Carpeting Rehearsal Room Walls

Upgrading a studio in Clacton

Ed had made the mistake of carpeting all the walls and putting in very little by way of reflective surfaces or mid‑range absorption. While this reduced high‑frequency reflections, it had almost no effect on the mid‑range, resulting in a 'honky' room sound.

SOS reader Ed V had set up a couple of rehearsal rooms, along with a Pro Tools studio for his own work, in an industrial unit in Clacton-on-Sea. Ed and his friend Gary J did all the work themselves, which is impressive — but Ed immediately noticed that something was badly wrong with the sound of his control room...

He'd treated all the walls by putting a layer of carpet felt over plasterboard and MDF panels, then covering the whole of the walls, floor and ceiling in thin cord carpet. He'd added some glass‑fronted pictures in frames, to provide a bit of high‑frequency reflection, and the studio furniture did the same to some extent, but at the listening position he found that the stereo imaging was very poor, and there was a noticeable mid‑range honk that made everything sound very boxy. So, in search of help, Ed contacted SOS...

First Impressions

When we arrived, our initial listening tests confirmed what Ed had already described to us. The control room itself was 5.3m long, 2.4m wide and 2.3m high — so it only just managed to avoid falling foul of the 'two cubes in a row' geometry, which in the world of studio acoustics, is only one step better (or less bad!) than a perfect cube. Ed had set up the monitoring to 'fire' down the length of the room, which was the right thing to do, because in narrow or small rooms, firing the monitors across the width of the room invariably leads to a very unstable low end.

Using carpet as a sound‑deadening material seems intuitively quite logical, until you look at what is actually happening! The effectiveness of a porous sound absorber depends on its thickness relative to the wavelengths of sound being absorbed, and in the case of carpet — which is very thin — only the highest audio frequencies are affected to any significant degree. Adding the layer of underfelt made the absorber thicker, but it is unlikely that the combination would have had much effect on any sound below 3‑4kHz. At the same time, the plasterboard structure that Ed had employed allowed quite a lot of low‑frequency sound to pass through, while some was absorbed by the flexing of the plasterboard.

What Ed had ended up with was a room that seemed adequately damped at the low end, was somewhat overdamped at the high end, and virtually uncontrolled in the mid‑range. The practical outcome of this was that the mid‑range dominated the room, giving rise to the boxy tonality he's noticed.

The other potential problem we spotted was that he'd fixed speaker brackets to the rear edge of his left and right equipment racks, so that his Mackie 624 Active monitors were 'looking out' over a large expanse of reflective surface, which was producing comb filtering. Comb filtering occurs when the listener hears the direct sound from the speakers plus a slightly delayed sound due to a strong reflection from a nearby hard surface, the result being an unnatural tonal change and a lack of sonic focus.

We devised a practical strategy over coffee and chocolate Hob Nobs, but also suggested that Ed introduce some further high‑end reflective areas further back in the room, using old album sleeves stuck to MDF, or more glass‑fronted pictures. We've also used unwanted CDs as high-frequency reflectors with some success in the past: anything with a hard surface will work. The front of the room was slightly less problematic at the high end, as Ed had a control-room window to his left and several pictures to his right, though we thought some hard, irregular scattering surfaces on the ceiling above him would improve things further, to add a touch of diffusion.

Our plan needed to be both simple to implement and pragmatic, so the approach was two‑fold: firstly, we'd control the excessive mid‑band reflections at the monitoring position, using two-foot-square, two‑inch-thick acoustic foam panels; and secondly, we'd relocate the monitors onto foam speaker pads on top the rack boxes, so as to reduce the amount of reflective area in front of them.

Works A Treat(ment)

We used some two‑inch thick foam panels and corner traps to add some control over mid‑range and low frequencies. This, combined with a few more reflective surfaqces, helped to restore some sense of tonal balance to the control room. We'll leave Ed to do something about the lighting!

Universal Acoustics had sent us a kit of their foam panels, corner traps and speaker pads to look at, so we opted to try them on this project as a practical test. One potential problem with the room geometry was that control‑room window was located exactly where we needed to place the left‑hand 'mirror point' absorber. The solution turned out to be to have one fixed foam panel directly forward of the window, and a second one further back that could be slotted into the window space when mixing, but stowed elsewhere when tracking. We suggested that this movable panel might also come in useful when recording vocals in the control room (which was quite possible, as Ed's computer was so quiet), as an absorber to hang behind the singer's head to mop up 'over‑the‑shoulder' wall reflections.

The setup we ended up with comprised two panels each side of the listening position, at the 'mirror points', with a couple more facing across the room just behind the control‑room window. Two further panels were placed behind the monitor speakers, and as we had four corner bass traps with us, we fitted a pair in each of the two front corners of the room.

On checking the sound again, we found that it was much improved at the listening position. The mid‑range honk was virtually gone, the top end was clearer, and the stereo imaging was now nicely defined. Our forward speaker placement left a very stable phantom centre image between the speakers, which was lacking before, and listening tests using a busy bass part showed that the low end had no seriously 'hot' or 'cold' spots. There was still a lack of bass when standing close to the centre of the room, but at the monitoring position and for clients sitting on the sofa at the rear wall, the balance was actually pretty good.

Ed then asked about fitting a wooden floor, which, with so much carpet elsewhere, we felt could only be an improvement — a judgment based on experience of my own studio, which isn't that different in size to Ed's. I have a laminate floor, and it sounds fine: reflections from the floor tend to be directed away from the listening position so the usual outcome is that the sound gets slightly more lively but not excessively so, and in this room the extra HF reflection could be beneficial. However, because the client sofa was right at the back of the room, there would be a risk of comb‑filtering effects there from floor reflections combining with the direct sound from the monitor speakers. A thick rug towards the back of the room might prove advantageous in this case, and is something that is easy to add if it proves necessary.

Mixing Tips

We experimented with different mics on Ed's guitar amp and cab. His SM57 gave plenty of bite, but not the sound he wanted. A Shiny Box ribbon mike gave a much fuller sound, but at the expense of some of that bite and top‑end. Either combining the two mics or EQing the ribbon mic signal gave a result Ed was happy with.

Ed's background was as a musician (he's a really good guitarist) and in live sound, so he was a little unfamiliar with some of the studio-specific plug‑ins at his disposal. He was particularly concerned that his mixes didn't sound very loud next to commercial mixes, so when we suggested he record at an even lower level to avoid the clip LED coming on at the main mixer output, he looked a bit worried. We went on to explain the concept of a limiter as a fast compressor with a very high ratio that could arrest peaks before they clipped, where the output ceiling level could be set by the user. Ed had a limiter plug‑in that worked in exactly this way, so we set it up and demonstrated that by using the limiter to trim 3 or 4 dB off his mix peaks, the end results would sound adequately loud without the sonic damage clipping causes. Hugh then went on to explain a bit about the 'loudness wars' and how mastering engineers were getting fed up of being told by record companies to, in effect, make loudness a priority over audio quality, leading to some very unpleasant‑sounding records being made. The irony is that if you use excessive compression and limiting to make a mix sound over-loud, it becomes very fatiguing to listen to so the listener's usual reaction is to turn it down!

We then turned our attention to mixing vocals, and explained the idea of automating the vocal track level to eliminate large-scale level variations, so that less compression is needed to create a homogeneous sound that will stay up‑front in the mix. If you use only compression to try to level a vocal track that has a lot of level variation, you run the risk of over‑compressing the louder parts and ending up with compromised sound quality.

Ed was also a little unsure as to what vocal reverb to use, and had already discovered that reverb has the effect of pushing the vocal back into the mix rather than bringing it up front. One of the solutions we explored in the context of his existing rock songs was to use a short (around 800ms), fairly bright reverb with 80 to 90ms of pre‑delay. Then, if you increase the level of the early reflection part of the reverb and decrease the level of the reverb tail, you can create a very in‑your‑face sound that still retains a sense of space and power. We tried a few experiments on mixes Ed was working on and he agreed that this approach was going to be useful.

Recording Advice

Ed in his newly re‑treated studio.

Ed then moved onto the subject of recording the electric guitar. To date he'd been using an SM57, but had found that the results were a little 'lightweight', and didn't reflect what he was hearing from his Marshall amp (which drove a 4 x 12 speaker cabinet) in the room. As luck would have it, I had with me a Shiny Box ribbon mic which had come in for review (see page 24 of this issue), so we decided to set up a practical comparison. We set up both mics alongside each other, initially aiming them midway between the centre and edge of the cone of one of the speakers and about three inches from the grille. We recorded the two mics onto two tracks of Pro Tools and then auditioned the results.

Ed could barely believe the huge difference in sound between the two mics. His SM57 captured the 'bite', but made the guitar sound much brighter and less punchy than we heard it in the room. By contrast, the ribbon mic captured all the punch and depth of the sound but was a little dull at the high end, although this was easily rectified by adding a few dB of EQ boost at around 3kHz — and of course, there were also some interesting sounds to be had by combining the outputs from the two mics. We then went on to try the SM57 angled so that it was around 45 degrees off‑axis to the cone, a position which softens the high end somewhat — again, Ed found this a useful experiment.

Our final recommendation was for recording vocals, which could be done pretty well in the rear half of his control room if he didn't want to use the studio. We explained that the mic position is best somewhere between the wall and the middle of the room rather than being too close to either, and that an absorber behind the singer is a good idea, as when using a cardioid mic most of the unwanted spill comes from reflective surfaces directly behind the singer's head. We had a couple of sE Reflexion Filters left in our stash (thanks to Sonic Distribution), so we fitted one of these up in front of the mic to shield the rear and sides from spill and to reduce the level of vocal sound getting out into the room. A combination of Reflexion Filter and either foam or duvet behind the singer produces far better results than a badly designed vocal booth. As usual Hugh showed Ed how to re‑jig the mounting hardware to get the assembly closer to the centre of gravity of the mic stand, though Ed had a spare non‑boom mic stand, which he could use to hold the Reflexion Filter directly via its threaded mounting point, and Hugh then demonstrated how a boom stand could be arranged behind to suspend the mic in the Reflexion Filter from above. That accomplished, we packed up our tools and left Ed to try out his newly treated room. We never did get to see the sea, though!