With the speakers mounted on their stands, reflections from the computer monitor and gear racks were interfering with the sound. By moving the speakers forward onto the desk, Hugh reduced this problem, and he also placed bricks under them to put the drivers at a better height for seated listening — these bricks also reduced the resonance of the table top on which the speakers were standing. For long-term use, Paul suggested routing holes into the table top and fitting the original speaker stands through the holes.
The team are back in London to help some high-profile remixers sort out the monitoring problems in their new studio.
This month saw us back in London at the premises of a team of dance-music composers and remixers who go by the name of Above & Beyond, and who own and operate the Anjunbeats record label. This group comprises Jono Grant, Tony McGuinness, and Paavo Siljamaki, and they've already enjoyed a healthy amount of commercial success, with regular appearances in DJ Mag's top 100 and remixes for Madonna, Dido, Ferry Corsten, Delerium, and Zoe Johnston. They are Radio One regulars and regularly DJ at serious clubs such as Cream, Crasher, and Godskitchen. The latest news is that their nomination for the BBC Radio One Essential Mix has just been announced the outright winner, which is probably why the guys brought in two packets of Hob Nobs rather than just the one!
The Above & Beyond studio had just moved into North London premises previously occupied by another small studio, and it comprised a single room with a vocal booth built into one of the rear corners, plus adjacent office space. Although the previous owners had applied a generous amount of acoustic treatment, this seemed to comprise a fairly thin layer of Rockwool covered by fabric and it wasn't in the best place to be effective. For example, the side and front walls were damped only above the head height of a seated engineer; the rear wall and ceiling were untreated; and the vocal booth had the same thin treatment on the rear wall, ceiling, and one side wall, with a reflective door, window, and side wall left untreated. Apparently the original owner's monitoring system had been soffit-mounted Tannoys (inset into the wall), and the vacated soffits had since been completely filled with Rockwool, which meant they now functioned reasonably well as bass traps, albeit without intention in that direction.
For monitoring, the guys had a pair of Mackie HR824s augmented by a Mackie subwoofer, while the main recording system was Logic Pro v7 running on a Mac G5 and viewed via a splendid Apple 23-inch LCD monitor. However, it turned out that the team have very strong ideas about audio quality and prefer the sound of analogue mixing to digital mixing, so they bring 24 tracks of audio out of Logic via three RME ADI8 units hooked up to an RME Hammerfall soundcard. These feed into a Soundcraft Ghost analogue mixer. They were also experimenting with some high-end preamps and had already come to the conclusion that their impressive collection of analogue synths sounded noticeably better when fed into the audio interfaces via a good preamp rather than being plugged in directly. They were experimenting with a loaned Neve VR preamp while we were there.
Preliminary Listening
We were greeted by Jono, who explained that their monitoring sounded wrong, with an inaccurate low end and poor stereo imaging, and that the vocal booth was unusably boxy. After the obligatory coffee and biscuits, we first turned our attention to the Mackie HR824s, which were mounted on good-quality stands behind the custom-built furniture that Tony had designed, but this turned out to be less than ideal, as the two rack units and the large LCD screen were in front of the monitors where they could reflect and diffract the sound. It also placed the monitors at a fairly shallow angle to the desk surface, inviting surface reflections from the table top. The subwoofer was placed on the front wall off to one side in a somewhat inaccessible spot behind one of the built-in floor rack pods, though it eventually turned out to be close to the ideal position as far as the sound was concerned. A footswitch had been rigged up to bypass the subwoofer, which made setting up a lot easier for us.
In order to test the bass response of the monitoring system, a semitone scale sequence was created in Logic triggering a synth sound with the filter set low to remove upper harmonics. However, this proved unreliable in practice, as the levels of the upper harmonics varied unpredictably, so the EXS24 software sampler was used instead — its default sound is a straight sine wave, which was ideal for the purpose.
Listening tests soon showed the bass end to be rather too hot, and the stereo imaging was also extremely poor. As the studio was said to be producing 'bass light' mixes, the over-hot bass end didn't come as a surprise. We decided that the best course of action would be to drag the subwoofer into a slightly more accessible position so that we could reach the controls. We also decided to move the speakers off their stands and mount them on the desk surface, around a foot forward of their current positions. This would move them in front of the LCD monitor and the two side rack pods, which would reduce the diffraction and reflection problems significantly. However, the table top was far too low for them, so Sachi, the studio manager, managed to rustle up a few fairly clean bricks from somewhere to help add mass and dampen any table-top resonances, while also raising the speakers sufficiently to place the tweeters near ear height and angled towards the listening position.
During this relocation, we checked the HR824's rear-panel switch settings and discovered that the HF gain had been turned down by 2dB because Jono felt the speakers were slightly over-bright. This was fair enough, but we also found that one of the speakers was set to the correct 'half space' mode for use close to a wall, but the other was set to 'full space', which would have thrown the low end out of balance, at least when the speakers were used without the subwoofer.
Subwoofer Settings
We settled on starting with the speakers set completely flat and with the room-position switches set to 'half space'. Once the subwoofer was dragged into a position where we could get at the controls, we tested it with a two-octave chromatic sequence of bass notes played via a Logic synth with the filter closed down to give a sine-like tone. However, the varying level of additional harmonics from this sound source became a distraction, so we switched to using the EXS24 sampler, as its default waveform is a pure sine wave. The subwoofer crossover had been set to its lowest setting of 55Hz, but we reasoned we'd get more headroom from the system if we moved it up to 80Hz, as less low end would be fed to the main speakers. (The Mackie subwoofer crossover feeds a high-pass signal to the main speakers, and when the footswitch bypasses the subwoofer, these filters are taken out of circuit.) As expected, the subjective level of the scale got progressively higher as the pitch increased, but we noticed that the level climbed, dipped, and then climbed again, so we reasoned that something was happening around the crossover point.
A curious dip in the bass response led Hugh to suspect phase problems between the subwoofer and the monitors. A quick trip under the desk to flip the phase-inversion switch sorted out the problem, but further careful listening and adjustment was required to get the subwoofer level right in relation to the monitors.
Hugh set the phase switch on the subwoofer to invert, which cured this dip, so it was clearly the best setting for this particular subwoofer position. Because of the layout of the furniture and the cabling to the subwoofer, it wasn't practical to do the usual trick of hauling the subwoofer to the monitoring position, then crawling around the room to find the most even-sounding spot, but if we had, I've a feeling it wouldn't have been too far from where the subwoofer was anyway. It certainly sounded fairly even and well-balanced.
Getting the subwoofer level right was trickier, especially as the guys had turned the HR824 input gains quite a long way down to compensate for the unusually high output level from the Ghost desk. After much messing around, we decided it would be best to simply turn down the monitor output from the desk and then start with the main and subwoofer speakers set to their default 'detent' positions. When switching back and forth between having a subwoofer and no subwoofer, this sounded just slightly bass heavy at the upper end of our test scale where we reckoned there should be little difference with or without a subwoofer, so we backed it off slightly until the subjective level was the same. This added more depth to the very low notes but didn't significantly affect the higher bass registers. Tony found a sine sweep test signal and this again seemed to confirm that we'd got the settings about right.
To make the subwoofer easier to move on carpet, we suggested fitting it with self-adhesive PTFE glider pads that are sold in DIY shops as an alternative to castors. Hugh fitted these to his A100 Hammond organ and Leslie, which he can now move on carpet with just one hand, so they do work effectively.
Hugh suggested that, as the console monitor output control was now less than a quarter of the way up at normal listening levels, attenuator pads should be fitted on the desk outputs to reduce the signal by 10-20dB, in order to allow the desk control to be set to a more realistic position and also to avoid the monitors flying off their perches if someone turned the level up full! In-line balanced XLR attenuators are readily available from suppliers such as Canford Audio and Studiospares.
Stereo Imaging Improvements
Now it was time to improve the imaging by applying a relatively small amount of four-inch acoustic foam, kindly provided by Auralex. In small rooms where an 'empirical' approach to acoustic treatment is appropriate, the most successful approach seems to be to locate absorbers on those surfaces that are currently reflecting sound from the monitors directly back to the listening seat. You can verify you have the right places by getting somebody to hold a mirror flat against the wall and then move it around until you can see the monitor from your mixing chair. This position determines the centre of where the absorber needs to be. Do this for both side walls and the ceiling.
The stereo imaging was still quite messy even after the monitors had been moved to better positions, so acoustic foam was fitted to the ceiling and side walls to minimise interfering early reflections — the foam panels on the right-hand side covered a window, so they were fixed with velcro to make them removable when natural light was more important than monitoring accuracy.
We found that, in the case of the side walls, the foam absorbers needed to be placed below where the previous occupants had installed their shallow Rockwool absorber panels. On the left-hand side the ideal place happened to span a window, so Velcro was used to make that particular panel detachable. The permanent panel was fixed easily to the wall with spray adhesive, but the self-adhesive Velcro strips we used on the other panel were initially less than successful, as they kept peeling away from the foam. To solve this, we tried spraying a thin coating of adhesive onto the foam, waiting for it to go tacky, then sticking the Velcro to that — it worked a treat. All we needed in the end were three 2 x 4-foot panels, one on each side wall and one on the ceiling. One little tip here: if you get spray adhesive or mastic gun contact adhesive on the exposed side of the foam, you can usually get it off quite cleanly while it is still wet by dabbing at it with the sticky side of gaffer tape — but don't let it dry first.
Listening tests confirmed that the imaging had improved quite dramatically after moving the monitors and adding the new foam absorbers to the sides and ceiling — the last making a surprising amount of difference. The rear wall was partly obscured by a settee — always a good feature in a studio (both ergonomically and acoustically), but there was also a small window onto the adjoining office that acted as an unwelcome reflector. As the window wasn't really needed, Jono said that hanging a curtain over it would be a simple option.
To take the edge off the remaining bare wall, I applied a 2 x 4-foot sheet of self-adhesive, one-inch dimpled foam donated by Sonic 8 — this is too thin to be used as the sole treatment in a room, as it is only effective at high frequencies, but when it is used to augment existing treatment, it works fine and sticks well to flat surfaces.
A final touch was to cut a seven-inch strip of the self-adhesive foam and use this to cover a flat wooden strip directly behind the monitors.
A few offcuts of acoustic foam were pressed into service to reduce reflections from the wall behind the monitors.
We didn't feel that additional bass trapping would be necessary, as the filled soffits, the doors, and the double-glazed windows already acted as bass traps, though there was a void above the vocal booth that could be filled with Rockwool if further bass trapping were to be needed. The main purpose of bass trapping is to ensure an even bass response, rather than simply to soak up bass, and we'd already achieved that as well as we could hope for in a room of this size.
Vocal Booth Treatment
With the studio sorted out, it was time to take a look at the vocal booth, and we tested this by setting up a mic and getting Tony to do a spoken monologue while we listened in the control room. Sure enough it sounded dull and boxy. We tried again with Jono speaking, just in case Tony had a dull and boxy voice (which he didn't!), then set about looking for the cause and a possible solution.
A panel of thick acoustic foam reduced the boxiness of the vocal booth nicely, but there was still no life in the sound at high frequencies.
Most bad-sounding vocal booths are made that way through too much absorption in the upper mid-range and high frequencies, with inadequate absorption at the lower mid-range and bass end. This can be due to using too great an area of treatment and/or using treatment that is too thin. In this case it was clearly the latter, as the inch or so of Rockwool that had been used would become progressively less effective as the frequency fell below 800Hz or so. My own view was that we should use thicker absorbers behind the singer (on the rear wall) and also try to introduce a little high-frequency scattering or diffusion. We could also use patches of self-adhesive foam just to take the edge off the areas of completely untreated surface within the booth.
The upper part of the rear wall included a cavity for a light (which wasn't really necessary), and it seemed that this was contributing to the boxy feel. To test the theory, I wedged a piece of four-inch Auralex foam across the top of the rear wall and stood another piece on end at the bottom of the rear wall. A further 2 x 4-foot panel was glued to the inside of the hard door. This certainly dried up some of the 'honk', but the top end was still lifeless. We needed to improvise some high-end scattering, but what could we use? What does a small record company have lots of that are small and sonically reflective at high frequencies? Demo CDs, or course!
By taping old demo CDs to the walls of the vocal booth at around head height, enough high-frequency reflection was reinstated to provide a nicely balanced overall sound.
Just to test the theory, we taped a couple of rows of CDs to the rear wall and to the padded left side wall at about head height. It really made a difference, adding just the right amount of brightness and presence. The spoken monologues now sounded much more natural, with none of that 'broom cupboard' sound that had beset the booth originally, so we declared the experiment a definite low-budget success. The guys said they'd do the job more permanently and more neatly using spray adhesive. They quite liked the look — using discs with nice label artwork would look pretty good and could be made into a nice feature. There was also room to put a few CDs on the ceiling if necessary.
A few thinner pieces of foam were placed on the wall behind the singing position to further reduce boxiness.
We noticed that the door of the booth didn't seal too well and so suggested fitting a compression door latch that would pull the door up against the seal to make it a better fit. There was also no permanent connection panel in the booth (the mic cable had to be run through a gap at the bottom of the door) so this was also added to the 'to do' list.
The temporary speaker relocation on bricks had worked well, but was not a very satisfactory long-term solution from an aesthetic point of view. So we suggested cutting a couple of large circular holes into the desk using a router, directly below where we currently had the monitors set up. They could then use the original speaker stands, projecting up through these holes (the top and bottom plates can be easily removed to facilitate this) to support the speakers in a more aesthetically pleasing way, while also isolating them completely from the desk.
Once again we'd shown that you can make a big difference to how well a studio room performs for relatively little cost and with minimal disruption. Hopefully, the guys will now get even better mixes from their new studio than from the old one.