Mixing Live Events

In spite of far-reaching advances in audio technology in recent years, many concerts still sound abysmal. However, while s/he is often viewed as the sole culprit - and will sometimes turn out to be the main culprit - a live sound-engineer may have to cope with a number of problems that are nowhere near the mixer.

If you want to do your best in what can sometimes be quite difficult circumstances, you will do well to understand what can make things easier (or harder), and realise that your job doesn't start and finish at one end of the multicore. You may have to deal with

  1. The room. Rooms - even the best-designed concert spaces - seldom make amplified music sound better. The best-designed concert spaces, as far as the live sound-engineer is concerned, are those which don't make it sound much worse.

    One common myth is that you can ‘EQ the room’. Rooms have their own resonances (largely due to room dimensions and parallel surfaces), as well as surfaces that reflect some frequencies more than others. Sometimes there are areas within the room (the stage area, the area under the stage, the balcony...) that add their own resonant and reflective characteristics. If the stage area is effectively a wooden box, performers may comment that their monitors sound ‘boxy’.

    EQ can have some limited success in reducing the effects of reflection, but is much less effective in dealing with room resonance. The reason for this is that the room will tend to resonate at its own resonant frequencies regardless of the frequency of the original sound. If you clap close to an acoustic guitar, the strings and body will resonate. If you shout in a tunnel, the tunnel will resonate. In both cases, you can change the frequency content of the original sound (you can shout at the guitar and clap in the tunnel) without having much effect on the result. THE FREQUENCY AT WHICH A RESONATING BODY RESONATES IS LARGELY INDEPENDENT OF THE FREQUENCY THAT EXCITES IT.

    If a room has strong resonance at 200 Hz, you can cut 200 Hz by 12 dB on all your graphics without having much effect on the room. In contrast, it will have a substantial - largely destructive - effect on signal content.

    There may still be good reason to make fairly heavy cuts at resonant frequencies: if room resonance combines with instrument or microphone sensitivity, the system will be much more prone to feedback (although a notch-filter or parametric EQ will usually have more effect here than a graphic or non-parametric channel EQ).

    Assuming you can't actually rebuild or resurface the room, however, the best way to minimise the problem is by finding optimum positions for the speakers. Position and angle the speakers to minimise mid- and high-frequency reflection, and try to place subs to avoid low-frequency resonance. Keeping mid- and high-frequency reflections off the walls can be managed fairly well by eye (although protractors and lasers can also help). Damping low-frequency resonance works in a similar way to playing harmonics on a guitar or other stringed instrument: if you stop a string from moving at its centre, you silence the main resonant (fundamental) frequency. Also, if you change the point at which you pluck the string, you change the tone (i.e. you affect the frequency balance). Moving the bass speakers changes the position at which they excite room resonances, and - much like plucking a string at a different point - usually has a noticeable effect. In some mirrored positions, a pair of bass speakers will also act in opposite phase at room-width resonant frequencies. As with guitar harmonics, calculation of the complicated interplay between string density, tension and length is largely irrelevant: you always find the octave harmonic at half the string length, regardless of the actual string length or fundamental frequency. Similarly, optimum positions for a pair of bass speakers usually occur at 1/8 and 1/4 room-width positions, regardless of the actual room width or resonant frequency. A couple of minutes with a tape measure can make a noticeable difference.

  2. Multiple sound paths. Any multiple sound paths will have unwanted consequences for phase-relationships, and should be avoided wherever possible. As well as obvious candidates (lots of speaker cabinets, or multiple paths from source to listener brought about by room reflections), there can be multiple paths at source. Wherever it is practically possible, avoid using pickups AND microphones on the same instrument, or several microphones on one instrument. On instruments which produce sound over a large area (like grand pianos or marimbas) more than one microphone may be desirable, but even here use no more than two, and try to make sure they remain in a fixed position with respect to the instrument and each other. Where you use a pickup and a microphone on one instrument - for the acoustic guitar, for example - be prepared for some tonal colouration that cannot be cleaned up with the EQ. It may not sound natural through the pickup alone, but adding a microphone may not help.
  3. Instruments, voices and microphones. If it sounds horrible unamplified in free air, it will probably still sound pretty awful whatever the engineer does to it at the desk. A lot of instruments, voices, pickups, and microphones have colourations of their own. The best a sound-engineer can do with these is damage-limitation. Too much EQ generally sounds unnatural, and a lot of cut - even at frequencies that are much too prominent to begin with - can suck all the body out of an instrument's sound. The ‘quick fix’ that works best is to find the most obvious nasties and cut them a bit:
    • Boost the mid (about 3 to 4 o'clock on the knob)
    • Use the mid sweep to find the nasty
    • Change strong boost to modest cut (10 to 11 o'clock)

    Needing more cut than about 9 o'clock means you may need to think about other solutions than EQ. Microphone placement can make a lot of difference.

  4. The band. One of the main problems a band can cause is making too much noise on stage. If on-stage levels are too high, the front-of-house sound is much harder to control. Very big stages in very big auditoriums make the problem a bit easier to manage, but even on large stages spill from amplifiers and drums into vocal mics (or from monitors into every microphone and pickup) can severely cramp a front-of-house sound-engineer's ability to control what the audience hears.

    If you are a musician who wants the audience - rather than the band - to get the benefit of your performance:

    • Learn to manage without monitors. If you get monitors, it is a bonus, and you won't need silly amounts of monitor level. If you really can't manage without them, get a good in-ear system and get used to using it.
    • Learn to make the sound you want at modest volume. You can now get effects-units that will give you feedback-assisted sustain from your guitar using a 10W practice amp. You can get drums to sound good without denting the skins (and most professional drum teachers will tell you that if you are denting the skins you are using the sticks badly).
    • Instead of turning up whatever you can't hear, turn down whatever you can hear until the thing you need to hear is loud enough as it is. The ‘turn it down’ solution means you try to work with the lowest level you can. The ‘turn it up’ solution means everything gets louder and louder until the weakest link - often the thing you most need to hear - runs out of headroom.
    • Learn to trust the front-of-house sound-engineer. If - like many musicians - you have already had the benefit of working with ‘sound-engineers’ who managed to make you sound abysmal, either recruit an engineer to tour with the band or (if that is impractical) try to ensure that any venue on your tour engages a professional engineer you will get on with. If you liked the guy who was there last time you came, try to make sure you get him next time as well. Similarly, if you found the last one dreadful, try to make sure you don't get him again (this advice isn't sexist: if - good or bad - it was her rather than him, get or avoid her).
    • Learn to communicate with the monitor engineer. What you have in your monitor should suit you, and only you. If you can only get what you want by explaining it to someone else first, you need to describe it as clearly and unambiguously as possible. If you don't know how to explain it, don't blame the engineer. S/he isn't a mind-reader.

    If you are a front-of-house or monitor engineer, this probably didn't need much explaining. However, you'll stand more chance of explaining it to the band if you are polite and well-informed. Try not to create the impression that you object to the way they play their music.

    If the sound that members of the band want in their monitor mixes needs outrageous channel EQ, consider using an extra channel for each of the main monitor components (see below).

  5. Desk and stage layout. Often the band has a ‘usual’ stage layout. Most of the time, this will work well enough, but
    • Try not to put trumpets and trombones at the back. There, they will spill into vocal mics (which are almost always at the front). Put horns at the front of stage at one side or the other if you can.
    • Try to ‘aim’ every sound source (guitar combos, monitors) to avoid vocal microphones. Keeping the lead vocal mic away from the drum kit helps.

    Some bands also provide a channel list. This is certainly useful information, but remember that if you are the only sound-engineer you are not obliged to put everything in their stated channel order. If, on the other hand, you are working with another engineer (on monitors while s/he handles front-of-house, for example), it will only confuse everyone if you try to work in a different order.

    Although it is ultimately a matter of individual preference, there are common protocols for channel layout. The channels are often laid out (from left to right) in a similar order to that given for sound-check order on the sound-check page. Many sound-engineers regularly have to work with different bands (not all having the same stage layout) and different desks, and it makes it much easier to find the right channel in a hurry in poor light if the kick drum is always on channel 1. As a slight departure, it is also quite common to place vocal channels - the most important part of the mix - immediately to the left or right of the mixer's master section.

    It often also helps to use Y-leads or Z-leads to split a signal onto more than one channel. This can be useful for

    • Running monitor and front-of-house mixes from one desk. You can use one channel for the main mix and another for the monitor feed. This gives independent gain & EQ control for each, so you can EQ the monitor feed for the musician's personal taste without compromising the front-of-house sound. Also, having got exactly the monitor sound the very fussy lead singer wanted, you don't have to touch it when you make changes to the the front-of-house gain or EQ.
    • Using the same microphone (or DI) for two or more different acts. You don't have to change anything on stage. You set up the gain, EQ & monitor mix for the first act on one channel. Then mute that channel, and set up the gain, EQ and monitor mix for the second act on another channel. Note, however, that connecting a microphone to two inputs halves the load impedance. Most mics will drive two channels without obvious problems, but some mics and DI boxes will struggle with more than two channels (this is most noticeable at lower frequencies: if the load impedance is too low, it will start to sound ‘thin’).

    On many digital desks you can route the same input to more than one channel without needing external splitters, but the input preamplifier gain will then become a shared setting unless you disable it on some or all of the shared channels.

    How you lay out channels for multiple acts is more open to individual interpretation. However, as long as you have some PVC tape and something to write on it with, you can easily:

    • Use white PVC and a Sharpie to mark channel positions on the desk. It is much easier to go to LEAD VOX than remember what channel you put it on. Most modern digital desks recognise this by allowing you both to name a channel, and order and group the channels for your own preference;
    • Use coloured PVC or coloured Sharpies (or your digital mixer's facilities) to identify obvious groups (e.g. Band 1 and Band 2, or Drums, Keys, Guitars, Vocals).
  6. Sound-engineers. Sometimes you may be just one of a number of engineers dealing with a production, and may not be the one with the final say on global things like system layout or house EQ. Also, you may have to work with one or more other engineers on the same part of the production. It is common on larger productions to have separate desks (requiring an engineer for each) for monitor and front-of-house mixes. On some of the biggest productions - running upwards of 60 channels - there may be more than one engineer to a single desk.

    The common aim of front-of-house engineers is fidelity (the origin of the word ‘hi-fi’ was High Fidelity). You want everything louder, but you don't want anything to sound different. If you can, compare the original sound with what the sound system is doing to it by lowering the channel fader to the point where the effect of the PA is scarcely audible. The tone of the original sound shouldn't change much as you raise the fader and the loudspeaker system takes over (obviously this doesn't work in larger auditoriums).

    Important factors in getting the band to sound right are:

    1. Get the gains right. All modern PA equipment is designed to operate within a limited range of signal voltages, and if all signals are at similar peak and average levels they will sit together reasonably well with the faders in a fairly straight line. Vocal intelligibility is the most important feature, so a few extra decibels on vocal channels is usual and desirable. See the section on gain structure for a little more detail.
    2. Realise your limitations. You won't turn a £100 violin into a Stradivarius by twiddling a couple of knobs. There is no magic knob, magic ear, magic finger, or 30-second treatment that will make a £100 violin and a Stradivarius equal, so don't expect an equaliser to do it. Get the best you can with what is available. Don't imagine you can turn what is available into something it isn't. Fidelity means a horrible noise sounds like a horrible noise.
    3. EQ as little as possible (don't turn a Stradivarius into a £100 violin).

      ‘Equaliser’ is a generic term for a range of frequency-altering audio filters. The term ‘equaliser’ came about from early use: frequency-altering filters were used to ‘equalise’ sound-systems that did not reproduce all frequencies equally.

      Modern sound-systems have come a very long way from those early days. Good microphones, microphone preamplifiers, mixer EQ circuits, power amplifiers and loudspeakers have usually been developed by very well-qualified professionals, using a vast array of professional equipment (and often fairly mind-boggling research and development budgets). A state-of-the-art loudspeaker system (developed by a team of audio engineers with very good hearing and post-graduate degrees) probably doesn't need radical equalisation. Whilst there is certainly a degree of personal taste involved, you shouldn't start with the assumption that you can get it to sound better in thirty seconds by making the graphic EQ look like a relief-map of the Himalayas.

      Even where they are needed, equalisers themselves introduce artefacts to the original sound. It isn't physically possible to change one part of the frequency spectrum without changing phase-relationships near the area of change (as well as - in cheaper units - some way from it). EQ doesn't just equalise: it adds its own colouration. That is one reason why most EQs have a distinct ‘sound’.

      Having a good ear for frequency (knowing what different areas of audio frequency sound like, and how to identify them) is an important factor in using EQ productively, as a little EQ in the right place is much more effective than a lot of it in the wrong place. If you need EQ at all, cut, don't boost. Rolling some bass off everything except bass instruments (e.g. kick drum & bass guitar) can improve clarity and separation.

    4. Get it done in the time available. You don't have the luxury of taking all day to tweak that one channel. The show has to happen on time.
    5. Don't do anything big or sudden. Other than during the sound-check, move gains, faders and EQ settings so slowly that unless you knew they were changing you probably wouldn't notice any difference in the sound. Even in the sound-check it is best if you act in moderation. If you need to un-mute something, check it doesn't have a huge signal on it before you hit the un-mute button (and if in doubt, lower the gain, fader or auxiliary settings first, then bring everything up slowly after you have un-muted it). If musicians on stage want more or less of something in their monitor mix, add or subtract it in small increments (if the knob is at one o'clock, don't go further than about twelve o'clock or two o'clock until they ask you for even more or even less).
    6. Listen to the lead vocals. If you can't hear the words, nothing else matters. If you can't make the words intelligible by increasing the vocal level, gradually reduce the level of everything else (a little of each channel, group or VCA at a time), until you can hear them. If the backline instruments are loud enough without the PA, take them out of the mix altogether.
    7. Keep your ears open. Throughout the show, listen out for booming, ringing, or any untoward sounds. If one of your drivers has a ripped cone you should be the first person to notice. Similarly, you should know before anyone asks you that the bass-player's strings are rattling, and ‘that buzzing noise’ isn't being caused by your mixer, power amps or loudspeakers.
    8. Keep your eyes open (but also read the next paragraph). Watch the band for signs they mean you to act on, or for changes of instrument (you may need to mute the fiddle and un-mute the mandolin; otherwise the one will start feeding back when s/he puts it down, and the other won't make any noise when s/he picks it up). Also, keep your eyes open for signs on the console that something needs attention (channel peak lights - or, if you are lucky enough to have them, channel meters - showing excessive signal levels are one common sign). If you need to adjust channel gains and use the front-of-house desk for the monitors, remember that changing the gain will change the monitor level. If you bring one up as you bring the other down (or vice versa) it may enable you to adjust it without the musicians becoming aware of it. Use your headphones and the auxiliary AFL: if YOU can't hear it change, you can be fairly sure the band won't.
    9. Close your eyes and just listen for a few seconds from time to time. You might be surprised how much better you can hear what is going on with your eyes shut.
    10. Look after your hearing. Get a sound-pressure meter, and be aware of sound-pressure levels at the desk. If you regularly engineer loud shows (even 100 dB - which isn't remotely earth-shattering - will damage your hearing if you do it every night), get some good ear plugs and use them. There are several makes that allow all frequencies through at restricted volume (so you can still hear everything, but at lower levels). If you wait until you start to suffer hearing loss, your career as a sound-engineer won't last long.

      If overall levels get too high - 110 dB is very loud - the whole mix can degenerate into an indistinct mush of noise. Apart from looking after your own hearing, remember that your audience can suffer hearing damage, and that only a few diehard adolescents who have probably mistaken testosterone for intelligence actually prefer it to be deafening. Keep overall levels sensible. It is only a matter of time before the ever-creeping tide of state intervention will force it on you anyway.

If you would like more information than is available here, or are interested in combining theory with practice, you could consider enrolling on a live sound-engineering course.

[Top of Page]