There are lies, damned lies and specifications, to paraphrase an old BBC2, Sunday afternoon, black and white film. In part two of X-Ray Specs, Ben Duncan concludes his layman's definitions of the terms that try us in the age of manuals.

MOTHERBOARD

A main circuit board used as a chassis or template, into which any subsidiary boards (cards) or modules may be plugged. In a complex FX unit this sort of assembly can greatly speed servicing time and thereby save on labour costs. No longer hip.

mSec

Milliseconds = thousandths of a second; Milli, a 1/1000th.

0V

Stands for 'nought volts', pronounced "Oh-vee", this is synonymous with the signal's earth, ground or reference point, often the deck or chassis of the equipment or the screen connection on a lead.

OP-AMP

An analogue chip. Short for Operational Amplifier, the original chips were designed ca.1961 for analogue computing. Digital technology soon made analogue computers obsolete but the op-amp lived on. Today they provide the fastest, most accurate and compact means of designing and packaging the complex signal processing inside consoles and outboard gear. A typical console uses upwards of 100 op-amps each equivalent to a 10 transistor circuit which would otherwise occupy ten times the real estate on a circuit card. Result? A 48-channel console needn't be the size of a baseball stadium. Op-amps are used as amplifiers, filters and impedance convertors.

PAD

A fixed attenuator, usually switchable which cuts down the signal by a known amount. Prosaic reason: to help mixer front-ends cope with Elephant herds as well as Butterflies. Poetic reason: the pad is like the storm shutters on an Arkansas farmhouse.

PAN POT

Short for panoramic control potentiometer, it essentially consists of a stereo volume control, cross-wired so the left and right outputs fade in opposite directions to produce an image shift between the speakers.

PARAMETER

Buzz-word for a dimension, quality or facet of a unit, preferably one that can be delivered in figures for greatest specmanship on behalf of the manufacturer.

PHASE SHIFT

Something which happens in ye time domain, phase is difficult for ordinary mortals to comprehend not least because it's often confused with polarity (see below). If you look at a phase shift on a 'scope (one of those funny TV screens), it's easy to comprehend the shift in timing that phase differences entail. Essentially, if phase is shifted uniformly, at all audible frequencies, the sound arrives later or earlier, but later or earlier than what? As there's no referral point, the ear can sense no difference, but more commonly in equipment, phase-shifts are bigger at some frequencies than at others. This has the result of putting the lows and highs out-of-sync, though unless the phase shift is made purposely large, it's only usually in the order of a few hundred uS (microseconds) or mS (thousandths of a second), making the effect subtle. But the FX of much bigger phase shifts can be gauged by making a few transatlantic phone calls — that weird, drunken sounding side-echo is the sound of your voice, phase-shifted through thousands of degrees. Complexity and confusion enters when it's sussed that phase shifts of 90°, 180°, 270° and 360° have magical qualities. For example, 360° completes a twist of a sinusoidal sound wave (it helps to think of it as a spiral, unfurling in 3D), and places the phase-shifted signal back in apparent synchronisation; but that's only for a continuous signal; a musical signal is generally reckoned to be unique at each instant. Therefore a low frequency signal which takes several mS to cycle through each full 360° is the one most likely to sound wrong when phase shifted by 45° or more (that's a significant fraction of 360°). This is but one reason why our ears are most sensitive to phase funnies at low frequencies. Verdict: people vary in their sensitivity to phase-related FX, so don't panic.

POLARITY

For DC sources, like batteries, this is about not getting the + and - a-about-face. Audio connections also have polarity though a better word would be symmetry. There's no absolute polarity of course — an audio signal is forever alternating between positive and minus excursions — but when a group of signals in a chain are compared, and if we take the original impulse at the mike as our reference, then there's a relative polarity at large. A practical result of inverting polarity is that the kickdrum 'sucks' instead of 'blows' on the leading edge, ie: the speaker cone moves backwards instead of forwards. If the kick drum is back in the mix the loss is subtle, but on hot percussion recordings, reversing all the leads on your speakers could result in a more incisive sound.

Also, for multi-mike set ups, it's no good if out of two adjacent mikes, one is saying 'suck', and the other 'blow'. This can happen if one mike is wired back-to-front or it's going through a processor which inverts the polarity.

To this end, posh consoles have a phase reversal switch on every channel. This has nothing whatsoever to do with phase — it's actually a polarity switch, and acts like swopping over the mike or speaker leads.

POT

Col. for potentiometer which in turn derives from words chosen to suggest a voltage measuring device. Strictly, it divides or apportions the voltage applied across it. Thus, when we turn down a volume control we're in fact adjusting the pot (behind the panel) to tap off a smaller percentage of the total signal voltage conceptually available. A superior choice of terminology would therefore be a tap.

PPM

Peak Program Meter. Strictly, a rapid-response 'swing-needle' meter developed by the BBC. The needle reaches the peak value of the signal fast enough to beat the ear but falls back slowly so the peak reading is visible in the ½ second or so it takes the eyes to wake up and send an "oops!" signal to the brain. Oddly, the official BBC PPM is scaled 0 to 7 ("Se'ven is an'tah'lay ahd'iquat'"). PPM is also the name for LED bar-meters which read peak levels even if they're not scaled to the BBC spec in dBu or dBV (which are arguably more sensible units).

QUAD

An abbreviation for four, as in:
i) Quadraphonics — four-channel sound, now obsolete.
ii) An IC package containing four chips (to save space).
iii) Short for Quadrangle (Vide 'Jennings & the mystery spider'), and,
iv) Nothing to do with four whatsoever.

The famous Quad amplifier derives its name from one of its predecessors. In 1947 The Acoustical Manufacturing Co. launched one of the first postwar hi-fi amps, and designer Pete Walker dreamt up a hideous acronymn, 'Quality Unit Amplifier Domestic' which quickly reverted to 'Quad'.

RF

Radio-frequency. Generally applied to any interfering old garbage above audible frequencies. Common RF nasties are CB on 27MHz (Megahertz), that's a frequency 1500 times higher than the highest crash-cymbal harmonics. Proximate TV stations, thunderstorms, fluorescent lights, machinery, police cars and nuclear explosions are other potential causes of RF aggravation. In general, there's least RF hash about in the countryside, but hark: One studio was beleaguered by a weird 'tititititititit' sound on the desk every time a veteran Lancaster Bomber flew over from the local air base. This particular black hulk was built in 1944 and therefore didn't qualify for prosecution under the 1955 act which forbade petrol engines without adequate RF interference suppression.

SID

An acronym coined by Walt Jung, a US audio guru, in the mid-70's. Standing for Slewing Induced Distortion, it's a crunch-up that arises at high frequencies only. In fact, it happens only when the signal exceeds the equipment's slew rate limit, a parameter giving in v/μS ('volts per micro-second'). A handy analogy is going round a bend, damned fast: a Capri has a poor slew limit (ie. loses control easily) and high speeds and/or tight corners will cause a (literal) crunch-up! For electronics, though, it's a combination of frequency and voltage that kills. Assuming frequency is limited to 20kHz, a slew-rate figure guaranteed to keep us out of trouble then depends on the output voltage. Therefore, what's a cool figure for a console (say 10v/μS) becomes a poor figure when applied to an amp. Rule of thumb: allow 1 v/μS for every volt of output. Symptoms: lassitude, nausea. A case of bad SID makes the top end sound gritty, hard and fatiguing.

SNR

Signal-to-Noise-Ratio. All about how much bigger the signal is compared to the hiss or hum — or maybe vice-versa... The table nearby provides a handy check on typical figures, but it's much easier to listen, and more accurate too. This is because SNR figures (like sound level readings) are all too often faked by weighting and averaging techniques. Besides, the ultimate relevance of SNR is whether hiss or hum intrude. For a given SNR, certain types of noise are less noticeable than other forms, and this can make a nonsense of SNR figures.

THD

Total Harmonic Distortion. Here we're discussing spurious signals which bear some harmonic relationship to the legitimate output. The idea of the total is not that we add them all up ("Now, 2nd plus 3rd plus 4th harmonic, let me see on this 'ere calculator"), but rather it's a reflection that basic distortion measuring equipment only gives the sum of all the spurious harmonics. Trouble is, the lower harmonics (2nd, 3rd, 4th) are cool, whereas higher ones (7th, 8th, 9th) are distinctly nasty. Thus a simple THD figure doesn't tell us anything useful about how the distortion will sound. Under some conditions, skilled listeners can wince at a THD of 0.1%; under different conditions, 5% or more may pass by without comment. For this reason the very low figures put on equipment (eg: less than 0.00015% 'distortion') are often an attempt to make it as low as possible, in the vain hope it will go away. Without really knowing how much distortion is acceptable, it's all a bit hit'n'miss. Advice: Good equipment will be specified with a distortion figure which holds at all audible frequencies and levels. For example, I'd trust a spec which reads "distortion = 0.1%, 20Hz to 20kHz, at all levels", but when a maker writes "distortion = 0.0001 % at 1kHz at -20dBU", it's a bit like telling us a 2.8 litre Granada does 35mpg — they forget to mention that at 20kHz (100mph) distortion (mpg) rises to 2% (10mpg).

Typical signal-to-noise ratios:

THIELE

Neville Thiele was an Australian who, in 1962, published the first-ever comprehensive formulae for tuning vented speaker cabinets. Prior to this, 'design' was by laborious cut and try techniques. Like all good things, Thiele's method was ignored by all and sundry except Electrovoice in the US and ATC in the UK, who took up the technique in the early 1970's. But by 1981, other speaker manufacturers had got their Thiele act together, and the world of bass amplification has become smaller, lighter and more efficient as a result.

TOP-END

Synonym for treble, or high frequencies.

VU

Stands for Volume Unit. The VU meter was devised in 1939. VU meters read the average signal level, and give a good idea of perceived (ie. actual) loudness, but are naff for recording purposes, because they're too sluggish to respond to spikey peaks which drive tapes into saturation and distortion.

Z

Shorthand for impedance. So Zin = input impedance, Zout is output impedance. Also 'High Z' = high impedance (etc).

ZERO LEVEL

NOT the same as zero signal. Because decibels are a relative measure and have logarithmic qualities, they can be either positive (bigger) or negative (smaller) than the reference, which itself must be bang in the middle of the scale and therefore labelled zero. Zero level is therefore 0dB, that's the referral point. So 'going to zero' means not nothingness, but simply returning to the point whence we came.