Hums, buzzes and noise were stopping Noor Ali recording his guitar parts, so SOS headed over to Worcestershire to set things straight.
Noor Ali is a guitar player with some history of recording in commercial studios, but he's since decided to buy himself a little multitrack setup and record his songs at home. He's a regular SOS reader and says it's helped him a lot with understanding recording techniques, but he called us because he was having problems recording his electric guitar. He felt that his system was producing too much hum, and, as so often turns out to be the case, the reality was that a number of problems were acting together.
Noor's studio is set up in one corner of a downstairs room in his Worcestershire home, and is based around a Zoom MRS1044 10-track recorder and a Joemeek VC3Q voice channel. Noor has a Joemeek JM27 capacitor mic and currently monitors through headphones, only checking the result on his hi-fi system after mixing to Minidisc.
The studio is home to a number of guitars, but most of our tests were made with a Gibson SG fitted with humbucking pickups. There are two guitar amplifiers, a 2 x12 Sound City combo and a smaller Laney 15W combo, both all-tube designs. Trying the Sound City first confirmed that it had a bad hum problem, even with no guitar connected, so I suggested that he get it serviced. There are numerous factors that can cause hum in valve amplifiers, including worn valves, incorrect output valve biasing and an imbalance in the heater voltage (most have a centre tap to ground). Furthermore, on an amplifier the age of this Sound City, any number of resistors could have drifted out of spec and time doesn't treat capacitors at all well either.
That meant we had a choice of recording via miking the Laney or DI'ing using the Zoom's built-in effects. We decided to check out the miking method first, and I was interested to see what miking arrangement he normally used. It turns out he'd been angling the mic downwards quite steeply and placing it close to the centre of the speaker, which had the effect of softening the tone slightly because the mic was being used off axis. I felt we might get more repeatable results if we went back to the more usual setup where the mic points directly at the speaker but is then moved outwards towards the edge of the cone if a warmer sound is needed.
After setting up the mic stand, we powered up the amplifier and plugged in the guitar. Sitting close to the amplifier gave the predictable hum problem, as the guitar pickups and wiring coupled into the hum field generated by the power transformer in the amplifier — even good humbucking pickups can only do so much — but by increasing the guitar/amp separation to around five or six feet, the situation was made better. The level of hum also depends on the angle of the guitar relative to the angle of the amplifier, so I rotated the amplifier to find a null and ended up with it being almost side-on to the guitar.
Once a good position had been arrived at, the level of hum and buzz was adequately low when the guitar strings were being touched, but became quite loud when the strings were released. This is often indicative of less than optimum screening within the guitar (around the controls and switches), but a practical workaround was to use a conductive wrist strap connected to the bridge of the guitar using a thin flexible wire and crocodile clip. Commercial variants of this idea can be purchased from UK electronics suppliers such as Maplin Electronics, where they are used by service engineers to prevent electrostatic build-up when working with electronic equipment that is sensitive to static electricity. Having proven the principle, we went on to do the rest of the recording with Noor making sure he kept one hand in contact with the strings or bridge at all times.
Aside from a little amplifier hiss, which wasn't loud enough to be a problem, we now had a working amp and guitar set up, though I wasn't at all happy with the overdrive sound the amp was giving — it was hard, unyielding and had a rather nasty barking quality that isn't normally a characteristic of Laney amplifiers. On discussing this, it turned out that Noor had changed all the valves in the amplifier (originally Russian valves that were easily driven into distortion) for some highly specified replacements, but I suspect that the biasing is less than optimum for these new valves, so the amplifier really needs to go in for a service. Noor had also upgraded to a different speaker, but that didn't seem to be giving any problems. The amplifier sounded quite well behaved when the gain was backed off to clean up the sound, so I suggested we use his Boss OS2 overdrive pedal in conjunction with the amplifier running fairly clean, and straightaway this produced a more comfortable 'British blues' sound.
At this stage, I asked Noor to make a test recording using his normal working methods, and on playback some unpleasant clipping distortion was immediately evident, even though he'd set a sensible recording level on the Zoom MRS1044. On checking the signal chain, it transpired that the clipping was taking place in the VC3Q preamp — the VC3Q has more headroom than many designs, but in this instance it was all being used up! The Joemeek mic is a sensitive small-diaphragm capacitor model, but neither it nor the VC3Q are fitted with pad switches, so the signal level entering the preamp was very high. I had to turn the input gain to its absolute minimum to get the level down to manageable proportions, and even then is was clipping the output stage on peaks because Noor had added a little EQ boost, and that was enough to push the signal past the limit.
One solution would have been to use EQ cut rather than boost to achieve the same tonality, but as I was starting from scratch I felt it would be better to zero the EQ settings, then use the mic position and amp controls to adjust the tone. Backing the mic off to around four inches from the speaker and moving it toward the edge of the speaker left us with a good sound and just enough headroom in the VC3Q to avoid clipping, even in situations where a little compression would be added. Noor had been using compression routinely, so I explained that, while compression can help even out the sound, every 1dB of gain reduction translates to 1dB increase in background noise when you're not playing. This exaggerates any hiss or hum present in the miked sound, so setting as little compression as you can get away with is safest.
On repeating the recording test, we were pleased to discover that what played back from the Zoom was very similar to what we were hearing from the amplifier. All the recording needed was a little of the Zoom's internal reverb while mixing, as the Laney amplifier didn't have a spring reverb. At this stage we felt we had established a valid method for recording the guitar that sounded right and didn't include an unacceptable amount of hum, buzz or hiss, though Noor demonstrated that if he touched any metal plugs connected to the Zoom recorder, or if he touched the metal base of the unit, there was a nasty buzz in the headphones. He explained that this caused a serious problem if he tried to DI the guitar via the Zoom's internal effects section rather than miking an amplifier, though he felt that even if the buzz were cured, he'd still record using the amplifier, as it gave a more authentic blues/rock sound.
A quick look over the system revealed the source of the problem. The only two mains-powered items were the Joemeek VC3Q and the Zoom MRS1044, both of which run from mains adaptors that don't carry the mains earth through to the units themselves. This lack of a hard earth can help avoid ground loops, because, in a typical setup, the monitor output of the Zoom would be connected to an earthed monitor power amplifier, which would provide a central ground connection for the whole system. However, as Noor was using only headphones, there was no earth connection anywhere in the system, which meant the metalwork of the VC3Q and Zoom MRS1044 were effectively 'floating'. This was confirmed by using a piece of wire to make a temporary connection between the metalwork of a jack connected to the Zoom MRS1044 and the mains earth — the buzz problem disappeared instantly. For a more permanent solution, I suggested the option of fixing a tag washer to one of the screws in the base of the MRS1044 and earthing that, but pointed out this wouldn't be necessary if a monitor amp and speakers were added. Noor had already realised that the headphones didn't provide an accurate impression of bass, so was considering buying a monitor amplifier and a small pair of reference speakers.
As is so often the case, what started out as an apparent single problem turned out to be a combination of factors, but, once these were addressed in a logical order, it didn't take long to identify them and come up with solutions. As with any recording, a good result starts out with a good sound at source, and around half our effort was focused on achieving that. Had the amplifier been working properly, that particular job would have been a little easier, but amp position, mic position and quality of guitar screening all played their part. On the recording side, it was only when we'd sorted out the system grounding and unusual gain structure that we started getting really good results. However, the whole process took less than an hour and made a world of difference to the recorded result.