In the below we will proceed with following the audio signal from the artist through the various wires and equipment until it reaches the speakers. This is the usual procedure to apply whenever there is some problem with the system that you don't know the cause for - follow the signal, carefully checking each lead, plug and piece of equipment until you isolate the problem. A bit of background definitions and equations are given in the first section, but these really gloss over alot of the details, so see the references for more on this.

Background definitions

Measuring audio signals: The key unit in audio is the decibel (dB) where deci is from the Latin for one tenth and bel is from Alexander Graham. A Bel is a logarithmicly scaled measure defined as the logarithm (base 10) of the ratio of two numbers. Since 1 Bel has 10 decibels, the formula is decibels = 10 log(A/R) which measures the relative relationship between A and a reference R. The reason for using logarithmic scales here is twofold: the human ear responds to sounds in much more of a relative manner than an additive manner, and the range of measurements of various audio signals is so large that on a linear scale sufficiently large to cover the entire range, low signal levels would be indistinguishable from zero.

To calculate dB therefore, you need a reference level as well as a signal to compare to this reference level. For power levels (measured in watts say) a doubling of a signal corresponds to an increase of 3dB since comparing a signal 2A to a reference A gives dB = 10 log (2A/A) = 10 log(2) = 10 (.301) = 3 (approx.) and similarly halving the power corresponds to a decrease of 3dB. Note that a 10 fold increase in a signal corresponds to a 10dB increase. For dB to be useful, it's important to know the reference level and there are several different dB measures depending on what you are measuring and what the reference is. The above formula is for power ratios, while for voltage ratios to be measured in decibels, it is necessary to remember that power is proportional to the square of voltage (from Ohm's law V = IR and P = I^2 R ) P = V^2 / R where P is power, V is voltage, I is amperage, and R is resistance. Due to this, calculating dB differences between two voltages (or two sound pressure levels - SPL) is decibels = 20 log (A/R)

You will see lots of different dB measures including:
dBV for the case of voltage in which the reference R is 1 volt rms
dBm for the case of power output with reference R 1 milliwatt
dBu for the case of voltage with reference R 0.775 volts

In the above rms stands for root mean square and is useful in describing the average level of a varying signal such as a complex waveform. Note that for measuring sound pressure level (SPL) which is a measure of the force of air pressure provided by a sound system at a location, a doubling of SPL corresponds to a 6dB increase (here 0dB for SPL corresponds to the threshold of hearing in the ears most sensitive frequency range - about 1kHz). Another rule to keep in mind is the inverse square law - for a fixed sound source, for each doubling of distance from the source, the SPL will drop by 6dB since the power produced by the source is spread over approximately four times the area.

Microphones and other inputs

There are several different types of microphones, with hosts of manufacturers for each. The basic types are:

Dynamic: here the mic is like a speaker in reverse, since there is a diaphragm which vibrates according to the sound applied, causing a coil of wire to move in a magnetic field producing a very small electrical signal.

Condensor: here the sound is picked up by a capacitor, which must be provided power either from a battery, or from phantom power provided along the mic line (this is a DC current provided by either a mixer or by a separate phantom power unit)

There are also a wide variety of other types (e.g. ribbon mics, radio mics, electret condensors, etc.), but the vast majority of live sound work is with the above two types.

Key factors determining differences between various microphones are their frequency response and the pickup pattern. The frequency response is quite different for mics designed for use by vocalists than for those designed for various instruments, so different mics are typically used for these purposes (e.g. a Shure SM 57 for instruments and a Shure SM 58 for vocals). The pickup pattern for the majority of live sound mics is either cardioid, with a heart shaped pickup pattern around the central mic axis, or supercardioid, which is more directional than a cardioid, particularly designed for cases in which you want to reject some acoustic signal from the sides that a cardioid mic will generally pick up. Exactly what mic to use where depends upon what you have available, the artists preferences, and your experience with the particular instrument.

Other inputs aside from mics are direct line signals (line level is -10 to +30 dBu and is much higher signal level that mic level which is typically -40 dBu or lower) which are typically obtained from an on-stage amp, or from a DI box (DI stands for Direct Injection, though these are typically just called direct boxes). A DI simply converts an unbalanced, high impedance signal from an instrument pickup or amp to a balanced low-impedance signal. There are two kinds of direct boxes - passive, which is essentially just a transformer inside a shielded box for converting a high impedance to a low impedance signal, and active, which require a battery or phantom power to operate.

Lines - balanced and unbalanced

There are two basic types of lines used in audio:

Unbalanced lines have a single lead running down the middle, with a wire braid shielding around it. Here the hot signal (e.g. in-phase or +) is in the center wire and the braid serves as both ground and the cold side (e.g. out of phase or -) of the signal is carried by the braid. The end of the line typically has a quarter-inch jack plug with just a tip and sleeve. Unbalanced lines are used typically only for the relatively short leads from an instrument to an amp or DI.

Balanced lines have two center wires carrying the in-phase (hot) and out-of-phase (cold) signal, with a wire braid around them both which is the ground. The typically end of the line is a Cannon or XLR type plug with the male end sending the signal and the female end receiving the signal.

What you want to be sure is that all connections to the mixing console (and any snake going to the mixer from the stage) is with balanced lines. Otherwise noise would be picked up in an unbalanced line and dumped right into the mixer. A balanced line greatly reduces noise problems (due to spurious electrical transients produced along the length of the line) since the shielding dumps this to ground in the mixer.

The other caution is to be sure not to use any speaker lines for connecting the audio components prior to the amplifier stage. Speaker wires have two wires to carry signal, but have no wire braid shielding around these. This shielding is essential to reject radio frequency and other interference that would greatly compromise the low level signals being sent to the mixer.

The mixer

Gain setting

A very important factor in making a clean, even mix possible is an appropriate gain structure for all inputs. What this means is that all signals coming into the mixers internal circuitry are at roughly equivalent levels. This is necessary to ensure that no one input controls the amount of headroom (how many dB increase is possible above nominal operating levels) available from the mixer. Setting the gain (e.g. how much amplification goes on in the pre-amplifier stage of the mixer) for each input channel appropriately not only ensures that no one input overwhelms the mixer, but also ensures that the lowest possible noise level is achieved from each input.

Achieving appropriate gain structure is relatively easy, but requires carefully going through each input channel to set the gain (or trim as it's often called) for the preamp stage so that only the appropriate amount of signal is sent into the mixer. Exactly how to do this depends somewhat on the mixer being used. A standard approach is to sent the channel slider at center (0dB), and adjust the input gain on each channel while that channel is being used at the level it will be during the performance (by having the artist sing or play into it) so as to have the VU or LED meters on the mixer show 0 dB. It is often best to roughly adjust the channel EQ at this time as well, since this affects the level from that input going to the mix.

Channel levels and EQ

Once the gain level is set for each channel, there are two other main controls of the input signal - one is simply the slider (or fader as it often is called) which controls how much of that input is sent to the output of the mixer. The level here should typically be close to the center location if you have set the gain correctly, but will certainly be modified from this as the entire set of inputs are mixed together, and should be taken out of the mix completely when the input is not used (a mute button does when you don't want to have to remember or write down the slider position).

The channel EQ (equalization) allows adjustment of particular fixed frequency ranges for each input separately. This allows you to boost or reduce certain frequencies depending upon the needs for a particular input. The exact frequencies ranges used (there are typically Hi, Mid and Low EQ adjustments) vary considerably from mixer to mixer, as well as the structure within these ranges that is affected by the EQ. Some mixers allow you to adjust the frequency affected (particularly in the midrange).

House EQ

This is a graphic EQ that allows you to boost or cut (up to a certain dB) a variety of frequency ranges. The frequency ranges are set up logarithmicly, from typically 20 Hz to 20,000 Hz (10 octaves), so that each slider on the EQ affects an equivalent ratio of frequencies, though the bands covered by any two sliders will be quite different numerically (e.g. the first slider might cover only 5 Hz while the last one might cover 4000 Hz).

You typically set the House graphic based upon the room acoustics, and your expectation for how the room will sound when the audience arrives. Note that the audience can make a considerable difference in how a room sounds, so it is not a good idea to "over EQ" a room during sound check (e.g. cut out alot of frequencies) unless you know from experience that it is needed. There are a variety of methods to "ring out" a room to find harmonic frequencies that might make the sound harsh or indistinct. One method is to pass white noise (e.g. noise with equal power at all frequencies) through the house system and use a frequency analyzer in the house to pick out what frequencies are enhanced, and then reduce them using the graphic EQ. Another method is to simply place a microphone (preferably of the same type you are using on stage) in the center of the hall facing the stage, and slowly bring up the mic level until you get feedback squeals, cutting out the main frequencies of those squeals. You don't want to overdo this though, because you can greatly deaden a room.

Amps and speakers

The output from the mixer goes typically first to a house graphic EQ and then to an amplifier. The amplifier boosts the relatively low signal coming from the mixer to a power level sufficient to drive the speakers that you are using. Amplifiers are heavy and produce alot of heat. It is very important that they have plenty of air flow around them. An amp needs the most power for low frequencies, less for midrange frequencies and the least for high frequencies. It is very important to match the power produced by an amp with the sound requirements of the type of music and the venue, as well as the power that the speakers can handle. It is typical to run an amplifier wide open (e.g. at the maximum output level) so that all variation in output level is completely controlled by the input level to the amp from the mixer. Troubles arise when the input level is too high for a particular amp - this leads the amp to try to reproduce the signal at the appropriate power level, causing clipping. This essentially chops off part of the amplitude of a waveform signal, and causes the speakers to try to reproduce a much higher amplitude waveform than the amp is providing power for. This leads first to distortion, and then, if it continues, the speaker fries (e.g. the cones rip or the coils burn up).

Speakers are of several types, with the majority consisting of coils of wire in a magnetic field driven by the amplified signal causing a cone of material to vibrate and produce a sound wave of the appropriate waveform. Horns are a means to focus the sound in particular directions. Speakers are horrendously inefficient, in the sense that a very small fraction of the power supplied to them actually gets transmitted into sound. Much of the power is lost as heat from the coils. Speaker systems can include separate speakers for different frequency ranges, with different amps for each speaker (two speakers here would be called a bi-amped system) and active crossovers controlling what frequency ranges are sent to each speaker. The single cabinet speakers typical of home systems and smaller PA's have more than one speaker in each with a passive crossover which splits the frequencies between the speakers. Here passive means that you have no control over how the split occurs - it is hard-wired into the speaker.

Equipment provided by the artist

As prices of out-board gear have declined, more artists are carrying with them a variety of equipment that they wish to use. Typically, this involves microphones, effects processors, in-ear monitor units and recording devices. As you will generally not have a great deal of time to setup and deal with this equipment before a performance (unless you have worked with the specific equipment previously), it is essential that you go through in a step-by-step procedure every change you might need to make to the equipment during the performance. This includes how to pause a recording device, how to mute and un-mute a microphone including wireless receivers, how to adjust an effects unit, etc. The objective is to avoid at all costs the possibility that a performance will have to be stopped so that the artist can show you how to do something. Some units have timed settings which can power them down after some period of non-use and you need to be able to bring them back to life quickly if they are needed and reset them as appropriate.

In general, you should assume that the artist is well aware of the appropriate application of the equipment they are carrying. However, do not be bashful about making suggestions for issues such as mic placement of large-diaphrahm condensors, clip-on mics for fiddles, internal mics on guitars. etc. If something doesn't sound right during sound check, it isn't going to magically fix itself during the show. Take the time necessary to try out various changes, particularly if the equipment has been obtained recently by the artist, or has been borrowed.