This month, we solve amplification problems for a tea-chest bass.
Chris Walker is based in Cardiff, within walking distance of Torchwood's secret underground headquarters, but his problem was very much down-to-earth. He has a home studio in a spare downstairs room: a simple setup, based around Cubase, running on a specially-built quiet PC, and monitored via a pair of the new model Alesis Monitor 1 active speakers, mounted on wall brackets. The audio interface was a Line 6 TonePort, which included the Gearbox amp- and effect-modelling software.
The main challenge was to sort out an amplification and recording solution for Chris' tea-chest bass. We decided that this would be an interesting challenge as the tea-chest bass isn't something you come across every day, yet, when played correctly it is actually a very solid bass instrument with lots of character.
Chris plays the bass in a band that's been playing together for over 20 years. They get invited to some very prestigious venues, but Chris has always experienced problems with amplifying the instrument, both for live performance and for studio recordings. He's tried various mics in different positions but these tended to result in unacceptable amounts of spill from the other musicians, and on stage acoustic feedback has also been a major problem.
Chris told us: "I've been playing tea-chest bass with my group Railroad Bill for over 20 years now. The full line-up is myself on tea-chest bass, Dan Nichols on washboard, Geoff Coates on rhythm acoustic guitar, Geoff Haynes on lead guitar, and Andy Baillie on mandolin, ukulele, kazoo, occasional musical saw, and so on. We were once described as 'like Lonnie Donegan on Speed' and we've played our souped-up versions of country blues, gospel, jazz and western swing to audiences all over the world.
"The tea chest produces a surprisingly big sound that's a bit like a double bass. There's not much sustain, but it is excellent for chugging railroad, rockabilly and jug-band style music. It's fine in small acoustic gigs or on the street, and is particularly good on a wooden stage or other hard surface with a back wall, but it's always proves a headache to mic up for recording or on larger stages. Because I play it entirely by ear, I need to be able to hear it loud and clear on stage, and this is where the problems start, as it can swamp the monitors and cause horrible low-frequency feedback. We've had good results from time to time, but it usually provokes head-scratching amongst sound engineers."
Chris's challenge to the SOS team was to help him find a pickup or miking solution that would allow him to run the bass through a bass amp or desk, and have it sound the same as it does acoustically, just louder. Chris: "This would give me control over my on-stage monitoring and would also help to isolate the sound for recording. I also wanted some ideas of how best to mix it to capture the 'true' live sound better, as recent recordings have been a little muddy.
"I've already tried a PZM mic under the box, which worked to a degree, but was less than ideal on stage. I also tried this through a bass amp, but with terrible results. I've tried a conventional microphone pointed at where the string meets the soundboard, but this still requiresa lot of fiddling to get the sound right, particularly when it comes to monitoring. I've also tried putting the mic under the box, and at one time I tried a cheap bass transducer which was difficult to position correctly and sounded very bad."
With Chris's musical and technical colleague Geoff Coates, we discussed the problem over tea and Hob Nobs. My first thought was that we should try a contact pickup of the type used on some acoustic guitars. We could move the pickup around to achieve the best sound, and it should offer good isolation from spill. I had my own Seymour Duncan acoustic pickup system in my bag of bits, as I had anticipated that it might come in handy. Sadly, it seems that this particular model is no longer available, but we were nonetheless able to use it to test the principle of using a contact mic: if we were successful, Chris would need to buy a currently available system (and good ones can be quite expensive!).
The instrument itself is simply an upturned tea-chest, with a broom handle fixed vertically to one corner by means of jubilee clips. A nylon cord then runs from the top of the broom handle and is fixed through a hole in the centre of the top panel of the tea-chest, by means of a large knot inside the box. Chris wears leather finger protectors on his left hand, which he uses to control the string tension, and plays the instrument with uncanny accuracy. The general sound is not unlike that of an 'economy' double bass. He'd cut a hole, or port, around two inches in diameter in the top panel near one corner, with the aim of miking at that point, but his experiments had been very prone to feedback.
As a first test, I fixed the piezo pickup part of the Seymour Duncan system to the top of the tea-chest using double-sided carpet tape. The pickup connects to a small, battery-powered preamp that would normally be fitted inside the guitar body. We fed the output from the preamplifier into a small bass combo amplifier and immediately heard very encouraging results. By moving the mic around, we found there were huge tonal variations to be had, but the most natural sound was obtained by putting the pickup close to the point where the string entered the top panel of the tea-chest. This could have been the perfect solution, as there were no significant spill or impedance-matching problems. However, given the price of good pickups, we decided to explore some DIY alternatives.
I'd noticed a Maplin Electronics store on the way in to Cardiff, so I suggested we get hold of one of their little one-inch piezo sounder discs and try using that as a contact pickup (I've used these before to make drum trigger pads). Geoff said he'd already bought one for an as-yet unrealised project and suggested we use that. They cost less than a pound each so it was no great sacrifice. These piezo discs function best when running into a high-impedance input, and though they will work into a typical guitar or bass amp with no intervening electronics, a preamp designed specifically for piezo pickups can give better results. For recording, a high input-impedance active DI box or an interface with an instrument input would give better results than plugging directly into a line input, as the typical line-input impedance is low compared to that of the pickup and so would load it unnecessarily.
Again, we found that moving the piezo disc around gave very different tonalities, and we were able to get a very decent sound from plugging it directly into the bass combo. Again, we used double-sided carpet tape to hold it in place. It also worked adequately into the Toneport's Instrument input, which has the same nominal impedance as a guitar amplifier. The susceptibility to acoustic feedback depended on the position and subsequent bass output of the pickup, but we felt that it would be usable under most normal conditions. It didn't sound as good as the Seymour Duncan system but it was impressive nonetheless, and again there was no noticeable spill from external sounds.
We still felt, though, that a bit more depth to the low end would be nice, so I happened to mention that we might be able to use a small loudspeaker as a microphone if anyone happened to have one. The ever-resourceful Geoff produced a small (5-inch x 2-inch) elliptical speaker that he'd picked up somewhere for 50p, so we soldered on a couple of wires to hook it up to our jack cable, then plugged it directly into the bass combo. The advantage of using a loudspeaker as a microphone is that it has a relatively low output impedance and is capable of driving both instrument amps and line inputs without additional circuitry.
With the speaker plugged into the bass combo, we moved the speaker around, face down, over the top surface of the tea-chest and found that the sound was very bass light, unless we placed part of the speaker over the hole Chris had cut in the top panel during his miking experiments. In fact, by overlapping the speaker with the hole by only a quarter of an inch or so, we got a tonality that reflected that of the instrument played acoustically, and there was plenty of low end to work with. Chris liked the retro look of a speaker fixed to the top of his bass (and he also liked the budget!), so we took time out to walk down to his local music store to buy a jack socket that we could fit to the side of the tea-chest to create a more permanent installation. We fixed the speaker in place using four M5 nuts and bolts with washers, then wired the speaker to the socket using a short length of co-ax cable. We also connected the speaker chassis to ground via the cable screen, and fixed the cable using a self-adhesive pad and cable tie, to prevent it vibrating against any of the surfaces.
Subsequent tests via the amp showed that we were getting a much better sound than Chris has managed before, and with reasonable resistance to feedback — though of course any acoustic instrument will 'take off' if you crank the gain high enough. For live use, I suggested that Chris look at the Behringer DSP Shark unit, as this combines a mono anti-feedback device with a mic amp and costs around the same as an average overdrive pedal. During setup, you turn up the gain until feedback starts, with the Shark in learn mode, then the unit automatically puts narrow notch filters in the signal path which correspond to the detected feedback frequencies.
The speaker worked fine going into the Toneport's Instrument input, and also produced a sensible recording level when Chris was playing normally. We made a couple of test recordings and found that by using small pieces of Gaffa tape to cover the gap between the edge of the speaker and the hole in the tea-chest top panel, we could reduce the amount of bass end in a very controllable way. Another alternative would have been to fit the speaker in slotted holes, so that it could be moved closer to or further from the hole. However, Chris liked the deep, punchy sound we were getting, which contrasted well with the somewhat dull sound he'd been getting earlier, and the sound sat well in the mix he'd been working on. The speaker does pick up more spill than the contact-mic approach, as it is essentially acting as a microphone, but the level of spill seemed acceptable.
Having found a couple of viable solutions to the tea-chest problem, we turned our attention to the recording system, which was set up on a small desk close to one side of the room. This placed a reflective plaster wall immediately to Chris's right, which bounced a lot of sound back to the monitoring position, compromising the stereo imaging. Chris had already bought a couple of foam chair mattresses from Ikea to hang on the walls in a corner for vocal recording, so we moved one of these to the right-hand side wall at head height just to see if it would improve matters (see photo opposite) and, sure enough, the stereo imaging was cleaner and better defined. It wasn't practical to set up the system to achieve symmetry, but it worked well enough, with no obvious problems. The monitors are small enough not to provoke any serious low-end room problems, and they were set up in the correct position relative to the mixing chair.
Chris had been struggling a bit, trying to find good EQ and compression settings to work on the bass, so we decided to see if we could improve on what he had. Using the compressor in Cubase, it was possible to firm up the sound slightly by using the bass guitar preset Chris had chosen, then adjusting the threshold to achieve four or five dB of gain reduction on the loudest notes. Compression also brings up any spill that may be present, so it makes sense to use as little as you can get away with. Compressor presets can be useful but you will always need to adjust the Threshold control, as the amount of compression needed will vary depending on the level and dynamics of the signal being processed.
Regarding EQ, the bass component of the instrument can be controlled quite effectively by cutting or boosting at 70 to 90Hz using a parametric equaliser, though it may be wise also to use a steep shelving EQ below 50Hz, to cut out the really low frequencies that eat up headroom for no benefit. If bass boost is required, it can help to keep the sound clean if you apply a degree of cut an octave or so above the point where you applied the bass boost, which worked out at around 160Hz in this case. This prevents the effect of the bass boost spilling over into the mid-range and muddying the sound. You can often add a bit of bite to the sound by boosting at three to four kHz, but we found there was little going on up at these frequencies that we could boost, so we suggested that an exciter-type plug-in would be useful if more definition was required.
Chris played us some of his original recordings, which seemed to lack low-end punch, so we experimented using the sub-octave 'stomp box' in the Line 6 Gearbox software. We found that we could use it sparingly to add some subtle low end that helped bolster up the sound. However, once we'd fitted the speaker 'pickup', we found that our new recordings had all the low end we could wish for.
Another useful tactic is to use a 'fingerprint' EQ (such as Assimilator for the Powercore platform), which can analyse a sample of a sound (the target) that you'd like to achieve, such as a passage of solo double bass, plus a sample of what your instrument actually sounds like (the source). The plug-in calculates an EQ curve to make the source sound match the spectrum of the target sound. Again, these aren't perfect, but used with care they can get you a lot closer to the sound you want. Chris didn't have a fingerprint EQ plug-in but after we'd explained the concept he said he would consider getting one.
Though Chris plays the instrument surprisingly accurately, pitching can be tightened up further using Auto-Tune or similar pitch-correction technology. This can work well, providing you don't set the correction rate too fast. Using a slower correction rate, the natural slides of the instrument pass through as normal but sustained notes are gently nudged into pitch. I've used this technique on double-bass parts before now, and it seems to work just as well on the tea-chest bass.
"Despite (or maybe because of) the South Wales Chocolate Hob Nob drought, Paul and Hugh worked fast and within five minutes had a good sound from the Seymour Duncan pickup with preamp, but it seemed a large financial outlay for such a simple instrument. Geoff's 50p piezo from Maplin was a revelation: it sounded much better than I would have expected for something so cheap. However, though it may have sounded better through an acoustic pickup preamp, it was still not as clean or clear as the Seymour Duncan.
"I was really surprised that Geoff's old speaker produced such a tremendously accurate and full sound — precisely what I was looking for, and it also had that home-made retro look of something that could have been knocked up in the 1950s! The fact that the entire installation cost about a fiver was a bonus too, and it would have been even cheaper if I hadn't chosen a high-quality gold-plated jack socket! My next task is to build some sort of housing to cover the speaker, protect it from damage and conceal its inner workings from the curious. And maybe add a knob that goes up to 11, and some space-age fins...
"Paul's suggestions for mixing were also very useful. The addition of a little sub-harmonic activity beefed the sound up, and I had a very useful lesson in EQ'ing, where I picked up some great tips. One of Paul's suggestions was to use a fingerprint EQ to mould the sound to that of a double bass, so I'll do some research on the web to find an affordable plug-in. His idea of using Auto-Tune to tidy up the intonation on recordings sounds interesting too.
"I'm now looking forward to our next stage gig where I can just take a bass amp and let the engineer run a DI from it. You can catch us and the new Electric Tea Chest this year at Glastonbury Bandstand, Trowbridge Pump festival and Goodwood Revival, amongst others!"