CEDIA RP-22 small-room cinema calibration — what we measure and why

CEDIA RP-22 — formally the *Recommended Practice for Home Theatre Calibration* published by the Custom Electronic Design and Installation Association — is the closest the residential cinema industry has to a measurement-driven specification. The document defines the measurements that separate a calibrated room from a configured one; the practice of taking those measurements on every commission is what separates a serious cinema integrator from a brand-installer.

Below is a walk-through of the seven core RP-22 measurements we take, what each one tells us, and how the data shapes the design choices made weeks or months upstream — at room-dimension stage, at speaker selection, at finish specification, at projector and screen pairing.

## 1. Octave-band SPL coverage at every seat

The first measurement is sound-pressure-level coverage in third-octave bands from 20 Hz to 16 kHz, measured at every seat position and at the reference listening position. The target is ±3 dB from seat to seat in the speech band (200 Hz to 4 kHz), with no worse than ±6 dB at the room's modal frequencies. A pink-noise reference, a calibrated measurement microphone (Earthworks M30 or equivalent), and Room EQ Wizard or SMAART do the job. Half a day per room.

What this measurement tells us upstream: speaker placement, seating geometry, and the height of the cinema chairs against the speaker baffle line. A room that fails coverage at the rear-left seat is usually a room where the back-row chair height was specified after the speakers were procured. The fix is either chair re-specification, speaker re-aiming, or a second pass of dispersion modelling against the actual seat heights.

## 2. RT60 per octave band

Reverberation time at 60 dB decay, measured per octave from 63 Hz to 8 kHz at the reference listening position and at one off-axis seat. Targets vary by room intent: 0.30 s at 1 kHz for a reference cinema, 0.35 s for a premium room, 0.40 s for a media room. The decay should be linear and the band-to-band variation under 25%.

What it tells us upstream: absorber area, diffuser placement, and finish specification. The most common RT60 failure is a room that measures clean at 1 kHz but rings at 250 Hz — a sign that wall absorbers are 50-mm fabric panels without bass-traps in the corners. The fix is a quad-trap install at every wall-wall-ceiling intersection, sized against the room's modal frequencies.

## 3. Modal behaviour below 200 Hz

Below 200 Hz, room modes (axial, tangential, oblique) dominate the response. The measurement is a low-frequency sweep at the reference seat, plotted against the room's predicted modal spectrum (computed from L×W×H at design stage). A peak-to-trough variation above ±10 dB is a failure; a sustained null at one mode is a worse failure than a peak.

Upstream consequence: this measurement is the most consequential design feedback we deliver. A room that fails modal behaviour is usually a room with parallel walls and a small length-to-width ratio. The fix at construction stage is non-parallel walls (5° splay) or floor-to-ceiling bass-traps in the corners; the fix at calibration stage is sub-woofer placement (corner-loaded vs nulled), digital room correction (Dirac, ARC, Audyssey XT32), or — in extreme cases — a passive bass-trap structural addition.

## 4. Time alignment in milliseconds

Speakers must be time-aligned to the reference listening position to within 0.5 ms. The measurement is impulse-response per speaker, processed for the time-of-arrival at the microphone. The integrator then sets per-speaker delays in the AVR or processor so that all speakers arrive in phase at the reference seat.

Upstream: this is largely a calibration-stage exercise rather than a design driver, but it has one design consequence — the AVR or processor must support per-channel delay above 20 ms (some entry-tier units cap at 15 ms, which is insufficient for a 9.4.6 Atmos room with screen-channel speakers behind an acoustically transparent screen).

## 5. Screen luminance in foot-lamberts

Projected white at the reference seat should measure 16 foot-lamberts for SDR content, 30-40 fL for HDR content with appropriate metadata handling. A 3-chip DLP or laser-phosphor projector paired with a 0.9-gain matte-white screen at 12 ft throw delivers this on a 120-inch 2.40:1 screen comfortably; smaller screens need lower-output projectors to avoid over-driving the room, larger screens need higher-output to stay above the SDR threshold.

Upstream: this measurement decides projector selection and screen pairing, set against the actual throw distance and the room's ambient-light control. We will not specify a projector before the room has a calibrated screen target.

## 6. Ambient-light contamination

A reference cinema should measure under 0.1 lux at the reference seat with the room in its operational state (projector on, screen at black, ambient lighting at the lowest preset). Above 0.3 lux, perceived contrast collapses regardless of projector specification — the projector cannot deliver black if the room is leaking light onto the screen.

Upstream consequence: this measurement is the single most consequential finish-specification feedback. The most common failure mode is matte-black wall paint that measures 6% reflectance rather than the 3% claimed on the can, combined with an LED strip behind the riser that the owner asked to keep on for entrance lighting. The fix is darker finishes, baffled lighting, and an authoritative project briefing on what 'cinema dark' actually means.

## 7. STI for dialogue intelligibility

Speech Transmission Index measures dialogue intelligibility on a 0-1 scale. Reference cinemas should measure STI ≥ 0.62; below 0.50 is a room with intelligibility complaints. The measurement uses a calibrated source and microphone at every seat.

Upstream: STI failure is almost always a room with strong reverberation in the speech band (250-2000 Hz) — a finish problem rather than a speaker problem. The fix is the same as the RT60 fix at the speech band: more absorber area on the side walls, with bass-traps to control low-end build-up that masks the dialogue range.

## What this looks like as a deliverable

On every cinema commission we hand over a calibration report: each of the seven measurements documented per seat, the design choices that drove each measurement, the room's pass/fail against the RP-22 target, and any deviations the owner has authorised. The report is the document that anchors the room over its operational life — a future technician returning for a tune-up has the baseline to measure against, rather than guessing what the room was supposed to sound like.