Both, in a fixed order. Acoustic treatment controls what the room does above roughly 150 to 200 Hz and shortens decay time. Electronic calibration sets level, distance and delay for every channel and corrects the modal region below that crossover point. Calibration cannot shorten a reverberation tail and absorption cannot time align a subwoofer. Treat first, calibrate second.

That order is not a preference. It follows from physics, and the numbers below show where the boundary sits.

The boundary has a name: the Schroeder frequency. Below it a room behaves as a set of discrete resonances. Above it, reflections are dense enough that the room behaves statistically and you can talk about reverberation.

Schroeder frequency is approximately 2000 times the square root of RT60 divided by room volume in cubic metres. Take a fairly typical private cinema of 20 by 16 by 9 feet, which is about 82 cubic metres, with a decay time of 0.35 seconds. That gives roughly 130 Hz.

So in that room, everything below about 130 Hz is a modal problem and everything above it is a reflection and absorption problem. DSP is very effective on the first. It is close to useless on the second.

Three things, and only these three.

Decay time. Porous absorption removes energy from the room, which shortens the reverberation tail. A small critical listening room should land somewhere between 0.25 and 0.4 seconds, reasonably even across frequency. Rooms with stone floors and plaster walls routinely measure 0.7 seconds or more, which smears transients and buries dialogue.

Early reflections. A reflection arriving at the seat within about 15 milliseconds of the direct sound fuses with it perceptually and shifts the apparent image. Treating the first reflection points on the side walls, ceiling and floor between speaker and seat is the single highest value square footage in the room. The mirror method finds them: sit in the primary seat, have someone slide a mirror along the wall, and mark every point where a speaker driver appears.

Frequency balance of the reflected field. This is where thickness matters. A 2 inch fiberglass panel of standard 6 pcf density absorbs efficiently down to roughly 500 Hz and does very little below 250 Hz. Move to 4 inches with a 2 to 4 inch air gap behind it and useful absorption extends to about 125 Hz. Rooms treated with thin panels everywhere end up dull on top and still boomy underneath, which is the most common self inflicted problem we find.

Calibration handles the things that are pure signal domain and that no amount of absorption will touch.

Channel level. Every channel matched at the seat, typically to 75 dB SPL with a reference test signal, so a pan across the front stage does not change loudness.

Distance and delay. The processor time aligns channels so that a sound intended to arrive simultaneously does. On a subwoofer this matters more than most people expect, because the sub is usually several feet from the mains and its own internal DSP adds latency of its own.

Bass management and crossover. Setting the crossover, usually at 80 Hz for a system with full range mains and lower for very large mains, and confirming that the handoff sums flat rather than dipping or bumping.

Modal correction. This is where DSP earns its place. Below the Schroeder frequency, room modes create peaks of 10 to 15 dB and dips that are just as deep. Platforms such as Dirac Live, Anthem ARC Genesis, Trinnov Optimizer and Lyngdorf RoomPerfect measure at multiple positions and apply correction over that region. In the modal range this works extremely well.

Target curve. Flat measured response sounds bright in a small room. A house curve that is flat to about 200 Hz and then tilts down by roughly 1 dB per octave to 20 kHz is a common and defensible starting point. The final tilt is a judgement call, and this is exactly the point where a trained ear matters more than a graph. This is the core of our entertainment work.

Because a null is not a level problem, it is a cancellation.

When direct sound and a reflected copy arrive out of phase, they subtract. The classic case is speaker boundary interference. A speaker sitting 3 feet from the front wall creates a cancellation at the quarter wavelength frequency, which is 1130 divided by 4 times 3, or about 94 Hz. At that frequency the two arrivals are close to equal and opposite.

If you boost 94 Hz by 12 dB to fill the dip, you are boosting both the direct sound and the reflection. They still cancel. All you have achieved is 12 dB of amplifier headroom and driver excursion spent on sound that never reaches the seat. On a serious system that is how woofers get damaged.

The correct fix is geometric: move the speaker, move the seat, or absorb the boundary. Which is a treatment and placement decision, not a DSP decision.

Peaks are the opposite case. A modal peak is genuine excess energy and cutting it electronically works well and costs no headroom. The working rule most calibrators follow is to cut freely and boost sparingly, if at all.

They do a different job, and in most private cinemas they are the higher value move.

A single subwoofer excites the room modes in a fixed pattern, so bass response varies dramatically from seat to seat. Two subwoofers placed to counteract the dominant length or width mode, or four placed at wall midpoints, average out that variation. The result is that row one and row three actually hear the same thing, which a single sub plus heavy EQ cannot deliver because EQ is applied globally while the problem is positional.

What multiple subs do not do is shorten modal decay. A room mode at 42 Hz that rings for 700 milliseconds still rings after you add a second sub, and the ear hears that ringing as one note bass. That is what bass traps address: thick porous absorbers in the corners where modal pressure is highest, or tuned membrane absorbers when a specific narrow resonance dominates.

Practically: multiple subwoofers for seat to seat consistency, traps for decay, DSP for the residual peaks. Each tool for its own defect.

1. Geometry. Speaker positions, seat positions, screen distance. Free to change on paper, expensive to change afterwards.

2. Structure and treatment. Absorption at first reflection points, corner treatment, decay time brought into range. Drapery, rugs and upholstered seating count toward this and should be coordinated with the designer rather than fought over later.

3. Subwoofer placement and count. Measured, not guessed.

4. Electronic calibration. Levels, distances, crossover, modal correction, target curve.

5. Voicing by ear. The final subjective pass on reference material, both music and film scenes.

Skipping step 2 and asking step 4 to compensate is the most common sequence error in home cinema. It produces a room that measures acceptably on a smoothed graph and still sounds wrong.

At Ermine that last step, the voicing, is done by Anton Chilibi, credited on the Eurovision winning song Stefania (Kalush Orchestra). Measurement makes a room correct. A mastering ear makes it sound right. Keeping it that way over time is part of care.

Can I just buy a receiver with room correction and skip treatment?
Automatic room correction will set levels and distances and improve the bass region. It will not shorten decay time or remove early reflections, so dialogue clarity and imaging stay limited by the room.

How much absorption does a private cinema need?
It depends on volume and existing finishes, but the target is a decay time in the 0.25 to 0.4 second range, reasonably even from about 125 Hz upward. That is measured, not estimated from wall coverage.

Is acoustic treatment always visible?
No. Absorption can sit behind acoustically transparent fabric walls, inside soffits, above a stretched ceiling or within the millwork. In the projects we do on the Westside it is almost never visible.

Does treatment help a room used for music as well as film?
Yes, and usually more. Music exposes decay problems and imaging errors more clearly than film mixes, which carry their own reverberation.

Can you calibrate a system another company installed?
Yes. We measure, treat where needed and voice systems we did not install. Call (888) 818 9996 or use the contact page.