Glossary of PA Terms - Q
- If you have arrived here by clicking on a linked term on another page of this site, it may take a moment before your browser jumps to the definition of the term that you clicked on; thank you for your patience. (If there's still no movement after a few seconds, you may have encountered a broken link; please report it.)
- If you have arrived here from a search engine, or by clicking on an alphabet letter on another page of the Glossary, then click on your required term in the list below.
The glossary pages provide definitions for over 2680 PA-related terms. If you can't find the term you are looking for, or would like any of the existing definitions to be expanded, please email me − likewise of course if you find any errors in the links etc. Use of this information is conditional upon acceptance of the Disclaimer on the PAforMusic home page.
In the list below, the most commonly looked-up terms are in bold, lighting-specific terms are in pink, and video-specific terms are in orange.
Q * QTBMC * QTP * Quad cable * Quad twisted pair * Quality factor * Quantisation * Quantisation distortion * Quantisation noise * Quarter track * Quasi-balanced * Quasi-floating * Quasi-parametric equaliser * Quick-connect terminal * Quiescent current * Quiet
The definitions for these terms are given on the assumption of their use in the context of PA systems; many of the terms have more general meanings when used in a wider context. Where more than one definition is given for a term, the definitions are numbered (1), (2) etc.
Some of the definitions themselves use terms (such as "signal") in a specific way − most of these are links (just the first time they are used, in each definition), so just click on them to see the meanings that are intended.
A measure of the width of the range of frequencies that are affected by a peaking equalisation control, in proportion to the equaliser's centre frequency. A higher Q means that a smaller range (or 'band') of frequencies is affected by the equaliser. So, a control having a lower Q will have a greater effect on the sound (for the same amount of cut or boost).
The term 'Q' is an abbreviation of 'quality factor', and derives from the fact that the bandwidth of early equalisers, which used a passive design, was reduced by the use of better quality inductors, which have lower losses (see Q (2)).
Numerically, the value of Q is given by the centre frequency divided by the bandwidth. If we consider a graphic equaliser, we can assume that the bandwidth of each band runs from a point logarithmically midway between its own centre frequency and the centre frequency of the band below, to a point logarithmically midway between its own centre frequency and the centre frequency of the band above. On this assumption we can calculate that the approximate Q value for the bands of an octave equaliser is 1.4, the Q value for a 2/3 octave equaliser is 2.1, and the Q value for a 1/3 octave equaliser is 4.3. (The values for 2-octave and half-octave types would be 0.7 and 2.9 respectively.) So, the range of Q values that can be set by the Q control of a parametric equaliser would typically be 0.5 to 8. See also Proportional Q and Constant Q.
A measure of the effective series resistance of an inductor, in proportion to the reactance of the inductor at a particular frequency. (So, the Q of a particular inductor will change with frequency). A higher Q means that the inductor behaves more like an 'ideal' inductor, which would have no effective series resistance and therefore have an infinite Q. So, the term 'Q' is an abbreviation of 'quality factor'.
Note that the effective series resistance is not just the resistance of the conductor that forms the inductor, but also includes magnetic losses (which in the case of cored inductors can be considerable at high frequencies). Numerically, the value of Q is given by the reactance divided by the effective series resistance.
A value specifying the extent to which a resonant system (e.g. a speaker or driver) is damped. A higher value of Q indicates a less-damped resonance; in terms of speakers and drivers low Q values are usually more desirable as they help to provide reduced colouration of the sound produced. This more general definition of Q is the basis of the previous two more specific definitions.
Q values of this kind that are specified for a speaker or driver are more properly designated the Qes (electrical Q), Qms (mechanical Q) and Qts (total Q). Note that these values of Q are completely unrelated to the directivity Q (see Q (4)) value for the speaker or driver. See also Resonant frequency.
A value specifying the directivity of a speaker or a driver. Note that this value is completely unrelated to the resonance Q (see Q (3)) value(s) for the speaker or driver. For further information see Directivity.
An abbreviation for 'quad twisted pair', though the full term is rarely used. In general, refers to any twisted pair cable that has 4 pairs, for example standard CAT 5e or CAT 6 UTP cable such as is commonly used for computer network interconnections.
However, in PA work the abbreviated term is more frequently used to refer to arrangements in which this same type of cable is used to carry four independent audio signals, most commonly either fully balanced line-level analogue audio or AES3 digital audio. When used in this way, certain restrictions must be applied in order to ensure the integrity of the signals being carried, for example care must be taken to avoid sharp bends or kinks in the cable (as usual) and nearby strong magnetic fields must be avoided.
As indicated above, most commonly unscreened cable is used, so only the hot and cold connections of each channel are carried. Where screened cable is used, the screen(s) are connected only at the sending end − so again only the hot and cold connections are carried.
As UTP cable is frequently terminated with 8p8c RJ45 connectors, a scheme is needed to allocate the 4 audio channels to the 8 contacts of these connectors. In order to avoid unacceptable crosstalk, it is essential to use a scheme that allocates each channel to a pair of the cable. This means that simply using adjacent pairs of contacts is not possible, because the hot and cold contacts for one of the pairs are spaced apart, with two other contacts in between. Thankfully the connections of the individual cable wires to the connectors are standardised, so all that is needed is to allocate each audio channel to the two contacts designated for each pair. It might be thought that the best way to do this is to follow the standard numbering convention of the cable pairs according to their colours i.e. 1 to 4 would be blue, orange, green, brown. However, unfortunately there are two different standards for terminating the cable pairs to the connectors, known as EIA/TIA 568A and 568B (though the 568B wiring is very much more common), and neither of these places the pair 'hot' conductors, numbered according to colour, in sequence across the numbered connector contacts (1 to 8). Therefore, a preferred approach is to allocate the channels to specific pairs of RJ45 connector terminals, regardless of the colours of the cable pairs. For example, numbering the channels' hot conductors in order across the even-numbered contacts (solid-coloured wires) gives the following allocation:
This provides a standardised method of connections to RJ45 sockets, and all is well when the cables plugged into them are (correctly) terminated with RJ45 connectors at both ends. However, where one end of a cable is terminated other than with an RJ45 connector (e.g. is directly wired, or is broken-out to 4 separate audio connectors) it becomes necessary to know which pair colour corresponds to which audio channel. This is only possible if it is known which of the two connector termination schemes is in use at the RJ45 end of the cable. Assuming the most common 568B scheme, the correspondence between the channels (allocated as above) and the pair colours will be:
Therefore, in this scheme, where a breakout cable breaks the channels out into individual cables coloured corresponding with the 568B scheme pair colours, then the channels (numbered 1 to 4) would be identified in the sequence orange, blue, green, brown. (Note that this reverses the order of the first two colours, as compared to the standard cable pair colour numbering system, so it has been suggested that to help remember this ordering the colours are referred to as 'sun', 'sky', 'grass' and 'earth'.) This matter, and other matters connected with the use of QTP cable for audio purposes, is addressed by standard AES-X246, which is not yet finalised. See also Format A.
See Star quad cable.
Quad twisted pair
See Q (2).
In analogue to digital conversion, the process of assigning a numeric value to each sample of the analogue signal. This may be done linearly (as in linear PCM) or using more complex schemes such as floating point coding (see Float (1)).
See Quantisation noise.
The added noise that results from the inaccuracy inherent in the assignment of a discrete numerical value to each sample of an analogue signal's instantaneous value, during analogue to digital conversion. The greater the number of bits used for each sample (called the bit depth), the more accurately its analogue magnitude can be represented in digital form and therefore the less quantisation noise is added.
Assuming linear quantisation, the level of quantisation noise voltage is halved for every additional bit employed for each sample. This means that for 16-bit sampling its level is 96 dB below full scale level (0 dB FS). So, 16-bit sampling is sometimes quoted as providing a 96 dB signal-to-noise ratio (SNR), but this is somewhat misleading on two counts. Firstly, it is only a theoretical optimal figure as it takes account only of quantisation noise, whereas in practice there will also be other sources of noise acting to reduce the dynamic range. This ambiguity is corrected by use of the term 'signal to quantisation noise ratio' (SQNR). Secondly, both these terms are comparing average signal level with noise level, and since the average signal level will be set somewhat below the 0 dB FS level in order to provide headroom, the actual SQNR will be less than 96 dB by that amount.
What about '96 dB dynamic range' − is that a better description of a 16-bit system? Well firstly, like '96 dB SNR' it takes account only of quantisation noise, whereas in practice there will also be other sources of noise acting to reduce the dynamic range. Secondly, it implies that a signal at −96 dB FS can be accommodated by the system, albeit at the same level as the noise. Whilst an analogue signal of that level at the input of the analogue to digital converter would indeed be at a level equivalent to the level of quantisation noise about to be introduced, a signal of such a level cannot reasonably be considered to exist in meaningful form in the digital domain, as it would have only one least-significant bit's worth of digital amplitude to represent it and therefore would not be recognisable as the original signal when eventually converted back to analogue. So it is very dubious indeed to imply that the system can handle signals of any level between −96 dB FS and 0 dB FS, in the same way that a purely analogue system with a 96 dB dynamic range can.
Quantisation noise may sometimes be referred to as 'quantisation distortion', but is not a form of distortion in the true sense. Note that the American spelling is 'quantization noise'.
Describes a particular type of electronically-balanced output, designed to simulate some aspects of the behaviour of an earth-free (floating) transformer-balanced output. It is alternatively described as 'servo balanced', 'cross-coupled' or 'cross-linked'. In particular, when one leg of the interconnection from such an output is shorted to signal earth, the signal level provided on the other leg is automatically increased by 6 dB (i.e. is doubled in voltage), so as to maintain the same voltage between the two legs.
This behaviour has two benefits:
- If the shorting to signal earth occurs as a result of connecting the output to an unbalanced input, the effect of the increase in level on the other leg will be that the same level is delivered to the unbalanced input as would have been delivered to a balanced input − but with a loss of 6 dB of headroom. In such a case, the shorting to earth is most likely to be provided by inserting a 2-pole jack plug into a 3-pole jack socket at the sending end; this will short the 'cold' output signal on the ring connection of the socket to the sleeve of the plug. Alternatively, the short may be deliberately provided at the destination end of the interconnection, in which case the 'cold' signal conductor will act as a 'ground-sense' conductor, causing the output to operate as a ground-compensated output.
- If the shorting to signal earth occurs as a result of a fault on an interconnection to a balanced input, the effect of the increase in level on the other leg will be that the level received by the destination equipment will be unaffected by the fault. In this case some degree of ground-compensation may be provided by the output, especially if the fault occurs close to the destination equipment. However, rejection of common-mode interference by the destination equipment's input is likely to be severely compromised by such a fault.
However, some types of quasi-floating output can become unstable or provide an excessively noisy signal if no connection is made to either its 'hot' or 'cold' legs, and the level appearing between the other leg and earth may not be as expected under these conditions. If connecting this type of output to an unbalanced input, it is therefore generally advisable for the unused leg (normally the 'cold' one) to be intentionally shorted to earth, as described above. (Though note that such intentional shorting is generally not advisable with other types of output.)
In the event of a disconnection fault occurring on one of the legs of an interconnection, the level provided on the other leg by this type of output should (with a well-designed output circuit) be unaffected, and so the level received by a balanced input of the destination equipment should fall by 6 dB (just as in the case of a conventional fully balanced output). However, as mentioned above, other unexpected results may be obtained under these conditions, notably a substantially increased noise level. See also One-legged.
A table comparing the most common types of balanced interconnections is provided under the 'Balanced' entry. Diagrams illustrating various different types of signal interconnections are available here (opens in a new window). Compare Balanced, Semi-balanced, Ground-compensated and Pseudo-balanced.
See Parametric equaliser.
See Spade terminal (1).
The DC current that flows through the power-handling devices in the output stage of a class AB power amplifier during the times that no signal is being handled by the amplifier. Many amplifiers of this type incorporate an internal trimmer for adjustment of the quiescent current; correct adjustment is important in order to minimise crossover distortion and ensure optimum operating efficiency. The term 'quiescent current' is usually reserved for transistor output stages, though in principle the bias adjustment of a valve output stage has an equivalent function. See also Output transistor. For more information see the Amplifier Classes section on the Amplifiers and Speakers page.
Describes sound or a signal that contains very little noise relative to the wanted sound or signal level, i.e. that has a good signal-to-noise ratio. Or, describes equipment (or a part of an item of equipment) that contributes very little noise to the signal(s) passing through it, and therefore has little detrimental effect on the overall system signal-to-noise ratio. The term is most often used of the kinds of equipment, or parts of equipment, that might be expected (if of lesser quality, or faulty) to contribute a significant amount of noise, such as pre-amplifiers. Compare Noisy.
A subjective term, describing sound judged to be of low loudness relative to an implied, specific or arbitary reference. This is the meaning usually intended by non-technical persons. See also Sound pressure level. Compare Loud.
There are no more definitions on this page. (The space below is to facilitate linking to the last few terms above.)
This page last updated 24-Dec-2020.