The Sound’s the Thing
Any of you who know me at all are aware that I’m very concerned about room acoustics when I’m recording. A while back, I wrote this article on the subject for a DIY recording blog my friend Dara Korra’ti publishes. Over the past couple of years I’ve been tackling the acoustics in my own studio.
I started with my isolation booth, reasoning that getting the best sound recorded was my first priority, and I could work around the problems in my control room later, since those issues did not affect the sound quality I captured, only my ability to hear it accurately later when mixing. With the booth fixed, at least I would have good sounding tracks to work with.
The basic problem in my booth was comb filtering, caused by sound waves bouncing around a very small space, and causing sharp peaks and nulls as they ran back into themselves. This results in a boxy, hollow, somewhat distorted sound. I solved that problem (very effectively) with a set of tube traps from Acoustic Sciences Corporation.
Properly arranged in the recording space, they create enough random reflections to prevent comb filtering. This was a huge win for me, and despite the rather steep price, I’ve never regretted the investment.
Solving My Control Room Issues
Recently, I decided to try to work on my control room. This presented an entirely different set of challenges. The primary problem here was not comb filtering, but uneven frequency response, and especially some overly long decay times at certain frequencies. Subjectively the room didn’t sound terribly bad, but it was not accurate. This made it hard for me to properly hear as I mixed music. Trying to mix in an inaccurate room is like trying to paint a picture under colored light. If the light is too red for example, you might use too little red to get the color you really want when blending paint. Likewise, if there’s too much of a certain frequency in the room’s response, you will tend to under-emphasize that frequency when mixing. This means that mixes that sound perfect in the room they were created in may sound off (or really bad) when played back elsewhere. This makes for a lot of back and forth – mix, listen in other places, correct, lather, rinse, repeat, until you get it where you want it.
My control room is pretty good sized, and one thing that had stopped me from really dealing with it was the expense to buy enough ready-made treatments. To effectively deal with my room, it could easily cost 5 figures (and not a number stating with 1). This meant building something myself. Now, understand, I am decidedly not all that handy, so this was a pretty daunting idea.
I decided the first step would be to take some audio measurements of my room, following this method to objectively evaluate what the problems were. I emailed my friend Gerry Tyra, who is an engineer (he can’t discuss what he does in his day job) who has a side interest in audio, to give me some advice on how to interpret the measurements. Gerry being Gerry, he replied quickly that he “had some ideas” on how to build some absorbers and diffusors to correct the room. We spent a couple of months refining his ideas, taking more measurements, and choosing and ordering materials.
The basic design we settled on was a box, just under 8″ deep, with a 4″ layer of Corning 703 rigid fiberglass for absorption, with an air gap of almost 4″ behind for bass and low mid frequency control. We’d build these in two sizes – large 6′ x 2′ units to hang on the walls, and small 2′ x 2′ units to mount on the ceiling. Additionally, some of the large units would have a masonite panel with a specific pattern of holes drilled in them, in front of the fiberglass to provide targeted diffusion of certain frequencies. In front of everything would be a frame with acoustically transparent fabric.
I’ll give a very basic step by step of what we did, in case you’d like to try to build something like this. I’ll be happy to answer questions for anyone who wants more detail.
So, first, a look at the basic materials we used. As I mentioned, the main absorption is provided by Corning 703 rigid fiberglass. This comes in 2′ x 4′ sheets. You can get it in thicknesses from 1″ to 4″. I decided to get 2″, and use 2 layers per unit. This meant I needed 9 boxes (at 6 sheets per box).
We decided to use MDF for the outer boxes. It worked well for three reasons – it’s fairly inexpensive, it cuts very cleanly, and one of the standard board widths (7-5/16″) was ideal for our purposes, so we wouldn’t need to do any additional ripping to get the width we needed. I toyed briefly with the idea of using a hardwood, like red oak for this, but the cost sent me scurrying back to the MDF. Another bonus was that the boards came with one side primed, so it would take paint nicely.
The other lumber in the above image is 3/4″ x 3/4″ poplar. This is what we would use to build the frames to mount the front fabric to, and also to frame in the fiberglass in the back of the units.
Speaking of fabric, this is what I used. It’s a specialty fabric manufactured by Guilford of Maine, that’s acoustically transparent. I decided to use two fairly bold colors.
The final major material was 1″ x 3″ pine board. This would be used to make french cleats to mount the units to the walls and ceiling in my studio.
Building the Units
Build – Day 1
We started by making the mounting cleats. The first step was to rip each board at a 45° angle.
Next, we blunted the sharp edge of each board slightly.
We noticed as we ripped and blunted the boards, that they were not coming out identically. This meant that for an accurate installation later, it would be vital that we use matching pairs of pieces. To insure this, we rubber-banded the halves of each rip together, and put the boards aside for later.
Stage 2 was to build the poplar frames for mounting the fabris and retaining the fiberglass sheets. We miter-cut the first end at a 45° angle.
Now we set about cutting the poplar boards to the right lengths for the frames. We would need two lengths. A short length for all the 2′ x 2′ frames and the 2′ side of the bigger frames, and a long length for the 6′ x 2′ frames. Gerry devised a jig we could clamp to the needed postions to give us the lengths.
The jig was mitered at 45°, and mounted to accept the ends we had already processed. Once it was set, and boards shoved into position, we could make the second miter-cut.
We ended up with a nice pile of poplar.
It was time to build the fabric frames.
We started with some wood glue on the ends.
We clamped each joint.
We pinned the joints with a nail gun.
We allowed each frame to set in the clamps for about 15 minutes.
Build – Day 2
Gerry and I split forces on the second day. I spent most of the day in the living room mounting fabric to the poplar frames. While I did that, Gerry wanted to experiment with the precise construction technique for the outer boxes. Before disappearing into the workshop for his experiments, he posed for some photos demonstrating the method for mounting fabric.
Step one was to cut the fabric to size. It came in 66″ widths, so I was able to get 2 pieces per linear length. I left about 2″ around each edge to fold over and staple.
Next, I clamped one side, and stapled the opposite side, pulling tight as I went.
Now I stapled the side clamped in the previous step, removing the clamps as I worked along the length.
Once I had two sides stapled, I could then staple the remaining sides, taking care to pull the fabric taut as we worked.
The final step was to glue the fabric to the corners of the frames, using wood glue:
And voilá, finished panels:
Gerry’s experiments with the MDF for the outer boxes resulted in our first change in plans. Originally, he had wanted to use miter joints, reinforced with wood dowels for strength and rigidity. However, he was concerned that it would be difficult to accurately drill the mitered ends of the MDF to consistently line up the boards. As it turned out, his concern was justified. Lacking a drill press that could simultaneously drill holes for all the dowels per board at once, it was going to be impossibly tedious to drill the holes for the dowels accurately. We decided that between the glue and nail pins on the joints; the poplar frames that would be attached at the front, and behind the fiberglass; and finally the pair of cleats that would span the width at the back, that there would be more than enough strength and rigidity.
Cutting the MDF for the boxes was similar to cutting the poplar. The first job was to make a 45° miter cut on one end. You can see how smoothly the MDF cuts.
We nailed the jig (remember that from poplar frame build) to the right position to cut boards for the 2′ side, and shoved the first mitered end into the jig to cut the other end.
Once all the pieces were cut, we were ready to assemble the boxes. We laid some clamps on a table, the hold the side pieces, and glued the ends on. It took both of us the get things lined up, before clamping them down. Once clamped, we used the nail gun the pin them:
We used wood putty to fill the nail holes, and any other gaps or dents.
Eventually, we had a lot of boxes.
This job stretched from day 2 into day 3.
Build Day 3
Back to the Mounting Cleats
Once we were done building the boxes, we chopped the long pine boards we had ripped way back at the beginning into the right widths to mount in the boxes.
As I mentioned earlier, we had not been able to get a consistent width on these boards when ripping them. That wouldn’t matter at all, as long as we made sure to always match the two pieces that came from the same section of each board. One of each pair would be mounted inside a box, and its mate would be attached to the wall or ceiling later. To make sure they matched, I numbered each pair as Gerry chopped them.
After all the cleats were chopped to size, we drilled each half that would be mounted in a box for pocket screws. Gerry had a special kit with a drill bit and a device to angle the drill for this. You can see that he’s working on box cleat from pair #20. To the right are some cleats that have already been drilled.
Next, we mounted one of the pocket-drilled cleats in each of the boxes. Gerry devised a clever jig for each size, so that we could always get them into the right positions. Unfortunately, I forgot to get photos of the jigs in the boxes to show this, but here’s the small jig by itself. It was sized to firmly fit in the box, with cross-bars that we shoved the cleats up against:
Toward the end of day 3, we were finally ready to paint the boxes. This turned out to take much longer than we had planned. We worked late into the night of day 3, well into the evening of day 4, but I’m getting ahead of myself.
Gerry has a paint sprayer, which is efficient, but obviously requires that the area be protected from flying paint.
The paint crew also needed protection, at cost of personal dignity.
It took Gerry a little while to get the hang of the paint sprayer. At first we were spraying too much paint, and the floor was thick with it. The boxes looked fine, but we were going through paint pretty quickly.
After a few panels, Gerry got the technique refined. However, we found that it was taking much longer for the paint to dry to point where we could handle the boxes than we’d expected. We had hooks on the wall in a different area of the garage, and we hung the boxes by the cleats to dry. We had assumed we’d be able to move each box after 10 minutes or so, but it turned out to be more like 30-40 minutes. Again, I forgot to take pictures of this (we were both pretty fried by now). We got 6 or 7 (of 25) done that night, and decided to turn in.
Build Day 4
More Painting and Assembly
The first thing I did on day 4, was to run out to get more paint. We then got to painting more units. While we waited for the newly painted boxes to dry, we assembled the ones we’d painted the night before.
First, we pinned the fabric panels flush with the front opening of the box with Gerry’s nail gun:
Now it was time to get the star of the show, the rigid fiberglass:
For the large boxes, we started by laying a full 4′ panel inside.
Now we cut a panel in half.
We laid one of the cut pieces in the box.
We started the second layer by placing the other cut half in the box. We offset it from the other small piece.
We finished the second layer off with a full sized piece.
Finally, we laid pieces of poplar on the fiberglass and pinned it to the box, to hold the fiberglass in place.
For the small units, it was pretty much the same, except we just cut one sheet of fiberglass, and layered the two halves in a 2′ x 2′ box.
And the final result:
At this point we had all the pure absorption units built. Gerry still had a bunch of drilling to do on the masonite panels to be placed in the six large units that would also function as diffusors. These panels would be laid on between the fabric panel and the first layer of fiberglass in their respective units, so they couldn’t be assembled until he was done. Maya and I headed home the next morning. Gerry and his wife Sandy would bring the finished units up from their house in southern California to our place in the Bay Area the following weekend for an installation party. Gerry planned to finish the diffusion panel drilling and assemble the units during the week.
Gerry and Sandy arrived the following Friday evening. Saturday morning, we got to work. The drilling for the diffusors ended up taking much longer than Gerry had anticipated, so he ended up doing a fair amount of it at my place on Saturday. Unfortunately, I completely forgot to take pictures of any of this part. Once he got that done, amd we assembled the remaining units, we began to mount the large units on the walls.
For maximum effectiveness, we wanted the units to form an airtight seal against the wall. We used foam weather stripping for this.
Next, we attached a pair of cleats to the wall. We made sure to use the matching numbered cleats for the pair attached to the unit. We used the same jig Gerry built to line up the cleats in the units to position them on the wall. I don’t have a picture of the jig on the wall, but here’s a picture of the jig itself.
The crossbars at the top and bottom are spaced to position the jig in the box, they sit up against the inside edge of the top and bottom. The two pairs of crossbars set the placement of the cleats. For the boxes, each of the two cleats is placed against the top bar, with its point facing downward. For the walls, the overall height is set by placing the vertical pole on the floor, and mounting a cleat resting on the bottom of each crossbar, with its point facing upward. Here’s a diagram of the orientation of the cleats on the wall and in the box.
Once the cleats were screwed into the wall, it was just a matter of lifting the unit and sliding it down so its cleats locked into the ones on the wall. The weather stripping worked beautifully to create a seal.
Here are some of the units mounted on the wall.
We finished mounting the wall units on Sunday afternoon. After dinner, we began mounting the small units on the ceiling. This was tricky, in several ways. We decided to mount them with a diagonal orientation, which made for some tricky measurements. Once we had the placement worked out, there was the question of working over our heads on the ceiling. The basic plan was to temporarily screw the small jig to the ceiling, mount the cleats, and then mount the unit. As an additional safety measure, we drove a screw through the box into the butt end of each ceiling cleat. It was somewhat slow going as we figured out the best way to go about this, and we only got three of the twelve small units mounted. Gerry and Sandy went home Monday morning. About a month later they came back up, and Gerry and I fairly quickly finished installing them.
We took some audio measurements and saw a huge improvement. The overall frequency response was much more even, and the decay times for the various frequencies were much shorter, and more consistent with one another. Most importantly, when I cue up music, it sounds fantastic.