A projector's white point is measured with a probe at the seats, and the standard says what it must be, so a room that reads wrong gets adjusted. Screen 2's probe reads warm, the colour of an old tungsten lamp, off a xenon projector that measures correct at the lens, and the screen is a sheet of white vinyl that adds nothing.
I am a projection engineer who does colour calibration for a group of cinemas, and the probe is the whole job.
How the measurement works.
You put a test pattern on the screen, a white field, and a spectroradiometer on a tripod at the reference seat reads the light coming off the screen. It gives you a colour temperature and a set of coordinates, and the standard says where those should be for the picture to look the way the film's colourist saw it.
If it reads off, you adjust the projector's colour processing until it reads right. Then you read it again from a couple of other seats to make sure the screen is even.
You can also read at the lens, with a different attachment, which tells you what the projector is sending before the screen has anything to do with it.
Screen 2 at one site reads warm at the seats. About three thousand two hundred kelvin, which is the colour of an old incandescent lamp, when the standard wants something close to sixty five hundred.
At the lens the same projector reads sixty five hundred, correct.
Between the lens and the seat there is a screen. A cinema screen is a sheet of white vinyl with a gain, which means it reflects light back a little more efficiently in some directions than others. It does not change colour. A white screen is white because it reflects every wavelength about the same, and that is what makes it a screen.
Sixty five hundred goes onto it. Three thousand two hundred comes off.
What has been checked, and I have been back to that room four times because it bothers me professionally.
The probe. Two of them, calibrated, and they agree with each other at the lens and in every other room in the building.
The screen. Replaced in 2022 for an unrelated tear. New vinyl, same manufacturer, same gain. Three thousand two hundred.
The room. Warm light reflected off wood or red walls could tint a reading if it were bouncing back onto the screen, and screen 2's walls are black and its seats are black and the house lights are off for the reading.
The picture. I have adjusted the projector so the seat reading is correct, which means sending a light so blue at the lens that the test pattern at the lens reads like a winter sky, and the seat then reads sixty five hundred, and the film looks right from the seats.
I have left it like that. The projector in screen 2 sends a picture that would look wrong in any other room, and the room turns it into the right one.
Where the warmth comes from.
A spectroradiometer gives you more than a colour temperature. It gives you the spectrum, the amount of light at each wavelength. A xenon lamp has a spectrum with particular features in it, spikes, that I know by sight.
The spectrum at the seats in screen 2 does not have them. It is smooth, the curve of a black body at three thousand two hundred, which is the curve of a filament, or a flame, or a fire. The light that arrives at the seats is not xenon light that has been tinted. It is a different light, from a different kind of source, coming off a screen that a xenon projector is shining onto.
I have shown that spectrum to two colleagues without saying where it was from. Both said it was a tungsten projector, and one asked which museum.
What is done about it.
Screen 2 is calibrated correctly at the seats and the paperwork is in order, and the film looks as it should, and the audience is fine.
The projector's colour settings for that room are so far from any other room's that the manufacturer's software flags them, and I have a note in the file that says compensating for room characteristics, and that is true, and nobody has ever asked what characteristics.
What I do.
I do the calibrations on the cycle and I keep the spectra.
The one thing I have changed is that I take the seat reading with the tripod and go back to the booth, rather than sitting beside the probe while it integrates.
A reading takes about a minute and the room is dark and the white field is on the screen, and you sit and wait.
In screen 2, in 2023, I sat beside the probe in the reference seat and looked at the white field for the minute, which I have done a thousand times, and the white on that screen was warm, the way the probe said, a warm white with a slight movement in it, a flicker at the edges, the way a picture flickers when it is being projected through a shutter off a flame, and I could hear, under the projector's fans, from somewhere in front of me and not behind, the sound of a mechanism with a claw in it pulling something through a gate.