In order to treat a marked room resonance at 110Hz, each speaker was first isolated from the desk, which was exacerbating the problem, using two heavy quarry tiles separated by the speakers' rubber feet. A further improvement was made by inserting a small piece of acoustic foam into each speaker's reflex port, raising the low-frequency roll-off point.
Musicality isn't everything when you're trying to make a decent-sounding recording at home, even if you're as talented as the multi-instrumentalist Harding family. So the SOS team travelled north to help them get their engineering techniques up to scratch.
Recently SOS's Sam Inglis attended a function at which the entertainment was provided by a family band comprising parents and their three daughters. Normally this type of line-up, especially when all three girls are still at school, is a recipe for hopelessly cheesy music and sympathy-based applause, but Sam (who tends to be very critical of live performances) returned full of praise both for the playing abilities of the band and their songwriting abilities. In a subsequent conversation with them, he discovered that they had bought a Yamaha AW16G recorder/mixer, a pair of Samson active monitors and three Shure SM58 microphones to set up their own studio at home. Furthermore, they even provided him with a CD of work in progress, the quality of which was surprisingly good and which would have put many a Demo Doctor entry to shame! However, the drum sound wasn't up to scratch, even though the playing most definitely was, and the overall balance and tonality of the mix could have been better, so Sam handed the CD over to myself and Hugh so that we could arrange a visit.
The Harding family band play a wide variety of styles of music including folk, jazz, rock, and pop. They have performed widely, as well as appearing on both local and national radio — they were even on BBC Radio Four in 1996 when they played for, and were interviewed by, John Peel. Tim Harding is a primary school head teacher and his wife Amanda is an archivist now working in schools presenting Living History workshops. They have been playing music together with their three daughters — Emily, Charlotte and Eleanor — since the girls were old enough to pick up an instrument, and the girls now play over 14 different instruments between them. Emily (aged 16) plays drums, flute, keyboards, percussion, violin, and sings backing vocals. Charlotte (aged 14) plays alto/soprano saxophones, clarinet, whistles and keyboards, and takes the lead vocal in most of the songs. Eleanor (aged 12) plays trumpet, violin, mandolin, whistles and keyboards, and also provides lead and backing vocals. Tim adds bass guitar, guitar, mandolin and trombone, while Amanda plays keyboards and violin. Emily and Charlotte currently write most of the material, and like experimenting with different styles of music.
Studio Treatment
Situated in a 17th-century house in a small village south of Kingston-upon-Hull, the band's studio comprised an extremely small control room (2.45 x 1.8m) adjoining a 4 x 4m live room that also doubles as a rehearsal space. Hugh and I were instantly suspicious of the monitoring accuracy of such a small control room, especially as there was no acoustic treatment at all, so while we partook of our coffee (and the obligatory chocolate biscuit!) we asked to hear a playback of some work in progress. Immediately our suspicions were confirmed, as the room obviously had a nasty resonance at around 110Hz that made the low frequencies very uneven, with almost a 'one-note' bass characteristic, tending to make mixes sound boomy. We agreed that we had to try to improve this before tackling any of the other problems, so we did some more listening using a couple of commercial recordings just to confirm the boom was a room problem and not a recording issue. It turned out most definitely to be a room problem.
After removing some cable storage hooks from the wall behind the monitoring position, some Auralex acoustic foam was attached to attempt to improve the acoustics in the control room. One corner of the foam needed to be customised, using a kitchen knife, to fit around an existing mains wall socket.Ultimately, we had to apply several remedies, each of which made an incremental improvement to the monitoring accuracy of the room. As the rear wall of the room was almost directly behind the mixing chair, we decided to treat this with a couple of panels of Auralex foam that we had brought with us. This necessitated removing a few hooks and fittings that had been used to store cables on the rear wall, and we also had to cut away the corner of one of the panels (using a serrated kitchen knife) to clear a mains socket. The panels were fixed with spray adhesive and perfectly matched the blue decor of the room.
We were also worried that having the speakers standing directly on a lightweight desk might cause resonance problems, so after confirming that the desk did indeed resonate at around the offending 110Hz when tapped, we decided to try raising the speakers off the desk by a few inches, first by improvising supports out of box files, then more permanently by using some thick clay quarry tiles left over from renovation work on the house. These Samson monitors come with detachable rubber feet, so we used two quarry tiles under each speaker with three of the rubber feet between the two tiles to act as an isolation mount (see main photo, above). This turned out to work pretty well, though we recommended placing some non-slip matting beneath the speakers as the tiles (what with being handmade and pretty ancient!) weren't entirely flat.
While moving the speakers, we also noticed that the mid-range frequency trims had been set to -3dB, so we reset them to flat and played our test records again. Sure enough, the 110Hz honk was less noticeable, but it was still there. I had an idea that the space beneath the desk was acting as a resonator and suggested packing some sleeping bags (one of the girls also suggested soft drum cases) under there as an experiment. We also tried putting some rubber car mats on the desktop to damp it. Again we heard a slight improvement, but it wasn't a complete cure, so another recommendation we didn't have time to try at the time was to cut holes in the desk sides so that it wouldn't behave quite so much like a tuned box. We felt it would also be worth trying to stiffen the desktop from below using something like 3 x 2-inch wooden braces.
Our final experiment turned out to be one of the most worthwhile. To reduce the degree to which the speakers were exciting the remaining room resonance, I pushed some scrap pieces of foam into the speaker bass port slots, the idea being to lower the Q value of the cabinets and start the bass end rolling off at a slightly higher frequency. This simple and easily reversible fix tightened up the bass considerably and, though the bass level in the room still wasn't perfectly even, individual bass notes could now be heard quite clearly and they weren't being obscured by the 110Hz honk. Given the small room size, we felt this was about as far as we could go in improving the monitoring environment, and everyone felt the sound was significantly tighter and more accurate.
During these tests we'd also discovered a low-level, but nevertheless annoying, hum on the monitors that didn't respond to any of the usual ground loop fixes (such as lifting screens). In the end, Hugh discovered that this was not caused by a ground loop at all, but that it was induced from a Korg X5D keyboard wall-wart power supply which was lying adjacent to the mains cables of the speakers and hard disk recorder. By moving the wall-wart away from these cables the hum disappeared, so we ended up simply plugging the wall-wart into another wall power socket, whereupon the hum was completely eliminated.
Sorting Out The Drums
The band had been working with their bass drum filled with bubble wrap, on the advice of another drummer. However, this was not producing a suitable sound for recording purposes, so it was removed and replaced with some heavy blankets. The front head was also removed, because the hole cut into it was too small to allow much flexibility in mic positioning.The drum kit turned out to be a rather nice Premier kit, fitted with decent heads and pretty well tuned. The band had been recording the kit using just their three SM58s, one on kick, one on the snare and one as an overhead. Their demo recording had a particularly tubby kick drum sound, and though the SM58 isn't an ideal kick drum mic, due to its lack of extended bass response, the real problem was traced to the way the drum was tuned and damped. A friendly drummer had recommended damping the drum by filling it with yards of bubble wrap, but it didn't seem to be doing nearly such a good job at damping as the traditional folded blanket!
Unfortunately, the front drum head had only a small hole cut in it (one of the less fortunate current fashions as far as recording is concerned) which meant there was insufficient room to get a blanket into the drum and no way to get the mic into the proper position, so we had to resort to removing the front head and working without it. My preferred option is to use a front head with a large hole cut in it, as that not only provides better access, but also leaves less of the head to resonate — something that can lead to a messy or flabby kick sound. Leaving the front head off in the long term is not recommended, as lesser drum shells may distort under the uneven tension, although I've never known this to happen. A more likely problem is that you may lose the fittings, or the tensioner nut boxes may rattle — as it was, we only had to spray the pedal with WD40 to cure a squeak.
Even when the front head had been removed and the bubble wrap replaced with blankets, the sound still needed more attack and less ringing. A little tissue gaffered to the skin and an ID card taped over the beater contact point helped here.Removing the head and putting a couple of folded blankets into the drum tightened up the sound, but I still felt the drum was tuned a little too high, so I dropped the pitch slightly, which gave a more satisfying splat at the attack of the note. The pedal was fitted with a hard felt beater, but for recording a solid wood or plastic beater usually gives a better tone, so we improvised by sticking a plastic ID card to the head using gaffer tape, positioned so that the beater hit the centre of the plastic card. Everyone agreed that this was much more like the kick sound they were after, as it had both depth and definition.
The other drums were tuned pretty well, but needed a little more damping, which was provided using small amounts of folded tissue held in place using masking tape or black gaffer tape. (Which wise man was it who said that gaffer tape is like 'the force' in Star Wars? It has a light side, a dark side and it holds the universe together!) It is very important to damp the lower tom heads slightly, as these will ring whenever the kick drum is played, and in most instances this ring will be picked up and recorded, making the kit sound boomy and ill-controlled. Although this can be cured to some extent by the use of gates, setting up can be tricky, and it's easy to make matters worse rather than better.
As luck would have it, I had brought a set of Samson drum mics with me, so I set them up on the kit, close-miking each individual drum and using the two included capacitor mics as overheads. Because the AW16G has a limited number of mic inputs, and because only two of those are on XLRs, we used the band's Peavey XR800F powered mixer as a submixer for the drum mics, and for the purposes of our tests these were fed directly to two tracks on the recorder. However, for maximum flexibility when mixing, it would be better to use two more tracks to enable the kick and snare mics to be kept separate, as this allows the snare reverb to be controlled independently and also makes it possible to adjust the kick and snare balance when the rest of the mix is up and running.
A pair of condenser mics used as stereo overheads really helped give a more professional sound. Positioning them a little further back gave a less cymbal-heavy balance, which meant that only a little of the each close mic was required to focus the sound of the drums.With just a little mid-range EQ cut on most of the close mics, we ended up with a more than respectable drum sound, and the best results were achieved when we used the overhead mics as the basis for the sound, then brought up the close mics just enough to create a good balance between the drums and cymbals. Using the overheads to provide the main body of the sound also lessens the effects of ringing drum heads, as these are picked up mainly by the close mics. Close mics on their own can sound a little dull and ponderous, while in this case the overheads alone were bright and lively, but noticeably cymbal-heavy. We positioned the overheads slightly towards the rear of the kit to reduce the volume of cymbals that they picked up. Getting the ideal balance required making several test recordings, then listening to the playback, as the acoustic isolation provided by the light, ill-fitting 'historic' door between the control room and live room was more psychological than real! I didn't mention neoprene seals and heavy fire doors as I didn't think this was the feel they wanted...
Having proven that we could get a good drum sound, we changed the rig to use two of the band's own SM58s (kick and snare) plus a single Samson C03 budget capacitor mic in cardioid mode (with the -10dB pad and high-pass filter switched in) as a mono overhead. We didn't use any close tom mics in this instance. Our overhead mic was pointed roughly at the drummer's lap so as to avoid picking up too much cymbal and to give the toms a chance. The reason for using a single overhead mic and for bringing back the SM58s was that the band said they could afford to buy one capacitor mic for recording vocals, brass/woodwind and drum overheads, so we wanted to see what results could be achieved that way. Obviously, you lose the stereo feel of the drum kit, but a little stereo reverb can help restore some of the missing spatial ambience.
Once a successful sound had been achieved with stereo overheads and individual drum mics, Hugh set about achieving a comparable sound with a more modest setup of two dynamic mics and one condenser.After a few minutes getting a good balance, we ended up with an acceptable drum sound that had good definition and tone, but where the toms were a little quieter than ideal and the kick drum lacked the depth of the dedicated kick mic we'd used earlier. Nevertheless it wasn't bad at all, and was a great improvement on the band's original recorded drum sound.
Sax & Vocal Miking
For live performance, the sax is currently miked using a small clip-on mic fixed to the bell of the instrument, though studio recordings (always made as separate overdubs) had been captured using one of the SM58s pointing at the bell of the instrument. The sound wasn't bad, but it lacked definition, and some of the more subtle nuances of the sound had been lost. We hung up as many blankets, duvets and sleeping bags as we could find in one corner of the live room to provide a non-reflective backdrop, then used the Samson C03 mic (again in cardioid mode) around 14 inches from the body of the sax and aimed at the body of the instrument a couple of inches or so above the bell. With no EQ or further processing, this immediately produced a nicely moody sound that Lisa Simpson would have been proud of, with lots of air and definition combined with the warmth. A hint of reverb and we were all happy with the result. Given that the Samson mic is around the same UK price as an SM58, this was quite remarkable.
A similar setup was used to record the vocals, again with the mic pointing towards a duvet-treated corner. Previously the vocals had been recorded in the control room, with all its resonance problems, using an SM58 and a pop shield. Actually the results had been pretty good, but Hugh and I felt that we could improve on the clarity and also lose some of the room coloration by using a capacitor mic (again plus pop shield) in the larger live room. A test recording using the C03 confirmed that the vocals did indeed seem more lively that way, and again all that was needed was some compression and a sympathetic reverb to make them sound as though they belonged on a serious record. No EQ was really necessary for either the vocal or sax recording made in this way, though a little air EQ is useful if the high end needs flattering.
Effects Advice
Charlotte's vocal and sax sounds had been suffering from being recorded using an SM58 in the less-than-ideal acoustic environment of the control room. Using a comparatively low-cost condenser mic in the live room, with drapes and duvets damping the corner behind the performer, immediately offered a much more usable sound in both cases — hardly any EQ was necessary once these measures had been taken, although a pop shield was needed in both cases.Tim Harding admitted that he had little experience adjusting effect and processor settings, and so tended to rely on the Yamaha AW16G's presets. In our experience, presets often don't give the best results, so we created a few custom patches and explained what we were doing as we went along. Calling up a vocal compression preset is fine, but compressors only do their thing when the input exceeds a threshold, so they behave differently depending on how loud you record the track in the first place. If, for example, a singer is consistently loud and close to peaking all the way through a song, the compressor will work nearly all the time, whereas if the track was recorded with the peaks at around -12dBFS or so, the compressor might never kick in, even though the compressor settings are exactly the same in both cases. If you must use a compressor preset, then at least adjust the threshold control while listening and watching the gain-reduction display.
Normally the compressor should show little or no gain reduction on the quieter phrases and around 8dB maximum on the louder phrases, though you may need to compress a little more than this if the singer has a very wide dynamic range. Ideally you should compress just enough to get the vocals to sit evenly in the track. I find that a ratio setting of 4:1 is enough for most singers and that an attack time of between 5ms and 30ms with a release of around 0.25s usually works well. I tend not to use any vocal EQ unless the recording obviously needs it, and as most voices are different it's difficult to see the point of vocal EQ presets. Having said that, you may find the ubiquitous 'air' EQ trick of applying a small amount of broadband boost at 12-14kHz helps to add clarity and presence to a vocal that needs it, especially if you are using a mic with a less-than-sparkling top end.
I also set up a couple of custom vocal reverb settings, plus a drum plate. These used the presets as a starting point, and my first vocal treatment was based around a vocal plate: I added brightness and reduced the bass multiplier value to below unity to get sizzle without muddiness. The reverb time can be anything from 1.8s to 2.5s, depending on what the song needs, and I used a pre-delay of 70-80ms to create a sense of space. The second vocal patch was based on the Live algorithm, with around 40ms of pre-delay and again some extra top end. This patch also sounded very neat on sax.
For drums I chose a bright plate, and combined a delay time of around 1.8s with a shorter pre-delay — any more than around 40ms can sound odd on drums, unless you specifically want a doubling effect, and as a rule the shorter the reverb time, the more reverb level you can add and still get a natural sound. In most instances, you only need to add reverb to the snare mic, with a little also on the overhead(s), just to help knit the sound together and to take the dry edge off the hi-hats.
My final foray into effects tweaking was to set up a very gentle compressor that could be used in the master stereo mix insert point, and also a 'smile' EQ setting that could add a little sizzle and depth to a finished mix. This is no substitute for professional mastering, but can help give finished tracks a more polished sound. If you're having a track professionally mastered, though, it's best not to apply any global effects of this kind to the mix. Our compressor setting used a ratio of 1.1:1, with a threshold setting of -40dB combined with fairly fast attack and release times to gently compress the whole of the mix — more conventional compression only treats the peaks, but it does so much more assertively.
Our 'smile' EQ simply combined gentle bass boost in the 80Hz region with a wide and shallow mid-range cut at around 250Hz plus 'air' EQ boost up at 12-14kHz. This should be subtle enough so that you are just aware of the extra clarity when the EQ is switched in. It's hard to put precise cut and boost figures on this, as an analogue EQ may achieve with one or two decibels of cut and boost what some digital EQs need 7-8dB to achieve. In the case of the AW16G, we needed 4-5dB of adjustment at the high end to get the requisite amount of air into the mix, while around 3dB of adjustment was generally enough in the mid-range and low areas.
A Bit Of Everything
Because there were so many separate issues to address, we felt that we hadn't gone into anything in great detail or depth, yet the improvements we'd made to both the monitoring environment and the miking of drums, vocals and sax were significant. The control room boom had been tamed to manageable proportions, the Auralex foam had tightened up the imaging by cutting down rear-wall reflections and the ad hoc speaker mounts and foam plugs had improved the evenness of the bass end to a worthwhile extent. I felt that we'd also demonstrated the benefits that even budget capacitor mics (combined with strategically placed duvets, of course!) can bring, and we even managed to throw in a bit of drum tuning and effects tuition!
Comments From The Hardings
"This was a great day where we all learned a lot. Thanks to Paul and Hugh's excellent DIY skills, the control room is already providing a much better sound environment for monitoring recordings — the difference made by a few relatively simple adjustments was very noticeable indeed! Our number one priority for upgrading will be to buy one or more condenser microphones and, as we're funding a whole band here, we initially need a microphone that is adaptable for use on voices, a range of acoustic instruments, and drum overheads, as well as being affordable. Something like the Samson CO3 or one of the other low-cost cardioid condenser mics will be top of the list, probably followed by a set of drum mics, as budget allows.
"We'd had no real previous experience of miking drums up at home, so it was great being shown how to get a convincing sound. We'd never realised that gaffer tape, masking tape and tissues would make such a big difference and Emily was particularly pleased with her tom sounds once Paul had doctored the kit. Using the clip-on drum mics was fun, and it was amazing how much clearer the drums sounded. However, even when the drum mics were packed away, Paul and Hugh still managed to get a very reasonable sound using the mics we already had, the best feature of which was the plate reverb effect, which gave the kit a really live feel.
"The biggest change to the kit though, was the front bass drum head being removed. This coupled with more gaffer tape, tissue, tuning, and an ID card taped to the head gave the bass drum a fantastic, punchy sound. This gave us the sound and feel the songs needed, without Emily having to wear her leg out!
"Although Emily might stick with the bubble wrap for live performances (she really does like the sound!), we'll definitely be using blankets for recording, and we have already altered a mic stand specifically for use in the kick so the mic can get right inside the shell. The overhead condenser mic gave a real sparkle to the cymbals, which gave the whole sound more fizz.
"Normally, we've recorded the sax parts in the small studio, just with an SM58 hovering over the bell. Moving the mic away from the bell just a little bit, and aiming it more at the body of the sax definitely picked up a much rounder, fuller tone. The condenser mic also helped to define and clarify the sound, while placing quilts and curtains on the wall improved the quality of recording no end. The obligatory reverb treatment enhanced the whole sound and Charlotte was extremely pleased with the end result!
"Again, Charlotte would normally have recorded her vocals in the smaller room with an SM58, but as we found with the sax, working in the big room against the absorbent backdrop brought what the mic was picking up into focus, and again the condenser gave that bit of extra clarity. Paul's EQ and dynamics settings were very helpful and I'm sure they'll be used a lot! One of the things we'd wondered about was whether to add reverb while recording or add it later. Charlotte's vocals were better when she'd sung with reverb in the headphones, so we were glad to discover that she could have reverb in the phones when recording, but not actually record the reverb. Adding it later means we'll be able to experiment with the different settings more.
"The advice on effects and processors was particularly useful. Tim felt that there was obviously much to be gained from experimentation, although he also felt that the key to this was confidence and, already since the visit, he's investigated various effects based on Paul's guidance and everyone's noticed the difference. Particularly satisfying was the stereo 'smile' EQ setting which added a bit of magic to the overall sound."