The Birth of the Array: Part Two
Last issue we looked at early attempts at propagating sound coverage to large assembled masses. The Greeks were the masters of this technology, but by the 1900’s the centre of gravity was moving to the US - via Germany.
The Last 100 Years
Western General Logo
The Altec A7 in the early 50s. A milestone
Modern sound reinforcement began in 1915 thanks to political necessity when a telephone receiver and a phonographic horn were used to announce a USA presidential inauguration to a large crowd. The main rush of developments came from the motion picture industry, eager to make the silent movies talk with synchronised pre recorder sound cover an auditorium full of people.
Under the brand name Western Electric, the early horn-loaded public address and cinema systems appeared. The brand morphed into the company Altec, which “wrote the book” technically and practically on all aspects of component and horn design. Most speaker systems today rely on direct or partial utilisation of Altec Technology. Very little had been ‘invented’ since. Most modern developments have improved application and materials based on old Altec designs. This applies not just to box designs, but also speaker components and horn flares.
This competitive pressure within the growth of early cinema saw the development of a number of products that are still with us today in almost the same guise as their original designs. The early cinema history period saw the development of the following:
- The Compression Driver and Phasing Plug
- The Radial and Sectorial Horn
- The Amplifier
- The Cardioid and Condenser Microphone
- Folded Low-Frequency Horns
- Stereophonic sound
- Audio measurement systems, including the dB scale
- And much of the maths that could be used for design and prediction of performance!
Compression Driver and Phasing Plug
Radial and Sectorial Horn
Cardioid and Condenser Microphone
Folded Low-Frequency Horns
This is a 40s graphic chart. The right axis is measured in ‘sensation units’. I’m not sure what kind of sensation they are referring to but it looks like the louder the better. With the growing popularity of the electric guitar and popular music in the 50s, the sensation scale was revised. Maybe you can invent your own chart?
Living in the 70s
The next major leap forward came with the early rock concerts promoted in the Late 60s. Whilst back in Australia we weren’t far behind the pack. Woodstock and The Isle of White rock and music festivals saw the birth of the large-scale rock concert. For the fist time, public address was not just speech reproduction from distributed horns barking sports results or political dissertations. In Oz, we had Sunbury ’75.
Check out the early attempt at fold-back wedges
The large-scale live sound market in 1965 was virtually non-existent. By 1975, full frequency at high volume levels was the name of the game. The only qualification you needed was a friend in a band. Efficiency was the priority over high fidelity. A normal hi-fi speaker box is 3-6% efficient. These boxes could be up to 65% efficient in converting electrical to acoustical energy.
The Sound System as used by Led Zep in Oz.
The above sound system toured Australia with Lead Zeppelin at the start of the 70s. I believe Australian company Jands supplied it (?). It is still essentially early 50s technology altered to suit this application. It was loud but not particularly hi-fidelity. At the Melbourne concerts, there were noise complaints from five kilometres away. This is very impressive for the racks of relatively low-powered amplifiers that were available. The efficiency of the system was very high, and for not much power in, there was a lot of noise out. The system sounded like a giant guitar amplifier and it was exactly what the crowd wanted.
These early horn-loaded systems were hard to tune “flat” and there was no science at all in the way the system was ‘arrayed’. The will be more on this developing science of arrayed systems later – hold tight. In order to design the new types of systems that could cover large areas with high fidelity, methods of measuring and quantifying sound behaviour and propagation were required. The information had been around for 20-40 years, but a few design engineers only understood it as a largely theoretical study.
For the person today who must prepare, install, specify or use a concert system, an understanding of the terminology and principles is essential. The days of ‘a mate of the band’ mixing the show, ended with the 70s.
Then and Now – Please Compare…
The PA system, assembled for Iron Maiden, represented state-of-the-art technology in the late 70s. Sydney production company Jands provided this particular system. These days, this PA would be called “a big ugly cluster” but in 1977, if it didn’t “turn you on” then you had no switches. It consists of folded horn bass, horn loaded midrange and bi-radial horn high frequency elements thrown into a big pile. What it lacked in sonic uniformity it made up for in sound pressure level; which was pretty impressive. Nobody apart from a few heavy duty maths freaks had heard of interference effect or line (source) arrays at that time.
The Nexo Geo system pictured is an example of the art in 2004. How did we get there? It’s a long story and there are a lot of ‘everybody knows’ opinions that will be challenged! This flown array (2004) represents almost the same sound pressure output but it is the size of less than two of the mid high boxes from the 70s PA. When augmented with sub-bass boxes, this light-weight and super-compact system is very high-fidelity and will have almost no sound pressure loss over 150 meters due to its very clever application of tangential array and acoustic mirror technology. This type of system is often referred to as a line array. Actually, there are very few real line array systems and this system pictured would be correctly known as a “Tangential array”. What tries to pass as a line array these days is often a ‘line source’. Its principle of operation takes us the full circle to the Greek geometry of 350 BC.
The Tools and the Technology
It’s a long road with a lot of by-roads that stop at dead ends, so let’s start with the basics. The first tool of the audio designer is the dB scale. The decibel is a generally misunderstood term. It was first used in 1929 by Bell Laboratories and it refers to a unit designation was ‘the Bel’, a unit to quantify a transmission unit scale to interface European and USA telephone systems. The Decibel represents one tenth of a Bel.
The Decibel was the logarithmic form of a power ratio, (10 0.1 ). This can be used to describe a voltage ratio, (x-decibels above or below a reference voltage, i.e. dBV), 2/ a power ratio, (dBm, m stands for milliwatt), or 3/ audio sound pressure level, dBL. The dB Level Scale is generally used in the live audio industry for expressing acoustic intensity and acoustic power.
This Particular dB scale is calibrated to reflect environmental noise (dBa). This scale is more sensitive to midrange frequencies than the scale generally applied to music reproduction systems. This is because the human ear does not hear all frequencies at the same ‘volume’.
Practical Test You Can Do at the Gig:
Turn up a sound source in 3dB increments. No, the scale next to the fader is not faulty; the ear is a poor measuring device for volume (loudness and power).
For decibels or dB to be useful for predicting sound coverage, you need to understand the Inverse Square Law.
The Inverse Square Law
It is a ratio that a change of one unit more or less in quantity, will double the result. In other words, when applied to an audio noise source, if you double the distance from the noise source, you have four times the decrease in volume. The effect of change is called the Inverse Square Law Change in Level. The changes in acoustic power are measured in 3db intervals. An example in terms of power, 53dB is twice the power as 50dB (see picture).
A sound source in an open space streams out uniformly in all directions. Imagine a light bulb and measure the light intensity at 1 meter. At 2 meters the light intensity will be one quarter the light level at 1 meter.
Technical Note: You will also see dB measurements on the VU meters (voltage units), of most mixing consoles. It is easy to confuse doubling the voltage (dBv set to a fixed reference point), results in a 6dB increase and doubling the power results in a 3 dB increase. When the voltage is doubled, the power increase is four times. Consider voltage as ‘pressure’ and power as ‘work done’
Why So Many Speakers?
The extremely high output of modern sound reinforcement loudspeakers is required because most of them radiate as point sources and therefore obey the Inverse Square Law. That is, their output drops –6 dB for each doubling of distance. (Certain types of speaker box loading configurations will vary from this figure). A Line Array style system behaves differently, it theoretically diminishes at a rate of 3dB per doubling, but we will go into this type of array later.
As we can see from the chart, 120 dB at 1 meter is only 86 dB at 50 meters (165 feet). In a relatively short distance, we have dropped from the threshold of pain to barely above normal conversation. We can regain the lost volume at the rear of the audience by multiplying the number of loudspeakers around the stage, with each doubling of speaker count regaining 6 dB of output. Alternatively, we could add more sources located toward the rear of the listening area, and “integrate” them with delay and equalization.
The reason for this is the geometry of the expansion of a sphere. Since the surface area of a sphere, the available sound energy must “cover the surface,” so its intensity/pressure at any given point is SPL origin ÷ pD2. The square of the diameter (the distance from the source to the surface of the sphere, on which our measuring point must be located) is in the denominator, thus the Inverse Square Law. It’s more than just yelling 1-2 testing
Another Rule: A system engineer in charge of installing a concert PA system must understand how sound travels through air, how it behaves in reflective environments, how we hear it and how to measure it.
Air is an elastic medium. Sound is a vibration travels through the air at about 1,100 ft (366.7 meters) per second - through water sound travels at over 4,000 feet per second. A sound wave can be measured in both frequency, (pitch) and volume (dB). As you can see, to cover a large area with sound, You either need one very loud sound source that you can focus onto the area required, or you need a pile of smaller output boxes assembled together to give you the necessary output - An array.
Even then, applying the Inverse Square Law, you only have a limited ‘throw’. With a system powering to 130 db @ 1meter, you will be down to about 95 dB @ 100 meters on a good day. But there’s more… Worst is yet to come.
For an engineer the pitch is unimportant, but the wavelength is everything!
More on the Ear
The human ear also is not sensitive to very high frequencies or very low frequencies. This is the origin of the “smilie curve” you may see on some equalisers set by people listening to music. (Check out any graphic EQ that returns from hire to a karaoke gig). The tendency is to boost bass and high frequencies. It might make music sound good but it will make live speech harder to understand, especially in a reverberant or noisy environment.
For the benefit of designers, way back in 1956, a couple of guys called Robertson and Dadson developed a graphic representation of free field equal loudness contours of pure tones. It shows how we hear a range of pitches. It is a frequency response graph of our hearing:
In the development of the loudness contours, a person listens to a sweep of frequencies across the frequency spectrum and sets each to what they hear as the ‘same volume’. As you will see, what we consider the same volume in each frequency band varies considerably in power required to amplify each frequency to the same perceived dB level. This will vary from person to person. In other words, we do not all hear the same.
Another Practical Test:
An effective demonstration is to set a frequency sweep generator at fixed out put level and wind it through the full spectrum at high power. This is a great experiment to clear the venue of patrons when you want to load out the PA. When you hit 1K, the crowd will be ‘running for cover’.
Rule Number Three:
The successful audio engineer will work with the acoustic environment, not against it. To do this successfully, he will understand the behaviour of sound waves and how we hear…
Whether indoors in a gym hall or outside in a windy paddock, we as sound contractors are paid to get intelligible results. We have seen that the ear is a poor determiner of real volume, it has a lumpy frequency response and sound reflections or excessive natural echo cause a distracting loss of clarity. We have also learned that the ‘human hearing circuit’ has a mechanism to detect direction of sound sources. This is very beneficial for survival for threat detection with natural sound in a normal environment but with an array of speaker boxes, we have introduced a single sound being reproduced by multiple sources and at least two locations, (left and right). Added to this there could also be ‘delayed speakers’ and ‘infill’ speakers.
Plenty of air is being moved, but it was moving in too many directions. Sound is arriving at the listener’s location at too many different times because the pressure waves from multiple speaker sources overlapped. A practical example is analogous to throwing a stone into a pond. Even waves spread out in all directions, their energy dissipated uniformly. If they hit any object or barrier, waves are reflected back into the on coming waves causing little wavelets to disturb the even spread of waves.
If you throw a handful of gravel instead of a small stone, there will be a disturbed wave flow right from the first splash. Imagine one speaker source as the small stone and an array of speakers as the hand full of gravel. In the frequency domain, this uneven confusion of waves is called Lobing and Comb Filtering: it results in inconsistent frequency response across the coverage area. In the time domain it’s called multiple arrivals. The human ear is quite sensitive to multiple arrivals, especially in the horizontal plane, since that’s how we locate the source of sound.
In the natural world we use the phenomenon of comb filtering to detect the direction of a sound source. If a PA system is affected by Comb Filters it will sound unnatural.
Look closely at the two-speaker response line attenuated by 6 dB; the comb filters are much less. There is a practical clue for a sound guy about arraying a big PA; can you figure it out?
The human ear is perfectly designed to hear speech and detect the direction of a sound source due to its stereophonic nature. A sound that arrives at both ears at exactly the same time (remembering 1,100 ft per second) is right in front of you. Sound to one side will arrive at one ear slightly later than it arrives at the closest ear. This fractional time delay between arrival times is how we perceive direction. The ‘phase shift’ (time shift), caused by the differing time arrival changes the sound a little. Small cancellations take place and on a frequency graph you would see a series of deep narrow dips. This is called Comb Filtering.
We now have an understanding of the dB scale, frequency and wavelength, loudness contours, the inverse square law and comb filtering. Let’s apply this to a real PA array installation.
The Ear and the Big PA (Just two boxes together big)
Our first concert situation is an array of two full range boxes each side of the stage at an outdoor festival. It sounds like a simple situation.
We can see now that any single loudspeaker may have an even frequency response, and the sound quality of a single loudspeaker may be nothing short of hi-fidelity, but that will not be the case once arrayed. With these direct-radiator packaged speaker systems grouped in a cluster, the individual drivers cannot be positioned to minimize time delays between devices and the resulting destructive interference, cancellation and response variations. (Comb filtering). These variations are important because the coverage uniformity in the 1,000Hz to 4,000Hz range is critical for good intelligibility and uniform frequency response from 250Hz to 1,000Hz is important for natural sounding speech or music.
The empirical rule use to be, put up one box or 100 and anything between was trouble. This was found by practical experience that if you put up enough boxes, all the aberrations seemed to even out if you stood far enough away. (Too bad for the people who paid to hear and have the ‘dud’ seats. There will be plenty of them).
So let’s add more boxes to our two-box array? May be that will help.
We have gone from throwing four pebbles into the pond, (two per hand full), to eight. Let’s see what happens.
In the four-box cluster of unspecified Brand-X 2-way full-range boxes, there is approximately 1 meter between the box centers, both horizontally and vertically. (That's a half wavelength at 225Hz, one wavelength at 450Hz, and two wavelengths at 900Hz). At the crossover point, where the horn and the 12" bass speaker in each box are operating at equal levels, there are a total of eight devices operating at the same level with multiple wavelength spacing between all eight sound sources. This is the offset distance vertically, horizontally and diagonally between all devices. Where the coverage of the boxes overlap, there will be significant lobing of the coverage and comb filtering caused by these physical-offset-induced signal delays.
The problem is compounded by the inability of the small format high frequency horns that are used in these boxes to provide high directivity, (narrow focus), in the vertical plane above the crossover frequency. The actual vertical coverage of the horn can be well over 100 degrees (and can approach 180 degrees) at the crossover point, ensuring vertical overlap of coverage between boxes, and destructive interference and cancellation. Above the crossover point, the comb filtering will fall in the critical speech intelligibility range, and the resulting response notches are often in excess of 1/3 octave wide.
Below the crossover point, the 12" bass drivers gradually broaden their coverage angle until the boxes each become almost omni-directional. In the frequency range between 250Hz and 1000Hz, very large variations in level will be found, with the highest uniform level exactly on the centre axis of the cluster, and spurious lobes all around that centre hot spot. This produces plainly audible sound quality variations with seating position, and may produce feedback prone positions under the cluster. (Enter the Cabaret singer who now wants to take full advantage of his new radio microphone).
Let’s Look at a Graphic Representation:
Wave C is now cutting across wave D out of time sync. The slight time variation (60 degrees) will mean that there is a pressure lag between the waves and they will be working against each other. It’s like a tug of war with everybody pulling at a slightly different time.
What about the fancy system controller?
The problem is related to the physical distance between individual loudspeaker components. The special signal processor/crossover that comes with some loudspeakers can only correct for time offset of the signal between the low and high frequency drivers along a single speaker system axis. The signal delay in the processor cannot adjust for every possible off-axis listening angle in a two-box array. The processor certainly can't fix the additional multiple signal delay variables thrown in by having additional boxes added to the cluster Bottom of Form
So why don't you just equalize it?
Consider the system equalizer that is intended to improve the quality of the sound. All four loudspeakers are connected to the same signal chain, so any change in equalization will affect all four speakers. Which of these frequency response curves would you correct for? A couple of the seats are not too bad, a couple of them (including the house mix position!) are very poor. If you tried to correct the bumps and dips in one location, you would actually be making them worse in other areas.
This is really a non-equalizable problem caused by physical distance offset of all the drivers (this is what is known as a non-minimum phase problem caused by a significant signal delay). Even though the loudspeakers are designed to be arrayed horizontally, there is no way to get around the physical offset problem vertically with small packaged two way direct-radiator systems. There are no delays or equalizers that will fix this problem (until someone develops a four dimensional signal processor).
Here we are back at our two-box frequency response plot. The only place to stand is right in front of one of the boxes for a flat response. The problem will be that the sound guy will have tried ti equalise out the lumps and because this is a mechanical cancellation, all he will achieve is to add more peaks and lower the intelligibility even further.
Remember the earlier question? An equalizer will not fix it, What about turning one box down a couple of dB? This trick will also smooth out a stereo mix when converted to a mono mix. Eg, sent to delay stacks or out fills.
With luck and experimentation, we might be able to steer the worst of the problem area to a spot where the crowd isn’t?
What if We Try a Different Type of Horizontal Array?
In this particular configuration, the comb filters and peaks are most severe below the crossover point, where there are three 12" loudspeakers and horns widely arrayed. The 12" speakers cannot be positioned to minimize the interference between them because the boxes are pre-configured with a horn and top end with differing dispersion. By the time you get the high-frequency horns pointed where they have to be pointed, the woofers are typically in the worst possible relationship to each other. The laws of physics and inappropriate box selection triumph every time. In fact in this particular instance, the only available correction would be turning off two of the centre box. This would of course result in a sonic hole between the boxes but the Lobing would be much less.
The reason for the extra boxes was not enough SPL to cover the distance. With all the losses and cancellations, severe EQ and headroom loss, we really have not progressed from the smaller array as far as sonic performance is concerned. Remember the dB scale and the human ear? For a slight increase in perceived volume, we must add 3dB, (double the power). So we need a lot more boxes to make a difference.
Forget the problems of the three or four-box array, even the two-box array is nothing to get excited about. And the view from above a two-box array: A predictable mass of ‘fingers’ ready to make the system lumpy and drive an army of complaining punters to the mixing desk.
By the time you get to the mixing position, it will be different again.
But this format of compact speaker box is very popular. I.e. a 2/3 way horn top end and a direct radiating 12” or 15” speaker with possible added midrange.
The original market for small packaged speaker systems grew from the small band and sound system rental market, where it was important that the systems be easily transportable, reliable, and could make music sound nice for the small venue. As single box solutions, the good boxes work really well and the boxes designed solely by the ‘marketing department’ are a poor compromise.
An ideal speech and acoustic music reinforcement system should be invisible to the listeners, the voice should sound completely natural and not sound amplified, it should just be louder to effectively move the talker and listener closer together. An ideal system would also preserve or improve the direct-to-reverberant ratio of the reinforced sound (critical distance), but that's another issue. The marketing departments have convinced many consultants, clients and contractors that these small systems are a panacea, a universal problem solver for sound systems. There are many examples of installations using this type of speaker system for speech and music reinforcement, and yet when you listen to a voice through the final result, the deficiencies are often very apparent. Somewhere in all of this marketing hype, people have quit listening to what the finished systems sound like. Or don’t feel confident or articulate enough to protest.
Don't get fooled, by ‘everybody knows’ generalisations and simplistic solutions.
Just because something appears to be popular, does not mean that it is good. Many of these small rock and roll speaker rigs sound very impressive, when demonstrated with a CD of pre-recorded music. Most listeners are much more forgiving of response variations when listening to music. If you're listening to a system intended for speech or complex music reinforcement, insist that it be demonstrated with a live microphone in the room at real world volume levels.
Is there reasonable gain before feedback? If the speaker system is hung above the microphone, are there bass lobes aimed at the microphone position that will induce feedback? Above all else, is the voice going to make it to the back of a crowded noisy room?
Some compact boxes may sound a bit ‘in your face’ when tested in a shop or small room. This could be because the mid range is very present. Many self-powered speaker systems have been pre-equalized to sound ‘nice’ but you will need those midrange frequencies back again if you want to project clear vocals to the back of the room. For a speaker box designer, those midrange frequencies are the hardest thing to get right because the crossover into the horn is often right in the middle of the midrange. Getting vocals right is the hardest thing because it is the only thing everybody agrees must be there and is use to hearing. This especially applies when you're buying a speech reinforcement system, listen for how well it delivers the speech intelligibility your audience needs and don't get fooled by marketing and technical hype.
None of this theoretical material will help you when the people who paid for expensive seats are complaining about audio issues that are out of your control. Maybe it’s time for another rule that might help your career.
Rule Number Four
No sound guy ever got the sack because the vocals were too loud.
Next time you have to put up an array of boxes, familiarise yourself with array behaviour with the following tests:
- Time delays simulated and real (phase shift):
Set up a two-channel array together with a digital delay on one channel. Show with pink noise sonic variations with a few milliseconds variations between boxes. Also dB output variations on one box and array behaviour.
- Polarity cancellation:
Hear polarity reversal on one channel electronically and acoustically. Especially try reversing the phase on one element of top end. The image will shift radically. Remember this sound and you will become the worlds best monitor engineer by spotting out of phase wedges instantly.
Physical alignment vertically and horizontally
Put up a poorly aligned array and test with pink noise. Move the boxes into a theoretically less conflicting alignment. Walk around and look for lobes and null points.
There is a lot of information on the Internet these days, some of it conflicting when you trawl the Major speaker company sites. You will have to make up your own mind but some sites worth visiting are (in no particular order):
…. Good Luck!
Next Week: The Line Array and Political Correctness.
Filter Home Page