Studio SOS: Grahame McGrath

The SOS team visit a well‑appointed home studio to try to cure a strange noise problem and help out with some acoustic guitar recording.

Grahame McGrath runs a small studio called Babydragon, in Worcester. He called us to ask for help with an irritating problem he had started to encounter while recording an acoustic guitar a few months earlier. For the most part, the sound was fine, but when the player pulled several strings simultaneously, there was a low‑level sound following the notes that sounded almost like a buzzing loudspeaker cone. It occurred both when miking and DI'ing, and on three different guitars. This sounded like an interesting challenge, and as Grahame's studio was just down the road from us, we decided to take a look. At least it would make a change from sticking foam onto walls!

When we arrived at Grahame's bungalow, we discovered that he'd converted one of the larger rooms into a small control room and an adjoining live area. He'd clearly done his homework, and had included generous amounts of bass trapping on the walls and ceiling to help make up for the small room sizes: the room was finished to very high standards and sounded pretty good. His Mackie HR824 monitors were set up on substantial pieces of granite kitchen worktop above rack enclosures on either side of his mixing console. The installation made the most of the limited space, although it did mean Grahame had to sit a bit inside the normal equilateral monitoring triangle, as the monitors were a little further apart than would be ideal.

Grahame hadn't skimped on equipment, either: four channels of his Soundscape recording system were fed from a pair of Lucid A‑D converters, with a Lucid master clock keeping all the digital gear in the studio sync'ed up and a Lucid D‑A converter in the monitoring chain. At the time of our visit, Grahame was using a Soundcraft Spirit 328 digital mixer. This was mainly for monitoring, but he has plans to replace it with an analogue mixer of his own design incorporating passive EQ, a few sends and the usual monitoring requirements.

What's That Sound?

Try as they might, the SOS team couldn't reproduce the strange noise that Grahame had heard on some of his recordings.

Apparently, the mystery noise on the guitar recordings showed up most when using Grahame's Focusrite ISA 428 mic preamps, but was also present to some extent regardless of which of his preamps he used. Although he had erased most of the 'faulty' recordings where the problem was evident, Grahame did manage to find a couple of examples, and although the noise was at a pretty low level, it was audible. From the sound of it, I was also able to rule out fret buzz as a likely cause. I'd describe it more as a very low‑level electrical 'splutter' at the end of the note decay.

We thought the best way to deal with this would be to get Grahame to recreate the exact circumstances under which the problem first occurred, so he enlisted the help of Gary, the guitarist he was working with when the issue arose. Gary also indulges in the odd spot of home recording, so he was just as intrigued as we were about the cause. He was also interested in the best ways to mic the acoustic guitar, so we suggested that once we'd located the source of the noise problem, we'd do a few miking experiments to demonstrate how apparently subtle factors can make a big difference to the sound.

Grahame had discovered that, with this particular guitar, he got cleaner results with the mic around four to five feet from the instrument. This is around twice the 'usual' distance but, possibly because the studio space was carpeted and very well treated (and thus a relatively dead space), it worked well. He'd used his AKG C414B XLS microphone for the original recording, so that's what we set up. Gary played the same song as before and we listened. Wouldn't you know it, the recording came out perfectly clean, with no sign of the mystery noise at all! We tried switching to a TL Audio Ivory preamp too, but the background noise didn't appear, and then we switched to a DI recording, again with no noise...

Obviously, this process didn't help us identify the source of the noise, but we did discover that the DI result sounded better with the guitar's preamp gain around three quarters up, rather than on full volume, as some distortion was evident during louder passages at maximum gain. We suspected that the guitar preamp was running out of headroom at high gain settings, possibly because of failing batteries.

Probable Cause

While Gary played guitar, Paul experimented with different microphone positions, monitoring the results on a pair of Beyerdynamic headphones.

Since the alleged noise was clearly an intermittent problem — evident on the previous session's recording but not easily repeatable during our visit — Hugh and I both suspected that the noise source might either have been some form of interference, or some dodgy signal connections. Given the time of year, it could well have been interference from a light‑up Santa down the road or some other temporary electrical installation somewhere nearby, but perhaps a more likely cause was Grahame's system of wall‑box and patchbay connectors.

When oxide forms on patchbay contacts, it can result in a degree of semiconductor action, which manifests itself as a non‑linear resistance (a resistance that differs between high and low signal levels). As the signal level falls, this non‑linearity can cause low‑level distortion of exactly the kind we heard on Grahame's example recording. An easy solution is to simply plug and unplug connectors into and out of sockets a few times. This can have the effect of rubbing away the oxide layer and can restore normal operation, which is what appears to have happened here — though a much better option is to treat the plugs and sockets with a contact cleaner such as Caig's DeOxit every few months, as that really helps avoid such problems. Grahame had a can of this on his shelf already, and said he'd give his system the once‑over after we'd gone.

In Position

Not wanting this to be a wasted visit, we decided to try to improve the recorded guitar sound, which, although not at all bad, did sound a little woolly at the low end for my taste. We routed the microphone signal, via the mic preamp, to some closed‑backed headphones in the studio, and I then moved the mic around relative to the guitar, to find the best 'sweet spot'. However, this turned out to be less easy than expected, as Grahame's Beyerdynamic DT150 studio headphones produced rather more bass end than his studio speakers, so whenever I adjusted the mic position to get what I felt was a good tonal balance in the headphones, the result heard in the control room was decidedly bass light. In this instance, I trusted the speakers more than the headphones (Grahame had them set for full‑space operating, which gives the maximum low‑end output), so once aware of the problem, I tried compensating mentally when positioning the mic, and checking the result on the speakers. We also double‑checked the sound on Grahame's diminutive Audioengine 2 speakers — tiny powered monitors with three‑inch drivers, which sounded surprisingly impressive.

On Reflection

The headphones and Mackie monitors were presenting significantly different results, so some double‑checking was needed in order to find the best microphone placement.

My first instinct was to move the mic closer, but in this small room that really emphasised the comb‑filtering effect that you get when the direct sound from the guitar combines with the reflected sound from the walls and floor. I got Gary to wear the headphones while I moved the mic, and he was genuinely surprised by the huge tonal differences that resulted. Even changing the height of the mic from the floor by a just few inches resulted in a dramatic change in the sound. Of course, comb filtering occurs in any room, other than an anechoic chamber, but I'd never heard it have such a strong effect before: clearly this is the reason why Grahame had arrived at the mic position he had, as it produced less obvious coloration than when working closer.

It seemed to me that the strongest reflection should come from the floor, and as this was covered in a thin carpet, the reflected sound would be attenuated at the high end. This was making the comb‑filtering effects more pronounced in the mid‑range frequencies, and therefore more obvious.

I always prefer recording acoustic guitars over a hard floor, so I asked Grahame if he could rustle up some tea trays and boards to place on the floor between the guitar and the microphone. This he did, and the result was a definite brightening in tonality. A further test proved that the floor wasn't the only significant source of reflected sound, however: when we changed the position of the guitar and mic within the room, keeping the same distance between the mic and guitar as before, the tonality changed again, with some places in the room sounding noticeably better than others. This is a useful lesson because we do tend to think of mic placement as being mainly to do with the relative position of the mic and the instrument, but where these are placed relative to the room boundaries also has a large effect. And in such a small studio as this one, you're never far from a room boundary!

Taking The Mic

The next test was to try some of the other mics Grahame had at his disposal: his Audio Technica AT3035 and a Neumann TLM103 seemed the most likely candidates. The AT3035 produced a decent result, but lacked the slightly inflated brightness of the C414B XLS, and so didn't sound as detailed or as lively. Putting up the TLM103 combined warmth with a very lively top end and, equally importantly, it seemed to play down the soundhole resonance, so the overall effect was of a much sweeter tonal balance. I think that result rather surprised Grahame, as he'd been using the TLM103 mainly as a vocal mic.

As a final test, we hooked up a Rode NT5 small‑diaphragm mic — which is significantly cheaper than any of the others we'd tried so far. The result was not dissimilar from what we heard with the TLM103, although it lacked the brightness, having a flatter frequency response. However, a little gentle shelving EQ at around 10kHz would soon add the necessary sparkle, so this was also a viable option.

Learning Experience

While Grahame is already pretty experienced at recording and understands the underlying theory, I think what guitarist Gary took away from this session is that, when it comes to recorded tone, finding the optimal mic position and the best part of the room in which to work is far more important than the actual choice of microphone. Even lowering the mic by just a few inches produced a far larger tonal variation than the difference between any of the mics we tried.

At the end of the afternoon, we headed for home expecting Grahame to call any moment to say the mystery sound had come back — that would at least confirm my theory that Hugh had simply scared the gremlin away! However — touch wood — so far, so good.