and you can hear some sympathetic ringing of the heads, but the acoustics are very dead. (The difference between the two recordings is subtle, and heard most effectively with good headphones).

This sample is the same thing, recorded in exactly the same way, after the ceiling treatments were installed. The room is slightly more lively, especially on the drumstick clicks. More noticeable is the increased drumhead resonance on the handclaps (which masks most of the actual reverberation). I think that is because more sound is being reflected, and thus reaching the drums.

Final Verdict

I was initially a bit disappointed that the increase in reverb time was so modest. But then I spent a bit of time playing some instruments. The difference is significant, although hard to describe - it's more a feeling that you get when playing. Here are a couple of specific example: for the first time the piano sounded better with the lid all the way up. And drums "fill the room" rather than merely assaulting you with volume.

Issues Encountered

If the panels are constructed too "tightly" (every piece of wood glued to every other one), they tend to warp, making fitting very difficult. If constructed too "loosely", there is a possibility that they will rattle or buzz when a loud bass note is played.
I ended up using a lot of glue, but I waited between various construction steps, to give the wood a chance to settle naturally. The use of Blu-Tack as part of the installation process (quite possibly a first in the world of acoustics) should also minimise any rattling.

Home

All content Copyright 2015 Trevor Magnusson