Theaters calibrate projectors to a fixed numeric target, not to taste. A technician projects standard test patterns, measures the light coming off the screen with a meter, and adjusts the projector until white at screen center reads 48 cd/m² at the DCI white point, with the red, green and blue primaries landing where the standard says they should. Opening night adds urgency but not a different procedure. The same measurements that govern a Tuesday matinee govern a sold-out premiere — what changes is how recently someone checked them, and whether anyone looked at the edges of the screen as well as the middle.
Table of Contents
- The number every auditorium is dialed to
- Why color setup happens in the room, not the factory
- Center brightness passes, edges fail
- What the standards do not tell a theater to do
- The sequence a technician actually follows
- What a filmgoer can do with this
- Frequently Asked Questions
The number every auditorium is dialed to
The anchor measurement is screen-center white luminance. The DCI Compliance Test Plan requires SDR white at screen center to read 48.0 ± 3.5 cd/m², with chromaticity coordinates of x = 0.314 ± 0.002 and y = 0.351 ± 0.002. Luminance is brightness per unit area; chromaticity is the pair of coordinates that fixes the exact shade of white, so a screen can be bright enough and still be the wrong color. That white is not eyeballed off a movie. It is measured on a defined full-white test pattern carried in a DCP — a Digital Cinema Package, the encrypted file set a theater receives in place of a film print.
The pattern has fixed code values of X′=3794, Y′=3960, Z′=3890, per the same DCI test plan. Sequential contrast is then calculated straightforwardly: white luminance divided by black luminance. SMPTE, the standards body for motion-picture engineering, states the same figure in friendlier units. Its 20F study-group report gives Reference White Level as 48 cd/m², or 14.00 foot-lamberts, with a theater tolerance of ±10.2 cd/m² — roughly ±3 fL. So an auditorium reading anywhere from 11 to 17 fL is nominally in spec, which is a wider band than the DCI number alone suggests.
Why color setup happens in the room, not the factory
Brightness is only half the job. The projector also has to be told what red, green, blue and white actually look like once the light has passed through its lens, through the projection-booth port glass, and onto that particular screen. A technician measures each projected primary with a colorimeter and types the resulting x/y values into the projector. Christie's CP4230 manual describes the workflow plainly: the primaries are measured from the center of the audience viewing location, and the projector derives its Target Color Gamut Data from the Measured Color Gamut Data file the technician creates. The measured file describes what the room does to the light; the target file describes where the picture is supposed to end up.
Christie's Real|Laser FAQ is direct about doing this onsite rather than trusting factory data — the screen, the port glass and the room all change the result. The same document notes that the projector automatically disables all color correction while the MCGD page is open, which is what makes the measurement honest: you are reading raw light output, not light that has already been corrected. There is a quieter decision buried in this step. The DCI target is selected by filename on the projector — Christie's CP2230 setup guide names "DC28_DCIXYZE_314_351" for the DCI white point — and content providers may specify which target applies. The meter reading does not decide the final look on its own; the file chosen alongside it does.
Center brightness passes, edges fail
A screen that measures perfectly at the middle can still be visibly dim at the corners, and the standards account for this. The SMPTE report cites ST 431-1-2006 for uniformity: luminance across the screen is nominally 85% of the center value, and must be no less than 75% of peak at the sides and corners. This is the measurement most likely to go unchecked, because it takes longer.
Center white is one reading; uniformity means walking the screen and sampling multiple positions. For a wide auditorium with a large screen, that is where falloff hides. For a viewer, the practical tell is a bright center with dimmer upper corners during a full-screen white or near-white shot — a snowfield, a bright sky, a title card on a light background. Letterboxed dark films hide the problem almost completely.
What the standards do not tell a theater to do
The standards fix targets. They do not prescribe a pre-premiere schedule, and that gap is worth being honest about. DCI's plan defers the specific screen-uniformity sampling positions to the ICDM's IDMS document and specifies no warm-up period before measurement. The warm-up and positioning conventions that technicians use — run the lamp roughly ten minutes, measure from the middle of the auditorium — come from practitioner guidance, not from a standard.
That matters because an unwarmed lamp and a warmed one do not read the same, so two technicians following the letter of the spec can still get different numbers. Nor is there one blanket figure for every room. SMPTE formed its 20F On-Screen Light Measurement Study Group precisely because measurement practice was inconsistent, and the resulting report separates three cases: review rooms and premium cinemas, general cinemas, and 3D. A 3D presentation loses a large share of its light to the glasses, which is why it gets its own treatment.
The sequence a technician actually follows
The order of operations is not arbitrary. Measuring color before the projector is in a fully calibrated mode produces numbers that describe nothing useful, because the color-correction process itself alters light output — a point Konica Minolta makes explicitly in its guidance on measuring cinema screen luminance.
A typical pass runs like this: Konica Minolta also advises that operators set a written periodic luminance-measurement procedure, based on the projector manufacturer's recommendation. That is the real difference between a well-run house and a poorly run one: not whether anyone ever calibrated the room, but whether there is a schedule and a record.
- Put the projector into a fully calibrated mode for white point and primaries before taking any reading.
- Warm the lamp — about ten minutes is the common practitioner figure — so output has stabilized.
- Project the full-white test pattern from the compliance DCP and measure at screen center, targeting 48 cd/m² at x = 0.314, y = 0.351.
- Measure red, green, blue and white from the center of the audience seating, and save the measured gamut file on the projector.
- Confirm the correct DCI target file is selected, since the content provider may specify which one applies.
What a filmgoer can do with this
If a presentation looks dim or oddly tinted, the useful complaint is specific. "The image looked dim" invites a shrug; "screen 6 looked well under reference white, and the upper corners were noticeably darker than center" names two separate, measurable faults a manager can hand to a technician. Premium-format screenings are not automatically better calibrated, but they sit in a different category in SMPTE's own framing — review rooms and premium cinemas are treated separately from general cinemas.
For a film whose look you care about, a 3D screening is the one most likely to disappoint on brightness, because the glasses take light the standards cannot give back. Finally, be aware of the tolerance you are living inside. Two auditoriums in the same multiplex can both pass — one at 11 fL, one at 17 fL — and the dimmer of them will be delivering noticeably less light than the colorist graded for.
Frequently Asked Questions
Is a theater required to recalibrate before every opening night?
No. The DCI compliance plan sets measurement targets but specifies no pre-premiere schedule or even a warm-up period. Konica Minolta's guidance recommends that operators establish their own written periodic measurement procedure based on the projector maker's recommendation.
Why is 3D treated differently?
SMPTE's 20F study group separates review rooms and premium cinemas, general cinemas, and 3D into three cases, because one blanket light-level number does not govern every auditorium.
Can a projector be calibrated using the manufacturer's factory data?
Christie recommends creating the measured color gamut file onsite for cinema rather than relying on factory data, because the screen, the port glass and the room all change the measured result.