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.
Gaff * Gaffer tape (Gaffa tape, Gaff tape) * Gain * Gain before feedback * Gain cell * Gain reduction * Gain riding * Gain stage * Gain structure * Galvanic * Galvanic connection * Galvanic isolation * Gamma correction * Gate * Gated reverberation * Gauge * GBF * Gear * Gel * Gender * Gender bender * Gender changer * Genderless * General MIDI * Generation * Generator set * Genny * GEQ * Get-in * Get-out * GFCI * GFI * GHz * Gig * Gland * Global * GND, Gnd * Gobo * Gone down * Gooseneck * GPO jack * Grams * Grand master * Grand piano * Graphic * Graphic equaliser * Gray scale * Grazing effect * Green Book * Grey scale * Grille * Ground * Ground-compensated * Ground fault * Ground fault circuit interrupter * Ground fault interrupter * Ground isolator * Ground lift * Ground loop * Ground potential * Ground sense * Group * Group bus * Group delay * Group fader * GRP * Guitar pickup * Guitar processor * Gun microphone * Guru * Gyrator
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.
Gaffer tape (Gaffa tape, Gaff tape)
Gaffer tape must not be used for securing heavy cables or other items overhead, as it may unexpectedly tear or become un-stuck. Also it must never be used as electrical insulating tape, as it is not designed for that purpose and even the black variety cannot be relied upon to have good insulating properties. Further, as most types are flammable, gaffer tape must not be used close to hot surfaces such as lanterns, or in the vicinity of any source of ignition, such as pyrotechnics. It gets its name from the gaffers (chief electricians) of film sets, where the use of this kind of tape first became common. 'Advance Gaffa' is a registered trade mark of Advance Tapes International Ltd. See also Duct tape. Compare PVC tape.
It is important to understand that, as regards audio signals, the amount of gain is not a direct indication of the loudness of the sound produced by the speakers. An increase in gain at some point in the signal path (with no corresponding reduction elsewhere) will produce an increase in sound level, provided that limits on the levels being handled (from that point onwards) will not be exceeded. However, a large gain applied at some point does not necessarily imply a high sound level. This is because the initial level of the signal may be extremely low, or because part of the gain applied at that point may be counteracted by a lack of it (or by a loss) elsewhere in the chain (or a combination of these two factors) − see Gain structure.
The gain provided by an item of equipment (or by part of it) gives no information about the maximum output voltage level available from the equipment, nor about its current-supplying capability or maximum power output. It only indicates the amount by which the signal level is increased by that equipment. Where the gain of the equipment can be adjusted, a gain figure quoted for it will typically be the maximum gain that it can provide.
In some contexts, however (particularly radio-frequency ones), 'gain' refers to power gain − that is, the factor by which the equipment (or an internal part of it) increases a signal's power level. Again, this gives no information about the maximum power output level of the equipment.
In general, the amount of gain provided by an item of equipment will be different for different frequencies that may be present in a signal, according to the equipment's frequency response. In cases where the frequency response is essentially flat over the whole operational frequency range of the equipment, quoted gain figures apply over that whole range. In other cases, where the gain varies substantially with frequency, this may be intentional behaviour (as in the case of equalisers) or undesirable. Either way, it is important to understand at what frequency, or over what frequency range, any quoted gain figures apply. In the case of audio signals, a 'reference frequency' of 1 kHz is commonly used.
In the case of a control marked "Gain" on PA equipment (on a mixer, for example), its function is usually to set the amount of initial amplification given to an input signal by a pre-amplifier within that item of equipment, in order to raise the signal to an appropriate level for subsequent processing. So, it allows the equipment to be adjusted to accommodate a range of different input levels − or we could say it controls the sensitivity of the input. In order to obtain the optimum dynamic range from the equipment, it is necessary to adjust such gain controls so as to provide the maximum amount of gain that is possible without excessive distortion (typically, clipping) occurring during the highest peaks of the signal supplied to the equipment. In order to assist with this adjustment, suitable metering facilities are usually provided (see below). On some equipment (e.g. by Mackie), the gain controls are labelled "Trim" (however, compare Trim (2)).
However, on backline equipment that is designed to provide the facility for intentional distortion, such as most types of guitar amplifiers, a control marked "Gain" is often considered to have the function of determining the amount of distortion, if any, produced. In this context the term 'gain' is frequently used by guitarists to refer to the extent of intentional distortion that the equipment is being requested to produce.
As making an adjustment to a gain control will change the output level of the equipment (assuming that the input level remains the same and that neither clipping nor limiting occurs), it is important that the gain is correctly set before the adjustment of other controls that affect the output level (such as the channel fader or Aux Send controls of a mixer). The correct setting of mixer gain controls is best determined by use of the channel metering facilities in accordance with the manufacturer's guidance − for further information see Gain or Trim control on the Mixing Facilities page and How should I set up my channel Gain controls? on the FAQ page.
The way in which overall gain is distributed throughout the signal chain is a very important factor in achieving the optimum overall signal-to-noise ratio and headroom − for further information see Gain structure. See also Ride (1) and Digital gain. Compare Loss.
Or, under given conditions of control settings, the amount by which a noise gate decreases its gain when the gate is fully closed (input signal level persistently below threshold), as compared to its gain when the gate is fully open (input signal level persistently above threshold).
In both cases, the gain reduction is usually specified as a (positive) value in decibels.
In a PA system, a fundamental aspect of achieving of a good gain structure is ensuring that the channel gain controls of the mixer are correctly set for each channel (see Gain). It is considered by some that ideally the remainder of the signal chain through the mixer, and through each item of outboard equipment, would provide unity gain, but as this may not achievable in practice a good guideline is to avoid very low or very high settings on faders or other level controls. (If this appears to be necessary, there may be equipment compatibility issues that need to be addressed.)
Setting the gain structure and the nominal working level in a methodical manner is collectively referred to as 'lining-up' the system (see Line-up (2)), and may be achieved by use of a test tone. A common approach is to ensure that the level controls of each item of equipment are adjusted such that the working headroom is the same (appropriate) value through each and every item. See also Metering and Standard operating level.
Some common examples of the latter method are given below; the figure before the slash is the number of strands, and the figure after it is the strand diameter in mm.
The thicker the conductors, the lower the resistance of the cable (per unit length) and therefore the greater the current that it can carry without overheating or causing an unacceptable loss of power. (Note, however, that other factors also influence its current-carrying capacity, such as the type of insulation employed and the ambient temperature when in use.)
The gauge of cable conductors is of special importance for interconnections between power amplifiers and speakers, because of the high currents that flow in these interconnections and the need to maintain a high damping factor. The round-trip series resistance of various metric gauges of speaker cable is given below; if the conductors get warm in use or are installed in a hot environment (e.g. in proximity to stage lighting), then the high temperature figures should be used. Note that resistance applies only at DC − at high audio frequencies the value of the cable's series impedance will be significantly higher than its resistance, because of inductive effects. For AWG sizes of cable, see AWG.
For PA speaker applications the minimum size usually used is 2.5 mm² (13 AWG), but the minimum acceptable size will depend on the power being supplied to the speakers, the combined impedance of the speakers being supplied by the cable, and the total length of the cable between the amplifier and the speakers. As a rough guide, to maintain a good quality sound the recommended minimum gauges are given in the table below, but be sure that the actual type of cable that you use has an adequate current rating), in your particular installation conditions. For very high power systems the use of powered speakers is recommended.
In this table, the cable currents indicated assume the speaker has unity power factor, and the minimum damping factors indicated (in brackets) are given at DC assuming cable conductors at 70 ºC and a value of 0.1 ohms for the total of the amplifier output impedance and the connector impedances. The mediocre damping factors indicated for a 4 ohm load can be substantially improved upon by selecting the next higher cable size, where practicable.
The extremely large cable sizes indicated for long cables and low impedance systems (10 mm² is larger than will fit most types of Speakon connectors) are an important factor in the desirability of siting amplifiers as close to their speakers as possible, and in selecting an acceptable system impedance − especially in powerful systems. It can also be seen that where an amplifier has multiple speaker output connections, it is always better to run separate cables from these rather than 'daisy-chaining' a single cable (of the same gauge) from speaker to speaker.
Note that for high power systems and for 100 volt line systems, the voltage rating of the cable is important as well as its gauge. For further information on connecting amplifiers to speakers see the System Assemblers page and the Amplifiers and Speakers page. See also AWG.
The smaller models (up to about 5 kVA) usually use a petrol engine, whilst larger ones are usually diesel-powered. For reasons of electrical safety, it is usually necessary to install an earth rod in order to provide a safety earth. CAUTION: Generator sets must only be used in well-ventilated areas outdoors, because of the hazardous exhaust fumes produced by the engine. Slang term: genny. See also Inverter.
Short, thin goosenecks are also often used between miniature clamp-on instrument microphones and their clamps, to enable the mic to be positioned and pointed in the optimum direction without adjusting the clamp.
The more sophisticated types may also provide feedback indication and/or suppression facilities. Digital types do not have a separate slider for each band, but allow separate control of each band from a common set of buttons.
The budget versions divide the audio frequency range into a small number of bands, whilst the professional units divide it into a larger number of bands, giving a finer degree of control. Graphic equalisers used in PA work are usually octave, 2⁄3 octave or 1⁄3 octave types; very occasionally 1⁄2 octave types are also seen. These designations indicate the ratio between the centre frequencies of any two adjacent frequency bands of the equaliser:
N.B. Although a PA graphic equaliser just looks like a larger version of the one you often find on a domestic hi-fi system, and what it does to the signal is just the same, the graphic equaliser on a PA system is provided for an entirely different purpose to the one on a hi-fi system. Whilst the one on a hi-fi system is (in practice) there to enable the listener to adjust the sound to suit his or her preference, the purpose in a PA system is to allow the sound to be adjusted in order to compensate to some extent for the deviations from a flat response that are introduced by the combination of the particular speakers used and the acoustics of the room, taking into account the placement of the speakers within the room. This means that for a permanently installed PA system, once the graphic is correctly set up it should never be adjusted again, unless the acoustics of the room are changed (e.g. by a change to furnishings or carpeting), or the type or position of the speakers is altered.
Note, however, that a graphic equaliser cannot alter the room acoustics. A highly resonant space will always tend to ring at its resonant frequencies and cause colouration of the sound. The only solution to this is treatment of the room acoustics, generally by an increase in the quantity or effectiveness of absorbent material present. See also Parametric equaliser, Spectrum analyser, Q (1), Proportional Q, Constant Q and Smiley face.
The extent of the additional reduction, which is called the grazing effect, and the frequencies of the sound which are most affected by it, is heavily dependent upon the nature of the ground surface across which the sound has to travel − a crowd of people would have much more effect than a bare wooden floor. The grazing effect can be reduced to some degree by positioning the speakers well above ground level; once this has been done, further advantage can be gained in an auditorium setting by banking the seating, so that the path of the sound to the people located at the back of the audience is well above the heads of most of the audience located in front of them. If the speakers are positioned at an insufficient height, however, banked seating will only make matters worse.
Grey scale (Gray
The output achieves this rejection by using one of the cable's conductors (called the 'ground sense' conductor) to sense the signal-earth voltage at the destination end of the interconnection. The output drive circuitry then superimposes that voltage on the signal voltage provided at the output's 'hot' terminal, so that the signal input voltage 'seen' by the destination equipment, relative to that equipment's own signal earth, is (in theory) just the wanted signal voltage.
So, the ground-sense terminal of the output connector is actually an input connection. This connection is usually made through pin 3 of a 3-pole XLR, or through the ring of a TRS jack. Pin 2 (or the TRS tip) carries the signal ('hot' conductor) and pin 1 (or TRS sleeve) provides the signal earth connection (usually made via the screen of the cable).
The input impedance of the ground sense terminal is arranged to be essentially the same as the output impedance of the 'hot' terminal, so that, when a ground-compensated output is interconnected with a balanced input, with the ground-sense conductor connected to the 'cold' terminal of the balanced input, the interconnection is able to operate much as a semi-balanced interconnection. (In such a case, the ground-sense conductor is not connected to the signal earth of the destination equipment and so no distant 'ground voltage' is sensed by the ground-compensated output. Therefore, no rejection is provided by the output; rejection of common-mode interference is provided by the balanced input, just as in the semi-balanced case. However, as any interference picked up on the ground-sense conductor will be added by the compensation circuit to the signal output provided on the 'hot' conductor, the interference level on the two conductors will no longer be the same and the rejection provided by the balanced input will be impaired.)
Although the reduction in earth-loop induced noise obtained by interconnecting a ground-compensated output with an unbalanced input may be comparable with that provided by a true balanced interconnection, such an interconnection is inferior to the balanced case in that:
When a ground-compensated interconnection is used with unbalanced destination equipment, the output can only provide compensation for earth voltage differences if the pin 3 / ring (ground-sense) conductor is connected to the signal earth of the destination equipment. In the case of destination equipment having a 2-pole input connector, a 2-pole plug should preferably be used and the cable's ground-sense conductor and screen both connected to the sleeve terminal of the plug (but see the next paragraph...). Inserting a 3-pole (TRS) jack plug into a 2-pole jack socket is unlikely to result in a reliable connection between the plug's ring contact and the signal earth of the equipment. [Note that this requirement is in contrast to the case with a 'normal' balanced output, whose pin 3 / ring ('cold') conductor should not be shorted to signal earth, as to do so would cause large currents to flow on that conductor due to the low output impedance of the line output; such currents could induce interference into other circuits − see Inductive coupling − and/or stress the line output components. When such an output is connected to an unbalanced input, the 'cold' conductor should be left unconnected, preferably at the destination end of the cable.]
It is debatable as to whether or not the pin 1/sleeve (screen) conductor should be connected to the signal earth at both ends of the interconnection. Such a connection would (in the presence of an earth voltage difference) cause an earth current to flow in that conductor. Although the voltage difference should in theory be largely rejected by the ground-compensation arrangement, the current flow can cause effects such as a voltage drop between the sensing point (at the connector terminal at the destination end of the cable) and the true internal signal earth of the destination equipment, due to impedances in the signal earth connections (especially if dirty or worn jack connectors are in use). [In the case of poor quality cables, the current flow in the cable screen can also cause a small differential voltage to occur between the inner conductors, due to differences in the inductive coupling from the screen to the two inner conductors, but this applies equally to fully balanced interconnections.] However, disconnecting the signal earth (screen) at one end is inadvisable, as it turns the screen into a long aerial which can inject radio-frequency interference at the connected end, and leaves the inner conductors effectively less-well screened at the unconnected end (due to the screen impedance). Sometimes a compromise is adopted by connecting a capacitor (e.g. 0.01 µF) in series with the signal earth at one end; this allows RF to pass but effectively blocks mains-frequency earth currents. However, the usual preferred practice is to have the screen connected to signal earth at both ends.
Note that the ground-compensated scheme cannot work properly when the output is feeding multiple inputs (unless the signal grounds of those inputs are connected by an extremely low impedance − for example when they are inputs on the same item of equipment).
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, Unbalanced, Semi-balanced, Quasi-floating and Pseudo-balanced.
In an audio group (also called a 'sub-group' or a 'sub-mix', because the group mix is usually a subsidiary component of the main mix), the audio signals from the channels selected to make up that group are mixed together onto a group bus. The resulting sub-mix may then be controlled in level by the group fader(s), and is usually available at a group output. An insert point is often also provided for it, to enable the group to be passed through an outboard serial effects unit before being returned to the mixer to be added into the main mix.
Note that, as the audio group is mixed after the channel pan controls, two group buses and two group faders are normally used to allow stereo operation − one for Left and one for Right. By convention, Left buses/faders are usually odd-numbered, and Right buses/faders even-numbered. (When an additional mono output is provided, to feed a Centre speaker, an additional fader may be provided for that mix − see LCR (1).)
When using outboard parallel effects units fed from channel post-fade auxiliary sends, a potential difficulty arises with audio groups: adjusting the channel fader keeps the effect level in proportion to the direct (dry) level, but adjusting the group fader alters the balance of effects and dry signal. This is because a group fader adjustment alters the direct signal level without a corresponding change to the effect level (unless the effect return is via a channel in the same group).
VCA groups are possible only on VCA-equipped mixers, and are generally found only on the larger types − but are becoming a more popular feature. They operate in an entirely different way to audio groups, in that the group fader does not carry an audio signal but rather 'remotely' controls the level of post-fade signal passing through each of the selected channels (in addition to the 'local' control provided by the channel fader). This 'remote control' is achieved by a DC control signal that is applied to a VCA (voltage-controlled amplifier) on each channel. So, no separate sub-mix of the signals from the 'grouped' channels exists, and therefore cannot be output from the mixer or processed via an insert socket. When more than one VCA group fader is assigned to a particular channel, all of the assigned VCA group faders have control over the level of signal passing through that channel.
VCA grouping does not suffer from the potential effects-imbalance problem experienced with audio groups, because a VCA group fader affects the level of the post-fade auxiliary sends as well as the level reaching the main (or audio group) mix bus. Also, because the group level adjustment takes place before the channel pan controls, only a single group fader is needed − even for LCR mixes.
Because each of these two methods has its own merits, mixers which provide the VCA grouping facility usually also support audio grouping (but often to a lesser degree than mixers without VCA groups). See also Matrix.
A system in which the phase response changes linearly with (linear) frequency over a particular frequency range (that is, a system having a linear phase response over that frequency range) will have a constant group delay over that frequency range, which is a good thing in high-quality audio systems.
Group delay usually remains fairly constant throughout regions of constant frequency response (e.g. well within the passband of a filter), but can change very significantly where the frequency response is rapidly changing (e.g. around a filter's cut-off frequency or frequencies). This has important consequences for the design of audio crossovers and in the design and use of equalisers.
In general, group delay does not refer to the absolute signal delay time through a system, but to the amount of delay experienced by a 'group' of similar-frequency signals relative to that experienced by signals of different frequency. However, provided that the group delay is reasonably constant over a particular frequency range, and that the frequency response is also reasonably constant over that range, then the group delay is the actual time delay that a 'group' of signals occupying only that frequency range will experience on passing through the system. See also Minimum phase. Compare Latency.
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 08-Jul-2017.