Imagine a J-array hanging in a 1,200-seat theatre. The front row is 8 metres from the rig. The back row is 56 metres. Without any intervention, the inverse square law delivers roughly 17 dB more direct energy to the person in row one than to the person in the last seat, and that is before you account for air absorbing high frequencies over distance.
Something has to give. The question is not whether to compensate. The question is what you are willing to trade to do it.
First, a bit of physics
A line array is not just a stack of loudspeakers. It is an acoustic tool that exploits mutual coupling: when closely-spaced drivers radiate at the same frequency and phase, their outputs combine constructively, narrowing the array's vertical dispersion and extending its effective throw. You get controlled coverage instead of a spreading blob of energy.
This coupling is frequency-dependent in a useful way. Long wavelengths (low frequencies) couple across the entire physical height of the array; short wavelengths (high frequencies) only couple meaningfully between adjacent elements. A well-designed line array naturally behaves like a longer column at low frequencies and a shorter column at high frequencies, a self-correcting property that gives you more consistent vertical control across the spectrum than any point source could.
The moment you start modifying the signal chain per element, as both shading techniques do, you risk undermining these properties. The question is how much damage each approach actually causes.
Gain shading (amplitude shading)
Gain shading is the simplest approach: turn down the elements pointed at the front seats. The lower cabinets in the J (the hook covering near-field seats) are attenuated; the upper cabinets covering the back stay at full output.
The logic is intuitive. The person in row three is already getting hammered by proximity, so pull those boxes back. Problem solved. Except it is not, quite.
The wavefront discontinuity problem
When you reduce the drive level on the lower elements, those cabinets still radiate, just at lower pressure. This creates a pressure mismatch at the boundary between attenuated and non-attenuated elements. That discontinuity is audible: it behaves as though a separate, non-coherent source has been introduced, producing transient smear and uneven frequency response at the transition zone. For listeners in that region, the array stops sounding like a single coherent source.
The effective line length problem
At low and low-mid frequencies, the array's vertical directivity depends on the entire physical height of the stack acting as a single distributed source. Attenuate the lower elements and you effectively shorten the acoustic aperture. You are sacrificing the array's most valuable property, its length, to solve what is at its core a high-frequency loudness problem.
The lobing problem
Mismatched amplitude between adjacent elements also alters the interference pattern between them. The smooth polar pattern you designed around develops secondary lobes, off-axis energy concentrations pointing at ceilings, floors and reflective surfaces. In a reverberant space this is particularly damaging, because those lobes excite room modes and generate flutter that muddies the direct sound for everyone.
Frequency shading
Frequency shading takes a more surgical approach. Instead of reducing the overall level of near-field elements, you apply different EQ curves to different zones in the array. Typically that means rolling off high frequencies on the lower cabinets (the elements covering nearby seats) while leaving their low and mid output untouched.
The key insight is that the tonal imbalance across a room is largely a high-frequency problem. Distant listeners receive less air-absorbed top end. Near-field listeners receive brighter, more direct HF because they are on-axis and close. The solution is not to turn anything down; it is to sculpt the spectrum differently per zone.
Why low-mids are the protected range
Frequency shading leaves low-mid frequencies untouched, which is precisely the point. Preserving maximum line length is what controls those frequencies' vertical dispersion. When you frequency-shade rather than gain-shade, the array keeps behaving as a full-height column for the frequencies where that height matters most. Coupling is preserved. Directivity is preserved. The wavefront stays continuous.
Addressing air absorption directly
Frequency shading also offers something gain shading categorically cannot: the ability to compensate for air absorption in a frequency-dependent way. By applying a gentle high-frequency boost to the upper (far-field) elements, or a high-frequency cut to the lower elements, you can narrow the tonal gap between front and back. Gain shading, which affects the whole spectrum equally, cannot do this without making the balance worse at one end or the other.
Practical implementation
In practice, frequency shading is applied by dividing the array into zones, typically two to four elements each, and applying a gradual taper with parametric EQ or high-shelf filters. The finer the zone resolution, the more gradual the taper and the less likely a listener will notice a step between zones.
For the most precise control, FIR (Finite Impulse Response) filters offer a real advantage: they can apply amplitude adjustments independently of phase, so the EQ on one zone does not perturb the phase relationship with adjacent zones. Systems with per-driver DSP and FIR capability can frequency-shade at very high resolution without introducing the inter-element phase errors that would otherwise compromise coupling.
A third option worth knowing: divergence shading
Gain shading and frequency shading get most of the airtime, but there is a third approach that deserves a mention: divergence shading.
Rather than adjusting level or EQ, divergence shading widens the vertical dispersion angle of the near-field elements. The boxes covering the front rows splay more aggressively, spreading their energy over a larger area, which reduces the SPL per unit area for nearby listeners without reducing the element's total acoustic output or changing its frequency response.
The wavefront continuity benefits are significant: every element still radiates at the same level, so there is no pressure mismatch at the boundary between zones. Several manufacturers have offered dual-dispersion cabinet designs specifically to enable this, matching the output of wide-angle and narrow-angle elements so SPL at the cabinet mouth stays consistent across the array.
What this looks like in practice
In real-world tuning workflows, using prediction tools like Soundvision, ArrayCalc or MAPP XT, the order of operations usually follows a clear hierarchy:
Gain shading has a legitimate role. It is just not the one most engineers reach for it to play. It is a finishing tool, not a primary coverage strategy.
Comparing the approaches
| Feature | Gain shading | Frequency shading | Divergence shading |
|---|---|---|---|
| Control method | Overall level reduction | Per-band EQ (parametric / shelving / FIR) | Adjusting element splay angles |
| Affected spectrum | Entire range | Selected bands (usually HF) | All frequencies (spatially distributed) |
| Low-mid coupling | Reduced, shortens aperture | Preserved, full length intact | Largely preserved |
| Wavefront continuity | Disrupted at zone boundaries | Mostly preserved | Preserved |
| Addresses air absorption | No | Yes, directly | No |
| DSP complexity | Low | Medium to high | Low (mechanical) |
| Risk of audible artefacts | Medium to high (lobing, steps) | Low if applied gradually | Low |
| Recommended role | Final trim only | Primary tuning tool | Design-stage alternative |
The bottom line
Gain shading is tempting because it is simple: pull a fader, reduce the level, done. The front row gets quieter and the numbers look balanced. But the simplicity is deceptive. You are solving a level problem by shortening your array, disrupting your wavefront and potentially throwing energy into reflective surfaces, all while leaving the tonal imbalance from air absorption entirely unaddressed.
Frequency shading is harder to do well. It needs an understanding of what is happening at each frequency across each zone, and the parameter space is larger. But it works with the physics of the array rather than against it. The coupling stays intact. The line length stays intact. And you finally have a tool that can compensate for what air does to sound over distance.
The professional consensus is clear: physical design first, frequency shading second, gain shading as a last resort. Understanding why, not just what the rule is but what happens to your array when you break it, is what separates basic tuning from professional system optimisation.