Things you should know before opening your laptop. ... speaker manufacturer without one, what's not to like? 2. J-Arrays improve the vertical coverage ... This bolsters the sales of DSP devices, but we can't fix directivity discontinuities with.
Designing & Deploying Line Arrays Things you should know before opening your laptop... Mark Engebretson
Part I - Line Array Architecture ‘Advancing science in the face of conflicting beliefs and practices is Engineering's greatest challenge’
Introduction The past several years have witnessed unprecedented growth in the complexity of sound reinforcement equipment and technologies. Today we have simulation programs to show complex loudspeaker coverage in three-dimensional spaces, measurement platforms to quantify detailed electro-acoustic performance, digital transport and networking schemes and a host of increasingly sophisticated computer-based control and signal processing devices. Virtually every new active element in the signal path includes some form of on-board DSP. Against this backdrop of ever-increasing complexity, the purpose of these papers is to provide easy-to-understand information and guidelines that will enable consistent high quality results from line array deployments. As such, they are neither an exhaustive treatise nor are they highly technical. They provide common-sense examination of fundamentals that govern all forms of sound reinforcement aimed at improving sound quality and listening experiences for audiences and audio practitioners.
Occam's Razor Occam's Razor is a principle attributed to the 14th-century English Franciscan friar, William of Ockham. It states that the explanation of any phenomenon should make as few assumptions as possible, eliminating those that make no difference in the observable predictions. When multiple competing hypotheses are equal in other respects, the hypothesis that introduces the fewest assumptions and postulates the fewest entities is best. The simpler a theory is, the better. When two theories predict phenomena to the same accuracy, then the one which is simpler is usually the better one. Be skeptical when facts and results are postulated from theories or assumption-based hypotheses. Break it down and keep it simple!
Line Array Myths Disruptive technologies develop their own mythos. These are reinforced by popular beliefs that interpret the technology and its meanings to all those affected. Unfettered, these beliefs become the common wisdom. There has been a litany of misinformation that has been promulgated about line arrays and their deployment. This has led to mixed results at every level. Line array technologies need to be properly understood because they underpin sound reinforcement today and in the foreseeable future. Following are the ten most widespread line array myths.
Ten Line Array Myths 1. Line arrays are a fad Not true. Line arrays offer cogent means to increase coverage and SPL while reducing temporal distortion and the architectural footprint of the loudspeakers. Unless you are a speaker manufacturer without one, what's not to like? 2. J-Arrays improve the vertical coverage J-arrays consist of two totally different loudspeaker arrays. They perform poorly because of the withering discontinuity where curved and straight segments join. 3. Down-fills are a good way to cover 'down front' Like the doomed pilot who runs out of altitude and ideas, we sometimes run out of both time and viable alternatives. But like the J-Array, using down-fill boxes splices
completely different loudspeakers to the main array, creating interference where coverage conjoins. 4. Split processing can optimize J-Arrays This bolsters the sales of DSP devices, but we can't fix directivity discontinuities with DSP. Curved and straight arrays have radically different vertical directivity characteristics and should not be connected together. 5. Simulations show the best way to configure line arrays Simulation programs are subject to the same human errors and manipulations that haunt every complex software, only more so. Most programs are based upon assumptions and constructions that fail to recognize the effects of discontinuities in array shapes. Use simulations with care, but don't recommend or construct an array that has a physical discontinuity just because a simulation shows appealing coverage representations – it isn't possible. 6. Configurable horizontal coverage improves spatial uniformity Configurable horizontal coverage seems like a good idea, but we cannot achieve directivity-matched transitions through crossovers while making waveguide mouth openings smaller to narrow the horizontal coverage. The result is irregular directivityfrequency and inconsistent frequency response in the critical middle and upper middle frequencies. Efforts to 'fix' the frequency response with filters come at the expense of spatial uniformity. 7. Each venue requires its own unique DSP This stems from the misconception that observed misbehavior is caused by room acoustics, and that DSP can somehow rescue the day. Both assumptions are false. Most misbehavior is either the loudspeakers or the arrays, and DSP has no effect on room acoustics anyway. Use equalization appropriately. 8. Tried-and-true audio practices perform well with line arrays Most 'tried and true' practices don't perform as well as we'd like to think they do, and they are less likely to perform well with line arrays. Best results will always come from careful cause-and-effect analysis before postulating solutions. 9. High sound levels from line arrays are OK If the sound leaving the loudspeakers is 'clean' there won't be any distortion, right? Wrong. Three types of acoustic distortion are more significant in live sound than all others and line arrays have a propensity for one of them. 10. Line arrays radiate cylindrical sound fields All that can be said in support of this thesis is that the near field of a line array is an interference field that roughly follows the frontal aspect of the loudspeakers. Inverse square law has not been abrogated. The far field acts like that of any other loudspeaker. All of these myths are fallacious They are either untrue, or require narrow context setting to be true. Some are the result of faulty understanding of the basic mechanics of how line arrays work and some are deliberate misrepresentations. Nevertheless, they are the cornerstones of line array folklore that are responsible for much of what sounds bad today. Fortunately, practicing good science will always be easier than fumbling bad science.
Why Array? Arrays serve to increase, decrease or re-shape coverage and/or increase the sound pressure level. Other than these, arrays have no useful purpose. Loudspeakers that are physically offset from one-another with conjoining coverage are a source of temporal distortion – combing and time smear in three dimensions that cannot be 'fixed' with one-dimensional solutions. That some loudspeakers can be mounted closer to one-another, and that some might have less coverage overlap than others only reduces combing and time smear. Minimum temporal offset results from small, tightly packed sources. Small sources have low directivity and 'soft' pattern edges, increasing coverage overlap and temporal distortion. High-directivity sources are big – separating the sources in space and time, which also increases temporal distortion. Line arrays optimize the ability to conjoin coverage of a plurality of like sources to produce minimal temporal offsets in the direction of coverage, but they cannot repeal the realities of time and space.
Why Line Arrays? Line arrays enable high sound levels, when compared to traditional multi-way systems. Due to their greater length, line arrays maintain high vertical directivity to much lower frequencies. • •
Improved direct-to-reverberant sound ratios in enclosed spaces Reduction of atmospheric interference effects out-of-doors
Line arrays can be constructed to provide optimally wide vertical coverage to meet special auditorium needs, e.g., balconies. They can also be shaped to provide tapered vertical coverage for very deep auditoria, long throws and low trim heights. Line arrays lack one dimension that is responsible for temporal distortion in large sound systems.
Designing & Deploying Line Arrays Things you should know before opening your laptop... Mark Engebretson
Part I - Line Array Architecture How Line Arrays Work CONVENTIONAL HORNS • Expanding wave fronts propagate normal to the boundaries of the horn. • At the mouth, high frequency wave shapes from adjacent horns will create destructive interference. • The practical upper frequency limit for summation is reached when the rise (a) of the resulting circular segment at the mouth equals a quarter-wavelength. • This makes conventional horns unsuitable for widerange line array use. LINE ARRAY WAVEGUIDES • Line array waveguides use various methods to normalize the propagation distance from the throat to the mouth. • In this illustration we see that each waveguide has been segmented into four identical small apertures, each equidistant from the inlet aperture. The result is minimal rise at the mouth approximating a planar wave shape. • This enables adjacent loudspeakers to sum without interference at high frequencies.
Segments & Line Arrays Concert Line Arrays consist of Segments Segments have Coverage Limitations Segment Coverage Limits are frequency‐dependant
3 Line Array Segments:
Straight vs. Curved Line Arrays
Vertical Coverage
Line Array Types Straight Arrays Straight arrays are best characterized by 'column' loudspeakers, wherein all of the array elements are oriented in a continuous, straight vertical line. Vertical directivity in straight arrays is a function of array size and frequency, with vertical coverage being inversely proportional to their product, and narrowing monotonically as frequency increases.
Curved (Arcuate) Arrays An arcuate array is curved by virtue of its diverging segment axes. When assembled from independent cabinet-type segments, the inter-box splay angles are held constant, which results in a constant effective radius of curvature. A relatively large array assembled from segments takes on the effective vertical coverage of its cumulative segment-to-segment splay angles and maintains constant vertical coverage between the overall size-imposed break point frequency and the frequency where beaming of the individual segments takes place.
J-Shaped Arrays J-arrays join two totally different loudspeaker arrays – a straight segment intended to cover distant seats, and a curved segment disposed to cover seating near the stage. J-arrays inevitably perform poorly because of the withering discontinuity where the curved and straight segments join.
Spiral Arrays Spiral arrays do well what the J-array was envisioned to do. Employing arithmeticallyincremented inter-element splay angles to decrease the effective radius from the top to the bottom of the array, spiral arrays provide constant vertical frequency-coverage graduating from high-to-low directivity. They are easily implemented and readily compensated.
Characteristics of Array Shapes Straight
Characteristics: •
Monotonic narrowing with f & l
•
Far field distance varies with f & l2
•
Directivity = N · segment directivity
•
Minimal temporal distortion
•
Very narrow HF beamwidth
Curved (Arcuate) 1° splay each segment Characteristics:
•
Constant frequency/coverage
•
Coverage = (N-1) (splay angle°)
•
Example: 8 @ 1° = 7°
•
Uniform far field boundary
•
Low temporal distortion
Curved (Arcuate) 5° splay each segment Characteristics: •
Constant frequency/coverage
•
Coverage = (N-1) (splay angle°)
•
Example: 8 @ 5° = 35°
•
Uniform far field boundary
•
Low temporal distortion
J-Arrays
Characteristics: •
Inconsistent coverage & DI
•
Coverage ≈ ∑ splay angles°
•
Example: < 1° + (4 · 10°) ≈ 41°
•
Irregular far field boundary
•
High temporal distortion
•
Propensity for Propagation
Distortion along the straight segment axis
Spiral Arrays Characteristics: •
Constant frequency/coverage
•
Coverage = ∑ splay angles°
•
Example: 1+2+...+N° = 28°
•
Far field transitions smoothly
•
Moderate temporal distortion
Reshaping Arrays (shading) Once an array is positioned and flown, there are a number of possible adjustments that will reshape the coverage: Divergence Shading means adjusting the inter-box splay angles to either increase or decrease the coverage. For safety and practical considerations, divergence shading normally requires landing large arrays to make the desired changes on the ground. Note: Do not make changes that will alter the shape of part of the array – proper divergence shading requires re-configuring the entire array to maintain array integrity. Amplitude Shading is the application of box-to-box level adjustment. It is typically used to attenuate the output of the lower elements covering the audience closest to the stage. Successful amplitude shading requires 1-2 dB maximum incremental changes in box-to-box drive level to preserve array integrity. This necessitates having sufficient electronics granularity, preferably discrete amplification channels per box. Delay shading 'bends' the virtual array by employing signal delays to change its effective shape. Like amplitude shading, delay shading requires sufficient amplification granularity to smoothly transition the effective shape. Typically, this requires delay steps less than a quarter period of the highest frequency. Note: Delay shading cannot 'bend' the coverage beyond the native coverage of the individual elements.
Multi-EQ Systems – Split Processing employs multiple DSP devices within an array as a means to normalize the response of different parts of the array. It is a vestige of the J-array wherein a straight line array segment conjoined a curvilinear segment. The withering discontinuity where the two different arrays conjoin cannot be corrected with DSP. However applied, split processing cannot correct directivity discontinuities. It can only reposition the interference. It serves no function in a coherent array.
Importance of Array Location Non-Conventional Thought Starters
Uniform front-to-back coverage isn't always desirable It comes at the expense of coherence & naturalness Uniform spectral balance in large spaces isn't natural Highs roll off with distance more than the lows Intelligibility is driven by signal-to-noise ratios The direct sound is signal - everything else is noise Most conventional arrays are inherently 'reverberant' Overlapping coverage creates 'time smear' Coherence requires simplicity Location for Coverage Uniformity Characteristics: • • • • • •
Uniform front-to-back SPL Uniform spectral balance Unnatural source location Likely need for front fills Reduced direct-to-reverberant ratio Reduced coherence
Location for Coherence Characteristics: • • • • •
Natural front-to-back SPL fall off Natural spectral balance shift Improved direct-to-reverberant ratio Improved coherence Reduced atmospheric effects
Subwoofers Subwoofer systems come in many different configurations: • • •
•
Closed Box, Ported, Horns and Hybrids Ground-stacked or Flown Conventional or Directional o Cardioid o Delay-Shaded o Forward-Steered Others as yet undefined types
Some do not integrate well with the main system
Subwoofers - Proscenium Applications Generalized choices of typical subwoofers used with line arrays:
Synchronizing Subwoofers 'A' shows a 'typical' rig with both the subs and the mains delayed to the back line. This works reasonably well with narrow venues. It ignores performance at wide off-axis angles, however. This can be problematic in arenas and in fan-shaped environments where the coverage needs to be very wide or 'in the round' applications. In 'B', we can see why this is so. The side flanking arrays have a different temporal relationship to the subwoofers than the front mains, while the back hemisphere arrays are in altogether different time zones. No single prescription is a panacea for every application. Remember that sound involves three dimensions – time has only one.
Steered Subwoofers A long line of subwoofers placed horizontally along the front edge of the stage can be delayed to create a virtual array that is much smaller and less directional. The center subs need to be several milliseconds ahead of the ones flanking them, which are several milliseconds ahead of their neighbors, continuing in this fashion until reaching the last subwoofers at the outside ends of the row. This helps to counteract the narrowing of the horizontal coverage with a horizontal subwoofer array. This works well out-of-doors. Use with caution in reverberant spaces. The delayed elements can exacerbate the effects of reverberation, contributing to an unintelligible 'roar' at low frequencies
Directional Subwoofers Cardioid subwoofers enable directional radiation at low frequencies. The directivity index along the major axis of a cardioid is 5 dB, while the output in the back hemisphere is attenuated by 2040 dB relative to the on-axis. Cardioid operation requires two loudspeakers, one behind the other separated by about a quarter wave-length. The nearest enclosure must be delayed by the propagation time to the rear and one of them inverted in polarity. This is only effective over about one octave and in vertical arrays. Use caution in reverberant environments.
Summary 1. Choose the array location and type. Use Arcuate or Spiral array shapes to minimize discontinuities in array shape. 2. Determine maximum box-to-box splay angles from Segment Coverage Limit calculation. 3. Avoid discontinuities in array shape and drive: o Spiral arrays, use 1- or 2-degree maximum splay angle steps o Do not mix cabinet types within the array o If shading, use small level or delay steps: < 2 dB; < 0.25Τ (where Τ = period of highest frequency of concern) 4. Avoid multi-EQ (split-processed) system configurations. Results are of dubious value. Difficulties are daunting. 5. Vertical coverage is the sum of the splay angles. 6. The vertical coverage limits are the vertical axes of the top and bottom cabinets. 7. For best results, point the loud end of the array toward the audience
