Why is most of what you hear not the speaker?

In an untreated room, most of the sound energy reaching your seat has bounced off something first. The direct sound — the part that travels straight from the driver to your ear — is the minority of what arrives, and in a small, hard-surfaced room it is a small minority.

You do not hear those reflections as separate events. They arrive within a few thousandths of a second of the direct sound, and your ear fuses them into one. What they do instead is change what you hear: detail smears, dialogue loses its edge, voices drift away from the actor’s mouth on screen, and the sense of space collapses.

So reflection control is not a refinement applied after the system already sounds good. It is most of the sound.

How is a first reflection point found?

Hang a mirror on the side wall of your cinema and sit down. You will see the speaker reflected at exactly one spot on the glass. That spot — and only that spot — is where sound from that speaker bounces off that wall into your ear. Slide the mirror along the wall and the reflection appears somewhere else on it. Move your head and the spot moves too.

Sound obeys the same rule. Angle in equals angle out.

Two things follow from this, and they are the whole argument:

1. The point can be calculated to the inch, before anything is built.
2. It is a different point for every combination of speaker and seat.

Why do row 1 and row 2 need different panel positions?

Take a worked example. 26′-8″ long, 16′-0″ wide, 11′-0″ ceiling. Two rows, the back row on a 1′-2″ riser.

Those two points are 4′-1″ apart. That is not a rounding error; it is wider than the panel.

Now consider the two things people usually do:

• Cloud centred over the seating. This is the intuitive choice, because that is where people sit. It catches neither point. Both rows keep an untreated ceiling bounce.
• Panels centred on row 1’s point. The front row is treated. The back row is not. Same room, same system, same evening — and the back row gets a narrower image and less intelligible dialogue than the front. Clients notice this. They usually describe it as “the sofa sounds better than the recliners” and assume it is the seats.

The band has to be dimensioned to hold both points, plus the panel width around each. In this room that means a ceiling band running 8′-0″ to 12′-6″ from the front wall, with the front edge fixed on the drawing rather than centred by eye.

How many reflection points does one room actually have?

The ceiling was the simple case. Every hard surface gets the same treatment, for every speaker, for every seat.

On the right-hand side wall of that same room, taking the centre seat of each row and the three front channels only:

Six points spread across a 9′-6″ band. Not a patch beside the speaker, which is where side-wall panels usually end up.

The points have a height as well as a position

Everything above locates points along the length of the wall. They also sit at a height, and the height moves for the same reason the position does.

Where the speaker and the ear are level, the reflection point sits at that same height. Row 1 is at seated ear height, so all three of its side-wall points land at 3′-6″. Row 2 is on a riser, so its ear is above the speaker and its points climb: 3′-9″ from the nearest channel, 4′-1″ from the centre, 4′-3″ from the furthest.

That spread — 3′-6″ to 4′-3″, nine inches across the six points — is why a side-wall band has a dimensioned height and not just a dimensioned length. Nine inches is easily covered. Raise the riser, add a third row, or bring in height channels, and the spread grows until a band whose height was chosen by eye starts missing the points at its edges.

Add the outer seats and that one wall carries 18 points. Across the four reflecting surfaces:

And that is the front three channels alone, before the surround and height layers are counted.

Nobody holds 72 coordinates in their head on site. That is the entire practical reason we draw the room first.

Position is half the job. Depth is the other half.

A panel in exactly the right place, made too shallow, still fails. It absorbs the top of the band and reflects everything underneath, which tilts the room’s tonal balance rather than correcting it.

The variable that matters is not the panel’s thickness. It is the effective depth — panel thickness plus the air gap behind it. A porous absorber works by slowing moving air, and air moves fastest at a distance from a hard wall, not at its face. Mount the same panel on battens and you move it out into that faster-moving air, where it behaves closer to a thicker panel than it measures.

Deeper reaches lower, and the relationship is close to inverse: double the effective depth and you roughly halve the frequency the absorber still controls. The figures below are the trade’s working approximations for specifying, not measurements — what a given panel actually does depends on the material’s flow resistivity and how it is mounted, and the only ways to know are its own test data or a measurement of the finished room.

A 100 mm panel screwed flat to the wall gives 100 mm of effective depth and controls to roughly 250 Hz. The same panel on 50 mm battens gives 150 mm, and reaches about 167 Hz — a third lower, from the same panel. Doubling the effective depth halves the frequency, so a full octave costs a batten as deep as the panel itself. That decision — battens rather than adhesive — is the step most installations skip. The panels are identical. The mounting is not.

Which is why effective depth belongs on the drawing and not only on the panel spec. It changes how far the assembly stands off the wall, and that has to be coordinated with the finish, the skirting and the seating line before anything is ordered.

What about diffusers?

Scattering devices follow the same logic through a different mechanism. The depth of a diffuser’s wells sets the lowest frequency it scatters; everything below that passes as though the device were a flat wall. A shallow diffuser is not a failed diffuser — it is a device with a high cut-off, and it belongs on surfaces where only the upper band needs scattering. A deeper one belongs where the scattering has to reach further down. Depth is a placement decision, not a quality grade.

One thing carries across only partly. Behind a solid, rigid-backed diffuser an air gap changes nothing you can hear: the wells are already the depth, and a rigid back on battens scatters the same as one on adhesive. That holds only for a solid back. A slotted or open-backed device with a cavity behind it becomes a resonant absorber as well as a scatterer, and then the gap is an acoustic decision rather than a styling one. Which of the two you are holding is worth establishing before the batten depth is fixed.

Why is it wrong to absorb everything?

This is where a lot of rooms fail in the opposite direction. Covering every surface in absorption is not the safe option. Your brain expects a room to sound like a room; strip the reflections out completely and the result is oppressive and airless, voices lose body, and the sound collapses inward as though you were wearing headphones.

There is no coverage percentage that answers this. Coverage is solved backward from the room’s target decay time, then checked surface by surface. You start from the decay the finished room has to achieve, work out how much absorption that volume needs to reach it, distribute that across the surfaces carrying reflection points, and verify the result by measurement rather than by area.

Two rooms of identical floor area can need very different quantities. A room with a plastered ceiling, a tiled floor and glass down one wall starts from a completely different place than one with carpet, a fabric ceiling and masonry. A percentage rule cannot see that difference. A decay target can.

Distribution matters as much as quantity. A room can carry the right total absorption and still have one bare wall doing all the damage, because a reflection point left untreated is untreated regardless of what the rest of the room is doing.

The rules that follow

• Ceiling first reflections are absorbed, always. A ceiling bounce arrives at both ears at the same instant, so it pulls the image into a narrow band above the screen instead of spreading it across the front wall.
• The centre of the back wall is absorbed. A scattering panel directly behind the main seat collapses the rear image.
• Scattering belongs in the rear portion of the room, roughly from halfway back, not immediately behind the speakers.
• Absorption and scattering are interleaved through that rear section. All-soft-at-the-front, all-hard-at-the-back is a well-known layout and a poor one.
• The front half is symmetric, left to right. First-reflection treatment on one side wall is matched on the other. Asymmetric absorption at the reflection points pulls the image toward the harder side, and it is one of the fastest ways to lose a centred image on a system that measures well otherwise.
• The rear half does not have to match. Scattering behind the seating can differ side to side. Feeding both ears an identical scattering pattern narrows the sense of width, and the rear half is the one place in the room where deliberate difference helps rather than hurts.
• Floor bounce is the one reflection point you cannot put a panel on. Carpet over underlay is the only treatment the floor can carry, and it is thin — 12–13 mm of carpet on 10 mm of underlay is about 23 mm of effective depth, which by the rule above means meaningful absorption only from roughly 1 kHz upward. That is not the whole floor bounce. It is the part carrying consonants and image focus, which is why a hard-floor cinema is fighting the room from the first day and why carpet is not optional. The rest of the floor bounce is not absorbed at all. It shows up as a cancellation notch, and in a room of these proportions that notch falls somewhere in the 300 to 500 Hz region — well below anything carpet touches. Its frequency is set by the speaker’s height and the ear’s height, which means it can be placed but not removed. That makes both heights a decision taken on the drawing rather than a consequence of whatever furniture arrives.
• Absorptive panels are hung as matched pairs. A single panel hanging alone is visible, and it is usually the sign of a count that was fitted rather than designed.

Why do the same rules not work for bass?

Below roughly 150 Hz the mirror rule stops being useful. At those frequencies the room stops behaving like a set of rays bouncing off surfaces and starts behaving like a resonator: fixed patterns of loud and quiet, set by the room’s own dimensions, standing still in the air.

The consequence is severe and routinely underestimated. A seat sitting in the wrong part of that pattern can lose 20 to 40 dB at a given frequency. A 40 dB loss is 99.99% of the energy gone at that note. Move the seat two feet and the answer changes completely.

So bass is a placement problem too — but what is being placed is the seats and the subwoofers, not the panels. Every room mode reaches a pressure maximum in the corners, which is why the corners carry the deepest traps: that is where they earn their depth at the lowest frequencies. Higher up, through the 100–250 Hz band, the wavelengths are short enough to be caught on the surfaces themselves — front wall, side walls, riser and rear wall — which is where that band is treated instead. And multiple subwoofers in calculated positions do more for bass evenness across a row than any quantity of absorption.

This is also why “we’ll add bass traps later if it sounds boomy” is backwards. By then the seats are fixed, the riser is built, and the simplest fix has been designed out.

What does a calculated design look like on paper?

A room that has been designed rather than decorated produces a specific set of documents before anything is ordered:

• reflection points listed per surface, per speaker, per seat — in feet and inches from the front wall
• treatment bands dimensioned on scaled wall elevations and a ceiling plan, not described in words
• panel type and effective depth against each position, batten depth included, so the installer is not choosing on the day
• coverage checked surface by surface against the room’s target decay time
• a visible flag against any position that had to move for a door, window or opening — moved on the drawing, never silently on site
• setting-out drawings for the site team, so what gets marked on the wall matches what was calculated

If a design does not produce those, the positions were not calculated. They were estimated, and the estimate will be found by the back row.

How much of that a project gets

That is a budget decision, and it should be an explicit one. A full set — every surface, every speaker, every seat, dimensioned and set out — takes design time, and design time is real cost. On a reference-level room it is a small fraction of the total and there is no argument for skipping it. On a smaller room it is scoped down: fewer seat positions calculated, elevations drawn for the surfaces carrying the most points rather than for all of them.

What does not get scoped down is the derivation. Reflection points are either calculated or they are not. A reduced drawing set still starts from calculated positions — it simply documents fewer of them. The failure this article is about is not a thin document set. It is no calculation at all.

A band calculated to catch both rows begins to miss the outer seats first as it drifts, which is why the setting-out is dimensioned rather than described.

Who is this approach for?

There is no shortage of people in India who will build you a beautiful cinema room. Most of them begin from the interior: the layout is set, the finishes are chosen, the lighting is designed, and the acoustics are fitted into whatever space is left over. That order produces rooms that photograph well and measure badly, because by the time anyone asks where the reflection points are, the seats and the ceiling are already fixed.

We work in the opposite order. Room dimensions, speaker positions, seat positions and riser height are settled first, because those four things determine every panel position in the room and every one of them is expensive to change afterwards. The interior is then designed to that geometry, rather than the geometry being squeezed into the interior.

This is not a choice between a room that performs and a room that looks right. It is a question of which one is decided first.

A cinema designed acoustically and then finished properly ends up as both. A cinema designed visually and then treated ends up as one.

So the filter is simple. If the finishes have to be locked before the acoustics are worked out, we are the wrong people to call — not because the finishes do not matter, but because that order forecloses every decision this article is about. If the room’s performance is allowed to set the geometry, and the interior is designed to it, that is the whole of what we do.

What three questions should you ask any installer?

1. Where are the first reflection points for the back row? In feet and inches, from the front wall.
2. What is the effective depth of the panels, and what frequency does that depth control down to?
3. How was the coverage arrived at, and how is it distributed across the surfaces?

If the answers are “we’ll sort it on site”, “standard”, and “as many as fit the walls”, you are buying wall panelling, not acoustics.

The same logic runs through every stage that follows — how a room’s decay target is set, how the room is drawn before anything is ordered, and how the finished result is verified by measurement rather than by ear. If your question is how much treatment a room needs in total, our sister brand covers that in detail: how much acoustic treatment a room needs.

A cinema room is an instrument you sit inside

Every panel position in it is derived from that specific room — its dimensions, its speaker positions, its seat positions, its riser height. Change any one of those and every panel position changes with it.

That is the whole reason we draw the room before anyone buys anything for it.