The SOS team return to Cambridge, turning another home studio upside down in search of improved vocal and guitar sounds.
Jazz and Alessia met while on an SAE course in Glasgow, and have now set up their own home studio in Cambridge to enable them to carry on composing and producing. They've already had some success in this field while studying in Glasgow — Jazz and his former production outfit had the same management company as Ms Dynamite, and did both composing and remixing work for the company. The work included remixes/work for Ms Dynamite, 50 Cent, Lisa Mafia, Da Endz, Working Title (for the film 'The Guru'), Lamya, and various Bhangra artists. Alessia writes, sings, and co-produces with Jazz. The pair are currently finishing their own album and hope to produce other artists as well as themselves in their own project studio.
The studio is set up in a first-floor room that is relatively long and narrow. After Jazz contacted us asking for help he sent over some photos of the room, and Technical Editor Hugh Robjohns and I both suspected that there might be monitoring problems to address. However, Jazz specifically wanted help with his vocal recording and mixing.
When we arrived, the desk supporting the computer monitor and a new pair of ADAM P11A active speakers was set up approximately halfway along the length of the room, the idea being to divide the room into a control-room area and an area for recording vocals. Jazz was toying with the idea of building a vocal booth in part of the available performing space. The studio setup itself was pretty simple, with a Mac G5 running a Digidesign Digi 002 rack interface, and Ableton Live was used for creating and assembling ideas before mixing in Pro Tools LE. For processing, Jazz was using only the plug-ins that come with Pro Tools LE, though it turned out that these worked pretty well for the type of music he was working on.
As the ADAM monitors were new, Jazz wanted to check that they were performing correctly, but after playing a selection of material through them, it became evident that the level of bass dropped away rapidly if you moved back from the speakers, and the computer monitor was also set up slightly forward of the speakers, which can cause problems with reflections, particularly at high frequencies. In addition to the low-end consistency problem, the bass was not as well defined as it should have been and the stereo imaging was quite poor because of the proximity of bare plaster walls to either side.
Fortunately, we'd brought three large Auralex foam panels with us (supplied by the UK distributor Audio Agency) that we felt would improve the imaging if placed on the side walls. We also suggested that a pair of foam isolator pads under the monitors would improve the solidarity of the bass end, given that the monitors were set up on a computer desk shelf rather than on rigid stands. We improvised a pair of these using the packing that the speakers came with and, even though they were not ideal for the purpose, the sound improved noticeably, so we arranged for a pair of Auralex's MoPad speaker supports to be sent on later so that Jazz could evaluate the final sound.
While our little fixes improved the general sound and imaging of the monitoring quite noticeably, the unreliable bass variation with listening position wasn't so easy to fix — we've found that most small rooms have a Bermuda Triangle (well, a Bermuda sphere really!) in the centre where bass sinks without trace if you mix from there, and if you put your monitors there too things get even more uncertain.
After talking about what Jazz and Alessia hoped to get out of the room, we persuaded them to try moving everything around so the monitor desk was at one end of the room, facing the wall. In long narrow rooms, it is invariably better to fire the speakers down the length of the room rather than across it, as the reflections from the close wall behind the engineer tend to be very difficult to deal with when working across the room.
We spent a few minutes removing everything from the bookshelves currently occupying the end of the room where we wanted to set up the desk, then began clearing that half of the room ready to set up the studio in its new location. I even vacuumed the floor before we moved the desk — this is getting to be a habit — while Alessia polished the bare cupboards. To reduce the risk of ground-loop hum, we used a 'star' mains system, which is to say we plugged a distribution board into one mains socket, then ran other distribution boards from this so that everything was powered from the same place. Hugh taped one of the distribution boards to the rear support of the computer desk to get the wiring off the floor and to give me yet another excuse to photograph him clambering under furniture!
Jazz re-connected the G5 and the Digi 002 as we tried to organise the rest of the room. In a very short space of time, we had a much more spacious setup with all the necessary gear within easy reach of the engineer's seat and with plenty of clear floor space for recording vocals and guitars. During this process, the impressive mound of top-quality chocolate biscuits provided by Alessia diminished greatly!
The studio's car-boot-sale Hammond organ (a bargain at £10) was repositioned to the other side of the room, and the bookshelves were moved to the back of the room, against the side wall close to the bay window. A new purpose-designed table was being built to support Jazz's turntables and DJ mixer, and it would sit neatly between the bookcases and the keyboard rack.
As the house belongs to Alessia's father, I suggested that we prop the Auralex foam up in a temporary fashion to find the best place for it, then Jazz could procure some quarter-inch MDF panels, stick the Auralex to that, and hang the panels up like pictures using hooks. This would save the mess that invariably ensues if you ever need to remove foam that has been stuck directly to the wall. Either spray carpet adhesive (as supplied by Auralex) or a Liquid Nails-type adhesive works well for this.
Both Jazz and Alessia thought this was a good idea, so we proceeded on that basis, putting one panel on each side wall with the vertical centre of the panel at about seated head height, and the horizontal centre of the panel just forward of the listening position. Fortunately a radiator on the left wall gave us a temporary mounting option in about the right place, while a shelf unit on the right-hand side was almost as obliging. The third panel was propped up centrally on the wall behind the speakers using the desk as a support.
Playing our test songs now confirmed that we had a much more even bass response as we moved back from the normal listening position, and the bass was also more solid and far less 'tubby' sounding. The foam at either side brought better focus to the stereo imaging as well. Jazz mentioned getting some foam bass traps, and we agreed this would be a good idea — he'd probably want to place them over the full height of the two front corners of the room.
With the monitoring now performing far better, we felt we could move on to the vocal-recording problems. The couple had recently bought a Neumann TLM103 vocal condenser mic, which we set up in its shockmount with a pop shield positioned a couple of inches in front of it. Alessia volunteered to do some singing so we could check our results, but before doing that we loaded in a couple of songs she'd already been working on to see what the vocal was like when heard in isolation.
Both Hugh and I were extremely impressed with Alessia's contemporary vocal style, but we could hear some room coloration clouding the lower range of the vocal mic. There was also a tendency for the mic to sound as though it was on the verge of popping sometimes. Alessia often records vocals when Jazz is at work, and she usually hangs a double duvet off the curtail rail that runs across the bay window. Duvets can be great low-cost problem-solvers, but there were so many bare plaster surfaces on the other walls that some room coloration was still clearly audible.
An impromptu solution was to pull the curtain across the bay window (it was fairly heavy and absorbed sound reasonably well) and then to hang the duvet on the adjacent side wall by tying loops of string to the top two corners, then looping these over the ends of the bay-window and side-door curtain poles — these were almost ideally placed for this purpose. We now had a corner that was reasonably dead on two sides, so if the vocals are recorded with the mic facing into the corner, most of the rogue reflections should be intercepted before getting into the mic (which has a cardioid polar pattern). Using a thicker duvet over the bay window curtain might also help. There was space on a roof beam to hang yet another double duvet to act as a screen between the booth and control room areas, and which would effectively form a three-sided vocal booth. Alessia also suggested hanging another duvet over the door to hide the reflective wood surface — every bit helps!
We felt that working like this would be better than trying to build a separate vocal booth, as these nearly always sound boxy and are quite difficult to treat in an acoustically neutral way.
To test our 'one duvet and adjacent curtain' setup, Jazz opened up a song they'd been working on and Alessia overdubbed a new vocal verse. This was also to serve as the test-bed for any processing we decided to try. She worked about six inches from the pop shield with her back facing the duvet. Just one take and we had a perfectly respectable piece of test vocal. We never did figure out why she was getting traces of popping before, as she always used the pop shield and also the bass-cut switch on the 002 preamp, although it was possible that the pop shield had been arranged too close to the mic itself — there should be a gap of a couple of inches of still air between the shield and mic. However, with our setup, we didn't get any hint of popping at all.
Listening back to what we'd recorded, we noticed straightaway that the room coloration was much reduced and that we had a good basic tone to work with before using EQ or other processing. Jazz had been using the Pro Tools reverb, but I felt a much shorter, brighter setting than the one he'd been using would better suit both the song and Alessia's voice. I hadn't used the Pro Tools LE plug-in reverb before, so I was interested to see what I could get out of it. I ended up using the medium plate with the decay time turned right down to 650ms and with a 70ms pre-delay. I left the filters at a fairly bright setting and turned the diffusion down to 45 percent to try to coax something closer to an ambience effect out of it. I didn't want an obvious reverb, but the vocal needed a sense of life and space.
This combination of parameters worked astonishingly well, and both Jazz and Alessia really liked the result. Jazz had been considering buying and using an outboard reverb, so we patched in a Lexicon MPX550 that I'd brought with me and set it up to give a similar sound to the plug-in reverb. None of us felt that it was any better in this application — which pleased Jazz, as he didn't want to have to budget for an outboard reverb if he could help it.
Although the new room arrangement and duvet hanging had removed most of the room coloration, there was the mildest hint of boxiness to the vocal sound which we tamed with a very gentle 200Hz dip using the Pro Tools LE four-band EQ plug-in. I also added a little 'fairy dust' boost centred at 10kHz to give a hint of 'air'. The amounts of EQ gain involved were only 2-3dB, but the vocal clarity improved and it felt much more upfront in the mix, which is what Jazz and Alessia wanted to achieve.
The final tweak was to add some compression. Alessia has really good mic technique, so compression wasn't needed to keep the level even at all. Instead, we were able to use it to add a little 'punch' and to create a more contemporary sound. We settled on a hard-knee, 8:1-ratio compression with an attack time of 3.7ms and a 100ms release time. The threshold was adjusted to give a maximum gain reduction of around 6dB on peaks and it was working pretty much all the time Alessia was singing.
I felt the result sounded really crisp and solid, and as a further demonstration of compression as an effect, I pushed the gain reduction up to around 15 to 20dB (by reducing the threshold) to show how compression could be used to produce a more assertive sound — almost like singing through a limiter. Of course, compression brings up background noise, breath sounds, and room ambience between phrases, so you really need good acoustics and a quiet room to get the best out of this technique. Using a gate or expander on the vocals prior to compression can also help.
Jazz then asked about panning strategies for backing vocals, as Alessia likes to layer a considerable number of different harmony parts. Hugh pointed out that extreme panning can lead to perceived balance problems when the mix is heard in mono, so we suggested keeping the main vocals in the centre as usual, panning the main backing vocals to around 10 o'clock and two o'clock, and saving extreme panning for any extra backing vocals that wouldn't cause the overall mix to suffer too much if they lost a little level when played back in mono. It's also possible to add space to a multi-layer backing-vocal mix by rolling out all the low end, say below 200Hz, and adding more reverb to those parts that are panned towards the edges. This retains transparency in the mix and also adds a nice gloss to the production.
Our final task was to try miking Alessia's Encore nylon-strung classical guitar. This turned out to be very easy, as classical guitars tend not to boom in quite the same way as metal-strung models, so you don't have to worry as much about aiming the mic too close to the soundhole. All acoustic-guitar recordings seem to benefit from some early reflections, either from a hard wooden floor, or from some kind of 'acoustic mirror'. The only suitable sized reflecting board to hand was one of the computer desk's side shelves, but this served the purpose when supported on a handy cable box and angled to bounce some of the guitar's sound back into the mic.
Jazz was considering replacing the studio carpet with laminate flooring, but I felt this would be a disadvantage, as the noise from the computer would become more evident with nothing to soak it up. A reflective floor isn't always great for vocals either, as it can colour the sound. The compromise solution is either to floor only a small section at the end of the room for recording acoustic guitars, or simply to put a wooden panel over the carpet when recording those instruments that need a bit of reflective help.
While wearing headphones, I moved the guitar relative to the mic until we found a sweet spot that came close to the natural acoustic sound of the instrument. I encouraged Jazz to move the mic around while listening on headphones to hear for himself how moving the mic small distances dramatically changes the recorded sound. In this case, the best result turned out to be achieved by aiming the mic from 12-18 inches away at a point midway between the end of the neck and the soundhole.
Although moving the whole studio around and re-plugging everything took a little time, I feel it really paid off in terms of more accurate monitoring and better use of the available space. The simple Auralex treatment took care of the worst local reflections around the listening position, and while our improvised speaker pads weren't great, they helped too. Switching these for the Auralex MoPads when they arrive should tighten up the low end even more, and adding bass traps to the room corners will further improve the evenness of the bass.
Jazz said that he might get a second computer monitor, in which case there wouldn't be room on his existing speaker shelf for both monitors and both speakers. One solution would be to cut off the metal vertical supports protruding above the current monitor shelf, fitting a wider shelf on top, but a more acoustically sound solution might be to use rigid speaker stands either side of the computer desk (provided that these don't place the monitors closer than around one foot from the side walls). Flat screen monitors aren't affected by the magnets in monitor speakers, so you can put them as close together as you like.
The simple duvet treatments really worked well for the vocal recording area, and adding more duvets, as suggested by Alessia, should improve things further. As stated earlier, this will almost certainly lead to better results than offered by a small, enclosed vocal booth. Our final experiments with processing confirmed that Alessia's already excellent contemporary voice could be made to sit better in the mix with just a little EQ, compression, and a short, bright reverb. Too much reverb can swamp a voice, but the treatment we chose actually helped pull it to the front of the mix. The final challenge before hitting the road and heading back to the Midlands was to eat a portion of Alessia's home-made pasta without coming back for seconds. We failed!
"Moving the furniture around has spread the bass evenly throughout the studio, and hanging the acoustic foam on the walls has also helped to give a clearer and wider stereo image. Another welcome side-effect of the reorganisation is that the room now appears larger and more inviting! Our vocal recordings are much dryer with a lot less coloration thanks to the correct mic positioning and duvet padding. It has made a huge difference overall — thank you SOS!"