PA System Set Up

There are a number of things to consider when setting up a PA sound system, some should be fairly obvious, like making sure that the speakers are aimed into the audience and away from walls and ceilings. I say should, because a lot of people do not seem think of this! By adhering to a basic guide, most systems will sound good and be well useable in a wide range of situations. These basics are covered here...

Limit The Frequency Response

PA frequency response should be limited from between 30-50Hz 24dB/oct at the low end, and 15kHz @ 24dB/oct at the top. The bottom end limit can often be set at 60hz or even higher without being noticeable, especially if response is given a slight boost in response before roll off. Certainly do not set below 30Hz.

Like our homes, the lowest frequencies we can hear is determined by the size of the room. A large indoor venue may comfortably handle the very deep frequencies down to 20Hz as can an outdoor venue. Smaller venues will not offer a real capability of the lowest frequencies anyway and it is also important to remember that deep bass travels the easiest. This is often the most important factor in planning for minimum sound spill from the venue.

Rarely is there any worthwhile input at these frequencies and is only audible according to theory. In practice there is little need to have a system response below 40Hz and by limiting it will actually help to produce a clearer and louder sound.

Restricting the bottom-end like this removes much of the low end noise that takes up large amounts of amp power. All of the other signals ‘ride’ on the longer lower frequency waves. The main signal can easily be driven into compression or clipping by inaudible low frequencies stealing the headroom, often in a non-musically related way.

At the high end there is no need for a response above 15kHz. If you are uncertain of this, tune into a good FM radio broadcast, these have a steep analog filter cutting severely above 15kHz, (actually a notch centred at 19kHz). See if you can really notice anything lacking here!

Filtering the top-end removes a lot of spurious rubbish due to radio frequencies (RF) and other interference that can get into the system via instrument pick-ups and other on-stage equipment beyond the control of the PA and sound crew. It may seem surprising but there are times when this can bring a significant increase in clarity.

Mono The Deep Bass.

Turntables are renowned for their subsonics and warped records produce a prodigious amount of near DC waves (watch those hi-fi cones wobble when you play vinyl through them). Fortunately much of this is out of phase, and when combined into mono cancels out.

Bass is the most power hungry area of the spectrum and deep bass is not really directional, only the harmonics are. Running the bass in mono means that the load is always equally shared by both channels, and this is important not only for amp and speaker power considerations, but also the bass sound distribution in the venue.


This is a sticky subject over which there appears to be a lot of controversy.

Many rigs feature a third octave graphic equaliser (30-32 bands). More often or not this is because it is fashionable.

In hi-fi circles graphic equalisers are regarded as something people use to make their new hi-fi sound as crap as their old one! With rather to many PA systems it seems that it is a device for making your PA system sound like a really loud version of that crap stereo you listen to at home... Its other main use is to compensate for the fact that the sound system hasn’t been set up properly.

Even when used properly it is often to apply a gentle room correction that could more easily done with a simple two or three band tone control. It can be useful for limiting the response extremes as described above. This is much better done with proper 24dB/oct filters.

A properly set up system should not really need any EQ apart from maybe minor adjustments at the extremes for room compensation. Unless you have well trained hearing and really know what you are doing they are graphic EQ's best avoided on a front of house (FOH) PA sytem.

If you really need sophisticated EQ this is best done with one of the modern processors available. These usually come with a measuring microphone and a computerised adjustments that make life easy. As they feature crossovers and limiters as well they can be a very useful asset.

Graphic EQ is best employed on the monitor mix as an aid to feedback prevention. Here they can be very useful.


Speaker Layout

One thing I find so odd with many PA systems, especially small to medium sized rigs (and some of the larger rigs as well), is the lack of thought given to the actual sound distribution from the speakers in relation to the audience and the room.

First of all, any speakers producing midrange and top-end should really be above audience head height. This one simple aspect ensures that most people will be able to hear the music clearly. Not doing so means that all of the sound is thrown at the people at the front who absorb and block the sound to those behind. It doesn’t take much common sense to figure this but it often seems to be done!

These speakers should then be angled inward and if possible, slightly downward so that they point into the audience not straight at the back wall or over peoples heads. There are less reflections from ceilings and walls to muddle the sound and more of the speaker energy ends up going where intended. All speaker systems in fact, big and small, should angle in toward the audience for best effect.

It is surprising how large an audience can be covered with even small speakers this way.

In most small to medium sized venues the bass end can comfortably lie closer to the ground and can, in-fact, utilise the ground effect of increasing the depth and volume of the bass.

This suggests the use of low crossover points and the bass speakers considered more in the manner of sub bass. 150Hz – 250Hz is a good area to cross-over. This of course, makes it easy to mono the bass end around the same point that the sound directivity is becoming mono anyway.



Unless you are using ‘one box’ solutions you will need to separate the speaker frequency bands with a crossover

One box full range speakers usually use passive crossovers unless they have built in amps, (self powered). Multi-cabinet systems will normally use an active crossover system feeding separate power amps for each bands although passive filtering is common with small bullet tweeters at 5kHz and up.

Active crossover filters are commonly available for two, three and four way set-ups, with some units configurable for any of these. These generally have crossover slopes of 24dB per octave. Some cheap units may offer slopes of 12 or 18dB per octave, but they cannot be recommended. 24dB types are cheap enough to buy (or make) and as well as a steep slope, they have the advantage that the frequency bands are in phase at the crossover points, an important factor in a coherant sound (see Speaker Phase)

First of all, it cannot be stressed too highly toward setting up the bass-end to cover just the bass up to around 200Hz. Mono the incoming signals and that’s sorted. If you have extra sub bass as well, simply drive them as appropriate (often 80Hz downward) from the mono feed. Any phase differences arising from speaker alignment is less noticeable at this frequency due to the wavelength.

An ideal would be to have a single midrange speaker that has a smooth response from 200Hz up to about 5kHz. This is the most important area of the audio range, where the information and auditory clues are, where our hearing is most sensitive, . From 5kHz up then being covered by the treble horns (tweeters). Whilst this is not impossible to achieve, having good dispersion around 5kHz is not always easy, and the range has to be split.

800Hz-1.5kHz is a common area in the mid to crossover, but needs care to ensure an even frequency response, peaks or phase problems here sound unpleasant, a small dip in the response here is tolerable though. There are upper-mid horns which have responses down to below 1kHz and need tweeters above 8-10kHz. Some larger treble horns have responses down toward 1 or 2kHz although most of them really need some augmentation at the very top end. This is the province of the ‘bullet’ and slot tweeters.

The mid horn lower crossover frequency depends partly on horn cut off point, the one octave rule (see below) and the top end response of the speaker in the range below. Just either side of 1kHz is a common point. Remember that apart from constant directivity horns the directivity of a speaker narrows as the frequency rises. Running bass or mid range speakers up toward the top end of their range

Tweeter manufacturers will always provide recommended crossover details with their products (some like the Celestion HF series of bullets are available with their own built in filters). These are generally the best points used, unless this is lower than required. Never ever try and feed them signals below their working range, you will kill them very quickly.

The One Octave Rule

When choosing the lowest crossover point for most horns, use at least an octave above the actual horn -3dB cut off point. This avoids the famous ‘honky’ horn sound that dominates a badly set up horn loaded system.

On the subject of horns it should be noted that constant directivity horns only maintain their wide directivity up the frequency range, is at the expense of a falling frequency response. A corresponding boost is required to maintain a level frequency response.

Setting crossover levels

The output levels on the crossover should not be treated as tone controls... They should be set to equalise the sound output between the different ranges, so that the response is level between them.

If they are not set properly, there will be an un-natural jump at the crossover point, which cannot be compensated for with EQ on the mixing desk or whatever, (I have often seen an odd 'jump' on the 3rd/oct graphic EQ that is often to be found feeding the main PA, this clearly to compensate for a mis-adjusted crossover!).

Speaker Phase

Correct speaker phase is where the sound waves from the speakers are vibrating in the same direction at any one frequency, that is, all cones moving outward at the same time and all moving inward at the same time.

All speakers within a given band need to be in phase with each other and with adjacent bands at the crossover point. Ideally all of the speakers will have been wired properly, and so will all of the cables and connectors. The amps and outboard equipment should all be of matching phase. Ideally though, is not always what happens the real world.

There are test devices available that help with setting up phase within a speaker stack and these can be very useful on larger rigs. They feed a narrow triangular pulse (called a witches hat due to it’s shape), into the system. A microphone picks this up and usually either an oscilloscope or a special piece of test gear indicates the phase.

Without such equipment phasing can generally be done by ear:

Out of phase bass speakers are noticeable in particular due to a lack of bass at the centre line between them. When out of phase most of the bass cancels out. One out of phase bass speaker in the same stack is like turning two off them off!

One good way to check, is to place two bass speakers closely facing each other, feed them with a bass sound and swap the phase on one of them, the difference is immediately obvious.

The rest of the range is not quite so straight forward but it is still important to get it right. When mid-range and treble speakers are out of phase all sorts of other problems arise.

If the left/right mid/tops are out of phase it can give super wide stereo effect with a low centre image when fed with a stereo source. This plays havoc on the ears sound localisation and with mono sounds there is a confusion as to where exactly it is coming from..., not good.

Mid/tops in the same stack and out of phase produces all sorts of cancellation effects dependant on the listening angle. Consider the famous phaser and flanger effect units but without the sweeping up and down, it’s not so immediately obvious (because it is not sweeping), but has a distinct quality.

Phasing between adjacent frequency bands often causes confusion, sometimes switching the phase on the crossover doesn’t produce such obvious results, especially if you have to run round to the amp rack behind the speakers and then out to the front to listen as is often the case. The main effect on music is a disjointedness between the affected bands. Between the bass and mids for instance the bass will sound separated and doesn’t ‘gel’.

The other symptom can be the near or even complete loss of certain musical notes when they hit on or near the crossover frequency.

To determine phase in upper ranges (mid to tops) a wide band signal is required. A suitable piece of music will often suffice, something fairly busy and full. A alternative and perhaps preferable is what is known as ‘Pink Noise’. If you mixing desk cannot supply this then there are test CD’s with pink noise, various signal generators which provide pink and white noise and many synthesisers have a noise source that can be used.

Don’t have the volume too loud, put one ear fairly close to the front of the speakers and move your ear between one speaker and the other. The sound should evenly shift from one to the other. Obviously if you are comparing two frequency ranges, the lower range will be dominant in one and the higher in the other, so listen to the sound around the crossover point.

If the whole sound seems to jump from one to another around a particular point between the two speakers then they are not in phase. Two adjacent speakers in the same range can be checked in the same way.

Sometimes inverting the phase between two ranges still lacks gradual change and still jumps. This happens for two reasons mainly (presuming the crossover frequencies are set within the speakers operating band that is...). One is that the crossover is less than 24db/oct, in which case proper phase alignment is not possible, or the speakers need time aligning. In this case there is often one setting that is a little smoother than the other, if nothing else choose that setting.

If the crossover has time aligning delays then adjust these until you have a smooth transition, if that doesn’t work the swap phase and do it again. Minor time align adjustments can sometimes be done by simply moving the upper speaker backwards or forwards in relation to the one below, again using your ear to determine the most smooth transition.

It can be a good idea when setting up a system for testing purposes to deliberately switch some speakers out of phase and become aware of the effects. There are often times when a cable is swapped or hastily repaired, or a switch is accidentally operated and a speaker ends up out of phase. This invariably happens at an inopportune moment and somethin with the system sounds weird but not obvious. Having heard the out-of-phase effects previously makes it much easier to pinpoint.

Limiting and Compression

The limiter serves two main functions, one to protect the amps and speakers from over loads and the other when there is a requirement to limit the sound volume in the venue.

The main limiter is set to a fast action and hard limits, that is, cuts off any level lincrease above a certain limit, squashes the peak there and then, the signal bumps into an impassable ceiling. Setting this to over-limit, that is, actually turn the level down, or giving it a long release time, is a very good way of detering over-zeolous DJ's and sound engineers...

The secondary limiter can be set at a lower threshold and the parameters eased off a bit. The effect is more that of hard compression above the threshold. This acts both as general venue limiting but also with a little care can be used to boost apparant sound levels and add some punchiness. Do not overdo this though, anything other than a small enhancement can squash the life out of the music and have the opposite effect. The purpose here is similar to adding loudness compression to a studio recording when mastering.


A Most Valuable Tip

The single most useful thing you can do to get a good sound is to know what a good sound is...

Get a decent home stereo, something that can warrant the label ‘hi-fi’ not the usual tacky cheap set of boxes that shout at you over a thumping boomy bass.

Maybe consider a pair of low cost studio monitors, fantastic deals are available on these nowadays. If you love music (if you don’t then why are you reading this?), you will do yourself a favour that you will never regret!

If you have a personal reference of what a good sound is, your ears will tell you the rest.

Share |