Recording an electric guitar may be done by simply plugging it directly into the mixing desk, maybe with one or more effects units - Rockman? - in between. To me, this is by far the least exciting guitar sound. I've been copying and analysing guitar sounds for several months, working to get a lead guitar sound out of my Mini-Moog (those of you, who've heard Jan Hammer's album Black Sheep or the Tom Coster solo stuff will know what I mean). It was quite fascinating to see how much the sound relies on the properties of the guitar, pick-up, cable, amplifier, and loudspeaker. Each of these components adds 'life' to the sound through 'irregularities' in its frequency response and through distortion (pick-up, amp, speaker). It is impossible to cut out one of them (eg. the speaker, by using the amp's line output) without deteriorating from the overall sound.

A guitar amp is a good example of how distortion is perceived. Many times, small amounts of distortion are perceived as sound 'colouration' rather than downright destruction of the sound. This also occurs with microphones, so a few basic words about distortion will be required later.

But first, let's look more closely at the chain of transducers that transforms the vibrations of metal strings into rock'n'roll's most distinctive sound.

Pick-up & Cable

The guitar's pick-up 'transduces' the string vibrations into an electrical AC current. The electrical properties of the pick-up, amplifier input stage, and guitar cord form a resonance circuit with a resonance peak between 1-5 kHz and 5 kHz, depending on the individual specs. Beyond this point, the pick-up's response curve falls off at a rate of 12dB/octave. This resonance effectively produces the guitar's power, as it falls within the frequency range to which the ear is most sensitive.

For years, tube (valve) amplifiers have been used for guitar amplification and, in this particular field, survived the introduction of transistor technology. It is common knowledge that valves 'overload gently', gradually adding distortion to the sound when overdriven. Transistor amps will also introduce large amounts of distortion just above the clipping level, but their distortion sounds altogether different.

There are various types of distortion: Harmonic Distortion (THD), Difference-Frequency Intermodulation (DFIM) and Transient Intermodulation (TIM). Because it is the easiest to measure, THD only is usually quoted in spec sheets. One should emphasise that THD doesn't tell the whole distortion story! What is even more interesting is that the same amount of THD may be perceived as a totally different degree of distortion! Here is the explanation why:

If a sine wave of a frequency 'f' is distorted, harmonics at 2f, 3f, 4f, 5f,... etc are produced, eg. a frequency of 1 kHz will create sine waves at 2kHz, 3kHz, 4kHz,... Now which particular harmonics are created at what amplitude (volume level) depends on the type of distorting unit. Tube amplifiers tend to produce even-numbered harmonics (2f, 4f, 6f...), whilst transistor gear will produce only the odd-numbered harmonics (3f, 5f, 7f...).

Odd and even-numbered harmonics do sound different. Odd-numbered harmonics are audible even below 1 % THD (hardly measurable with medium-priced instrumentation!), while even-numbered harmonics seem to be definitely less annoying. Psychoacoustical tests showed that THD-figures of a few percent were perceived as undistorted sound. (Interestingly enough, musicians tolerated even greater amounts of THD than sound engineers: up to 10%!).

Microphone Distortion

Microphones, when confronted with a very loud signal, tend to overload. When close-miking drums, brass, amps, or vocalists in a rock/pop context, tremendous SPL (sound pressure level) must be accepted by the mic, sometimes far above the physical threshold of pain. Figure 2 lists SPLs delivered by some instruments at typical microphone working distances. Of course, moving further away from the instrument lowers the SPL sensed at the mic's location considerably, following the inverse square law ie. doubled distance equals -6dB SPL.

The maximum SPL threshold for a certain THD value (0-5% is specified by the 'IEC standard) published by microphone manufacturers is of little practical use. It is only valid for a 1 kHz sine wave (except for AKG's C460). But the problem range (with condensers) is the bass. Condensers generally produce pretty ugly distortion and should be used with care in high SPL situations.

Distortion thresholds for dynamic microphone types are seldom given. These mics distort in a gentle way, introducing distortion gradually as the SPL increases and usually the mic's distortion will not be audible. A trumpet played fortissimo (loud), a guitar amp squeezing 100 watts into a pair of 12-inch speakers or a vocalist producing 140 dB SPL won't deliver 'clean' sounds anyway, but the sound colouration due to distortion is quite noticeable and it differs a lot from one particular type of dynamic mic to the other. One should always keep in mind that this is an important quality of a mic's sound that won't show up in any spec sheet, response curve, or whatever. One more reason to rely on your ears when choosing a microphone.

Speaker

Typical guitar amp speakers are built differently from PA speakers. They emphasise frequencies at about 4kHz, and don't deliver much sound above 10kHz. When miking up a speaker, one important factor to consider is its projection pattern. Full-range speakers (as all guitar amp speakers are) radiate treble frequencies in a narrow beam following an acoustical rule: as the wavelength of a signal becomes several times smaller than the sound source (the speaker's diameter), the radiation angle narrows. So, if you want to catch all the treble, aim the mic at the centre of the speaker cone.

Miking The Amp

Place one mic in front of the speaker, at a distance of maybe 1 foot and aim it at the centre of the speaker cone. Since the frequency range delivered by the guitar-amp combination is fairly limited (80Hz to not much more than 6kHz), a cheaper dynamic mic with a modest frequency range will do the job. Vocal mics such as a Shure SM58 are also often used.

If you want to beef up a rather bright, metallic guitar sound (eg. a Stratocaster with a Fender Twin amp) you might use the AKG D12 (Tasco did so to thicken the guitar sound of Eric Clapton-sideman, Albert Lee). Other mics often used in the studio and onstage for this purpose include Sennheiser MD421, Shure SM57 and the AKG D125, which has a certain punch to it having been developed for percussion and amplifier pick-up.

Next step: pick up the room sound (ambience). Use a second mic, which may be a condenser, for this purpose and try the following place positions:

1) In line with the speaker cabinet's main radiation direction, but further away, to the back of the room.

2) Rear-open enclosures (eg. Vox AC 30) may be miked up from the back. For small distances use a dynamic mic and reverse the phase of the signal on this mic.

3) Sealed enclosures (eg. Marshall cabinet): mike the cabinet from the side, to get the sound radiated from the wood.

4) If you have enough spare channels, put 3 or 4 mics around the guitar amp cabinet and combine their sound. Experiment with positions and phase reversal.

An interesting sound can be achieved by additionally miking up the electric guitar's acoustical sound (for rhythm guitar work). This is especially useful with semi-acoustic guitars.

Phase Reversal

If you don't have a phase reversal facility on your mixing desk, you may want to construct one. The easiest approach is to build a phase reversing cable (which must be marked, please), and for comparing the in-phase and out-of-phase sound you'll need a switch. It will help you not only in recording, but also to fast-check your mics, cables and speakers for proper wiring. Figure 4 illustrates the principle, but ready-made adaptors with built-in phase switches can be bought as an alternative, from good audio dealers.