Good sound system design fundumenatals.

As with any new venture the object is to get the best result with what we have at our disposal within the limits of our budget.

In the design of an audio system we need to provide a high level of clarity or legibility for all the recipients in our audience with an even distribution of sound throughout the auditorium. This must sometimes be done bearing certain visual restraints in mind as in the case of a church that has stringent anesthetic requirements as the result of the artist nature of the architecture. Once these restrictions are taken into account we may have to accept some severe compromises in the quality of the final result.

If we examine an ideal situation with no restrictions in a conventionally laid out hall and describe the techniques used in responsible design first we will gain an understanding of this prior to discussing environments with difficulties or limitations.

Assuming we have an area such as a conventional school assembly hall. We can overlook its large dimensions for the point of the exercise and consider the layout and the methods to be used irrespective of venue size.

Sitting in the center of the room space facing the riser or stage area we would see a loudspeaker to the left and another mirrored to the right and another central one slightly forward of the front edge of the riser and above it angled downwards and directed at the seated area two thirds of the way back towards the rear of the hall. The central speaker is designed to spread the sound evenly across the seated area rather than straight down the Centre.

The central speaker is fed by its own single channel amplifier, unless it is bi-amped in which case two amplifier channels are used, one for the high frequencies and one for the lower frequencies. An electronic crossover or filter is used to ensure that the correct frequencies are fed to each speaker component. This is done to improve the dynamic range. An electronic crossover is used rather than passive to improve the overall performance.

The central loud speaker is used to handle the speech and the two mounted to the sides are used to reproduce the music. However, it may also be used to control the balance of the music content by an experienced engineer in some circumstances

The reasons this layout is used is fairly easy to understand. The music speakers (situated left and right) are of course required to establish a stereophonic sound field and to do this two are required. We place these in an elevated position in order to achieve good "line of sight". Line of site means that our ears can 'see' the loudspeakers without any impediments or blockages. This means that all those seated in the hall are able 'see' the speakers because they are elevated. Doing this is the first step in getting a good result and a fundamental good design technique. The distance they are situated from the ceiling is a tradeoff between clarity and how well the sound reaches the rear of the building. This can be calculated mathematically. It is also worth mentioning that the music speakers do not necessarily need to be positioned on the rear wall and may in fact be brought forward and hung from the roof. This move will improve clarity but may cause a visual impairment that could very well upset your architect. Bringing them forward takes them out of the reverberant from field to some extent thus allowing the audience to hopefully hear more of the direct or near field and less sound of the reverberant field.

The clarity is improved because the loudspeakers are closer to the ears and so compete less with the reverberant field. The reverberant field is everywhere inside the room and it is caused by sound reflecting off smooth hard surfaces such as walls, floors and ceilings. The further away from your ears the speakers are, the louder the reverberation will sound by comparison. Therefore bringing the speakers closer to the listeners ears the clearer the sound will become. The closer the better but remember the other considerations. If you bring the speakers to close you will destroy the stereo sound field because a listener on the left will only hear the left hand side speaker and because it is so close he may not hear the right hand speaker. The left hand speaker’s sound pressure will mask the speaker on the right. This is an example of the kind of compromises the sound designer must face and why it is a good idea to employ one. They know what they can get away with. Also the visual aspects need to be considered as well. Speakers sitting away from the walls make look silly if they are too obvious unfortunately.

Angling the loud speakers downwards has two big advantages. Firstly we need to be aware that speakers generally sound at their best on axis or when we are looking directly at them rather than from the side. (Off axis.) This is because high frequencies tend to be more directional than middle and low frequencies. Because the sound is in general louder in front of the speakers we want to aim them at the listeners as accurately as possible. Everyone cannot of course be on axis so here we have yet another compromise. Sometimes more than one speaker per side is used to offset this effect and they are panned or angled to bring more of the audience into the direct field.

 The second and most important reason we aim the speakers downwards is that if we are steering them at the seated area only, then we are directing most of the sound at an absorbent surface and so are therefore minimising the reverberation effect that we already know is caused by reflections. These concerns are very important though basic considerations of good audio design technique.

The central voice speakers are located thus to eliminate a characteristic known as phase shift. This is caused when the same sound comes from two different locations simultaneously. If we are seated "off center" the sound from the speakers reaches our ears at different times and are said to be out of phase with each other. This so-called phase shift causes a slight muddiness or blurring of the sound quality and we do not want this aberration to occur if at all possible. A single central speaker tends ot help eliminate this.

By following the suggestions outlined above, we are hopefully well on the way to constructing a good audio reinforcement system but we have been working in an ideal situation and this is not always the case of course. There are an infinite number of different internal shapes found in buildings and we must consider the way we are going to handle as many as possible.

The best way of doing this, is for me to leave the basic shape the same but to start stretching the dimensions somewhat. All the rules outlined above still apply of course but we need to further discuss the various compromises and how best to handle some these tricky situations.

A Longer Rectangle

Let’s consider a longer shaped hall. It could be an old style cathedral that is three times as long as it is wide for example. In this case everything described above can remain the same but the sound will get rather weak towards the rear of the building and we must raise its level somehow. The answer is simple and all we need to do is add two or more pairs of loudspeakers on the sidewalls strategically placed further down the hall. This in itself is fairly straight forward but we have a serious problem already described above and this is phase shift again. The people sitting in front of the additional loudspeakers are going to hear the sound from both pairs (front and rear) and we have also forgotten the speech system.

We use the same technique on the rear speakers we used on the front and angle them inwards and downwards. This is a good start but how do we get rid of the nasty echo effect as the sound traveling down the room is now coming from two different locations simultaneously as discussed earlier.

 We add an electronic delay unit to the rear speakers to line up the phase again with the front speaker system. Delay lines, as they are called cost money but are worth every cent of it. The result is near perfection. You have to hear the effect they have in order to believe it. It’s quite amazing. The echo should disappear entirely.

A delay line unit is a 'one rack unit' (1U) electronic device that is patched or inserted  between the mixers output and the power amplifier that stores the sound data momentarily and sends it out again a short time later. This is measured in milli seconds and the further away the secondary speakers are, the longer this delay needs to be. Sometimes in the case of a really elongated shaped room, more than one pair of delay speakers is used.

In this situation two different delay times must be used and most delay lines have extra outputs to accommodate this requirement.

As we add music speakers so we must add more speech speakers in the same manner. Their location remains as per the front speakers, above and in the center. We must use a delay line for these speakers as well.

A Higher Stud

Consider the same building but now with a higher stud. Not much will change but the reverberation characteristics will alter and will probably increase substantially. This may mean more speakers and also bringing them closer to the audience. A lot depends on the acoustics themselves and on the amount of absorbent materials present. This can also determine whether or not the use of horns is prudent. A high stud and poor acoustics are usually very good reasons to consider using horns. The modern "line array" speaker system concept is often used these days in this type of acoustic environment.

A Wide Building

If we come back to our original layout but assume the building is much wider rather than deeper we will need to alter the design yet again to make things work reasonably well.

Opinions on how to go about this may vary between the various design professionals and not everyone will agree with my approach. With stereo systems we have an area that we refer to as the sweet spot. This region is at the point where the left / right coverage by the pair of speakers is ideal. This is of course dead center or close to it. The sweet spot decreases in quality the further to the left or the right we are sitting or standing. It would be fair to say that about one third of the total area covered by a pair of speakers provides a good stereo sound field so we need to extend this area as much as possible. We can do this simply by adding one or more pairs of speakers. The number we add will depend on the distances involved and the seating layout. A loudspeaker of conventional design has a coverage angle of approximately 90 degrees. The further away it is located the wider the area it covers but the further away it is mounted the more power we need and the less clear it becomes as it competes with the reverberation. The number of speakers used will therefore depend largely on the seating layout the positions available for hanging them and the distance they will need to be away from the audience.

The vocal speaker will almost certainly need to be a horn loaded design in order to get the throw needed to reach back or project the long distances involved. This would mean a cluster of horns set up in such a way to ensure an even coverage with a minimum of phase cancelation at the overlap. Horns do not cover all frequencies so thought must be given to how we are going to reproduce the lower frequencies as well. This is done by cone speakers and these are usually 12, 15 or even 18 inches in diameter, depending on the crossover frequency the designer elects to go with. The larger the horn the lower the cross over frequency can be, but large horns tend to be very costly and take up a lot of valuable space and this is not always possible. The cone speakers are usually mounted in bass reflex enclosures that do not have the same efficiency as horns or the same directivity but this does not matter too much fortunately. The clarity of speech is determined largely by the mid high frequencies and these are handled by the horns. Because horns are much more efficient than cone type speakers we will need a way to drop their level or efficiency down to match the output of the cone speakers. If we use two amplifier channels this is not a problem because we just turn down the volume level of one channel on our electronic crossover. With a passive crossover we will need to rely on a switched resistive pad network to do the job and these are a common source of failure in two-way systems. They however  an economic way of getting the job done, just not as well.

As I stated earlier there are other ways of coming at this problem but this is my way and it works very well.

A Very Low Stud

 Low studs place severe limitations on the sound designer as line of site becomes a problem especially if items such as lights or video projectors are going to complicate the issue. These areas make it difficult to do anything really creative as space becomes such an issue and large bass speakers are not possible either. The common method of doing these rooms is to use ceiling mounted speakers that face directly downwards. This is called a distributed sound system and they are very cheap if cheap is what you want. Low-level sound companies use this method as generally speaking they have no imagination and tend to have a bash bash get the cash attitude. Distributed sound systems are almost always mono in nature and very boring all round really. Speakers that face directly downwards create lots of little hot spots and do very little else. If the room is wide and we have no walls to work with sometimes a distributed system is the only possible way of doing the job as using cabinet type speakers and angling them downwards but towards the rear from the ceiling is seldom visually acceptable.

If the room is rectangular, wall-mounted enclosures can work very well and a stereo sound field can still be created. The main problem with a low stud is that the loudspeaker cannot be elevated very far and this means it will be difficult to achieve a good front to back ratio. The people seated near the speakers may get blasted while those further away may not get enough sound. Here we have yet another compromise.

Odd Shaped Rooms

The way we handle speaker layouts in rooms that are non-symmetrical depends to a large extent on the application for which they will be used.

Straightforward background and or low-level speech situations are easily sorted out with small wall mounted enclosures carefully placed to maximize coverage but avoiding intrusion. In these situations we may not be concerned at all with presenting a stereo Sound field, instead preferring to concentrate on a gentle background subtlety that provides a relaxing ambience and the ability to handle voice announcements as well. Speakers are mounted on wall surfaces and inside corners while avoiding outside corners and areas where people are likely to prefer silence such as reception areas or telephones and also making sure that are out of easy reach of people who may try to steal them. The wall speakers are angled on their brackets to provide the best coverage.

In entertainment venues we may however decide to provide a stereo sound field. This is usually done by dividing the overall area into sections that are associated with the seating layout. If booths are included in the design this can make things much easier as they provide natural square shapes that lend themselves to stereo sound, so if we get the speakers correctly positioned we can provide a reasonably large sweet spot. Speaker’s always sound best when you are looking into them rather than when you are hearing the sound from behind or from the side. This is because high frequencies do not flow around corners very well.

Sometimes we can greatly enhance the performance of such systems by the addition of a small hidden sub woofer to bring up the bass level for frequencies below say 100 Hz. It is hard to produce these low frequencies using just wall speakers on their own.

High-Level Systems

High-level systems mean high volume levels so even more attention needs to be paid to location. If care is not taken with these powerful devices they can cause hearing damage and this is not a good thing unless you are an ear doctor.

High level systems are really only used in night clubs or sports arenas or large venues that require live acts like Rock Bands. They should always be placed well above head height and certainly not directed at people who are standing close to them. These more powerful speakers are usually used in entertainment situations in conjunction with an even more powerful dance floor drive system. It is not possible to get reliability out of small low powered speakers in this situation as their sound output tends to be drowned out by the much louder system over the dance floor. DJs will often turn them up so they can be heard above the dance floor system and this will often result in serious damage to them due to over powering. Perimeter systems should not be able to be turned up by the master volume control on the disco mixer. Including this simple safety design precaution can save a small fortune in repair costs.

Placing speakers is definitely not white man’s magic but it does get a lot easier with experience. Remember that you can only get 90 degrees of coverage per speaker. Avoid placing them to close to people. Get them high and angle them downwards if possible. There are no hard and fast rules, we generally just focus on getting the sound spread as even as possible.           

Impedance Matching

There are two ways of matching the impedance when connecting a large number of loudspeakers up to an amplifier. They can be connected in series parallel or a transformer can be used. Either method will work well but the transformer mounted on each speaker saves on wasted power due to losses if the cable lengths are going to be substantial. The downside is that transformers can cause colouration and of course they add cost. If the power amplifier has an output power of 200 watts into a 4 ohm load you can connect up to 50 loudspeakers but you will only have four watts of power available for each one. The main output transformer that is attached to the amplifiers output must have a primary impedance of 4 ohms and a power rating of at least 200 watts. It will provide 100 volts and hence the term, 100 volt line system. Each little speaker transformer has a primary that will draw 4 watts of power and reflect this into its secondary which will be 8 ohms and deliver 4 watts of power to the voice coil of the loudspeaker.

The series parallel method has better sound quality and better dynamic range but there is a loss of power due to the resistance of the cable. Speakers are configured in groups that are wired in series or parallel and combined to match the amplifiers impedance by connecting the groups of speakers in series or parallel back at the amplifier. Any mismatches can be corrected by using wire wound resistors to increase the series resistance and give a better match so the amplifier does not become overloaded and overheat.  If you have power to spare and don’t mind spending a little more money on cable this is the better way to go in my opinion. Both these methods will work well with either ceiling mounted or wall mounting speaker enclosures.

Good sound engineers with some system design experience can walk into a building and get a feel for it almost straight away. They do not need to do a whole lot of mathematical calculations in order to get a great result. They can tell by 'feeling' the atmosphere what will work and what will not. Sound quality is very dependent on the acoustics of a room and this will change radically from when the room is full to when it is empty. At least one thing is certain, the acoustics will only improve as it fills up with people.

Large expensive buildings often need large expensive sound systems especially if high-level sound is required. The way these are designed and installed are largely dependent on what they are going to be used for. Unfortunately there is seldom a lot of room for hit and miss and mistakes can be very expensive.

Do not take on the responsibility for an expensive installation unless you are sure you know it is within your capabilities or unless you are prepared to take responsibility for any mistakes you may make. If in doubt call in the experts, don’t risk it yourself.

Dance Floor Systems

Dance floor systems are potentially very loud and must also sound very clear. They are usually at least a two way design and should preferably be a three way system. The mid high speakers should be flown in the best position above the dance floor and one or more sub woofers should be at floor level. I have often seen the mid high satellite speakers flown to close to the dance floor which is poor design indeed. This results in a small sweet spot and the people in the corners getting blasted. The satellite speakers should be mounted well outside the actual area of the dance floor and steered inwards and slightly downwards towards the opposite corners. The position of the subs is not as critical as bass tends to be omni directional in nature and so subs can be placed anywhere convenient as long as they face the dance floor and are not too far away from it. I have even mounted them in the ceiling facing directly downwards and this has worked surprisingly well.

Building a Robust Three Way Systems

• Add a single rack strip of horn fuses at the top of and at the front of the amp rack. These will mean never having to climb a ladder to get to the high frequency horn fuse holders.
• Do not use line connectors in the speaker cables. Substitute these for terminal blocks. Much more reliable.
• Make sure you construct the amp rack so you can get at the rear of it. Mount it on wheels if you need to but do not fix it in place.
• Leave one U space between each power amp and add an extra fan in the back of the rack. This greatly improves the cooling.
• Use amps with more power than is needed so distortion is much less likely to occur. Distortion can easily damage loudspeakers. Headroom is like having an engine that never over revs.
• Place the satellite speakers outside the edge of the dance floor. This gives much better spatiality to the sound and eliminates blasting.
• Horn loaded subs greatly increase the reliability of the eighteen-inch woofers. Horns have a characteristic loading effect that seems to protect the speakers from over excursion. We have never needed to replace damaged sub woofer drivers in horn enclosures ever.
• Using the 3 way design means that no passive crossovers are necessary. These are a source of major unreliability in speaker enclosures.
• Use heavy-duty speaker cables so that image imbalance does not occur. No wasted power across feeble speaker cables either.
• Use only power amps that are known to have very good cooling, large heat sinks and reliable fans. Clean the fans often if possible.
• Never use tweeters for the high frequency reproduction. Use only one or two inch horns or variants in between.
• Make sure the mixer selected has accurate output meters. These will give the D J's a good indication as to where clipping will occur.
• Find a way of limiting the DJ's microphone so that feedback cannot occur.
• Use a mixer that has a second output for a remote zone. This will isolate the perimeter system from the master volume control so when the DJ cranks up the volume he will not damage the perimeter speaker system.
• Use an electronic crossover with at least a 24 db per octave roll off slopes, as this gives good protection against power factor damage to the top horns.
• Place a rack cover protection plate over the crossover and the equaliser. These will help stop unauthorised tampering which is a very common cause of equipment damage due to incorrect settings.

There are several other techniques we use that fall into the category of trade secrets and these will not be divulged here but the above outline of good design practice will take you along way along the road to a good robust dance floor system that should offer several years of reliable service that you may otherwise not achieve. Given the choice and the budget I would never offer my clients a two way system and I would always try to build a system that was more powerful than actually required. One of the biggest nightclub systems we ever built was of course installed in one of the country’s biggest nightclubs, which was the Coliseum Club in Manukau. In two years we only got called out once and that was because the DJ had the band DJ changeover switch in the wrong position!

In a ten year marathon between 1980 and 1990 we did two hundred sound installations in entertainment venues right across NZ. Most of these were done using equipment we had manufactured ourselves, although I must confess the many dozens of SL1200 turntables we used were all imported.

Considerations for Church Systems

Church systems invariably involve visual issues and architects are usually very fussy and are rather  unsympathetic to the needs of the sound engineer. These problems can be avoided if only the architect were to involve the sound designer from the outset. Sadly this seldom happens and as the audio is usually the last consideration it inevitably is seen as an intrusion on the original design conception.  Positioning the central voice speaker in the ideal location is therefore likely to be a very difficult accomplishment if the architect has any say in the matter. This often results in a direct conflict between the objectives of the two parties who are both intent in realising their individual objectives. One wants good looks, the other wants good sound and combining the two is not always easy. In a completed building it is highly unlikely that loudspeaker enclosures are ever going to actually enhance the appearance of the interior of a modern church. The challenge is to make their presence as subtle as possible. This can be done by blending merging and hiding. Blending is done by making the speaker enclosures the same colour as the background directly behind them. Merging is done by making their shape fits harmoniously with what they are attached to such as a pillar, a molding or a sculptural form. Hiding is done by mounting a speaker inside a wall or stage front so that the only part of it visible is its grill. This is usually done with front fills, side fills or sub woofers and it can work very well. Depending on the layout this technique can also work well with the left / right music speakers.

If a central vocal speaker is to be used with a high stud, which is usually the case in a modern church building, making it disappear is a big asks indeed. The other issue here can be line of sight if a projector is going to be used. In this situation the need for a central speaker is often overlooked and the music speakers are used to reproduce the vocals also. The central speaker would be situated effectively in mid air above the pulpit like a UFO floating in space. If it were hung by a pair of thin cables that doubled as the speaker wire it might just be possible to get away with this approach if the colour of the enclosure allowed it to blend into the background and the design of the enclosure matched the architecture in a harmonious manner. Churches often have facades and conveniently placed panels that allow for the concealment of loudspeakers, but nothing beats considering the incorporation of the audio system prior to the confirmation of the buildings final design.

Considerations for School Assembly Hall Systems

Apart from the usual acoustic issues, assembly halls are good applications for the sound designer to place a well designed system into. Their natural layout is very conducive to the central speech and left / right music playback design. A rebated stage design makes this easy, because fixing for the speech system is ready made and the stage wings are the obvious place to mount the music playback system. This all changes however if we are talking about a more modern type of open stage or large moveable riser. Some schools even require the stage to be portable. The formula remains the same however, with the speech system in front of and above the stage edge and the music playback speakers to the left and right of the front edge of the stage.

If the hall is a long rectangle further reinforcement may be required half way down the sides.

If a mobile stage is part of the concept, a central speech system may not be possible unless the layout is returned to the preferred position for formal use.

Stand speakers are generally the best approach to sound reinforcement in mobile stage situations because they follow the riser around and can be used for other applications away from the hall when the hall has been returned to its standard layout.

Theatre and Opera Houses

• The main sound cue and music system. (Left / right stereo sound cue and music replay.)
• The secondary sound cue system. (As above but speakers placed throughout the void.)
• The central speech system. (Central horn cluster above the proscenium.)
• The mezzanine system. (A repeat of all the main systems for the mezzanine floor.)
• The fold back system. (Floor monitors placed on or around the stage.)
• The stage side fill system. (Composite speakers mounted on the stage walls for music.)
• The foyer system. (Perimeter speakers for background music.)
• The show feed system. (A system so the artists can hear the stage from the dressing rooms.)
• The comets system. (The intercom for the production crews communications.)

The above list is indicative only and not fixed or compulsory and some are even more complex but due to budget restraints theatres are generally underdone.

Presenting an advanced description of these systems and their setup would be an exercise in futility as theatre designs vary so much from one to the other. In such venues a consultant is invariably involved in all phases of the design and the associated technology.

Anyone interested in further information on this topic are more than welcome to contact me. Please call Paul Johansen on PH: 444 8776 or email him at [email protected]