Metering Infrared Film Through an IR Filter

Your meter cannot see what the film sees

Every handheld meter, every in-camera meter and every phone sensor is built to measure visible light. That is the whole job. Phone cameras go further and carry an IR-cut filter over the sensor, so any near infrared that sneaks in gets thrown away on purpose before it is counted.

Infrared film behind an R72 is exposed by light between roughly 720 nanometres and wherever the emulsion's sensitivity runs out. None of that shows up in a meter reading. So the number on your meter describes a different scene from the one landing on the negative.

Worse, the ratio between the two moves around. Picture a midday summer park. Sunlit leaves bounce back a huge amount of near IR, which is why they print almost white (the Wood effect). A clear blue sky holds very little, so it goes nearly black. Now take the same meter reading in a city square full of brick and asphalt, or on a hazy afternoon. The visible light can be identical while the infrared is a stop or two lower. Your meter will not know.

Why metering through the R72 does not work

Screw an R72 onto the lens and look through an SLR. The viewfinder goes black, or close to it, a dim maroon if the sun is behind you. Your eye is mostly insensitive past 700nm, and the filter passes almost nothing below 720nm.

A through-the-lens meter has the same problem. Bare silicon does respond to near IR, so you might get some reading, but meter cells are filtered and calibrated for visible light, and whatever IR response is left has nothing to do with your film's. The number is noise. Some people report it being close on one camera, then wildly off on the next. Don't trust it.

You cannot calculate around the filter with a simple filter factor either, at least not the way you would with a yellow or orange filter. A filter factor assumes the light behind the filter is a fixed fraction of the light in front of it. With IR, that fraction depends on the scene. Foliage changes it. Haze changes it. The angle of the sun changes it.

The method that actually works

Meter the scene normally, without the filter. Set the film speed on the meter to the box speed or, better, to a low effective rating you have chosen for that film behind that filter. Then expose, bracket, and keep notes.

Two ways to think about it, and they amount to the same thing. Either you rate the film low (Rollei Infrared 400 behind an R72 is commonly shot somewhere around ISO 3 to 25, depending on the light), or you keep box speed and add a stop count. Pick whichever you find easier to keep straight in the field. I prefer the low ISO, since it leaves one less sum to do on a tripod.

Then bracket. Not by a third of a stop. Shoot your best guess, one stop over, two stops over, and one under if the scene is bright. Bracketing wastes frames, but a roll of infrared costs more than most and an unusable roll costs the trip. Write down which frame was which. After two or three rolls in your usual light you will know where your sweet spot sits, and the brackets can tighten.

Lean toward overexposure. A thin infrared negative has no Wood effect worth printing; the leaves come out grey and the whole thing looks like a bad panchromatic shot. A dense one prints.

Which films still see infrared

The honest answer is fewer than you would like. The deep-IR films are gone. What remains is extended-red film: ordinary black and white emulsions whose sensitivity reaches a little past visible red.

The shorter the film's reach, the less light gets through a 720nm filter to expose it. SFX and Retro 80S can produce a real IR look on a bright day with lots of green in the frame. On an overcast one, expect long exposures and modest results. HIE had no anti-halation layer and gave that famous glow around highlights. Nothing on sale today does it quite the same way, and if someone offers you a freezer-kept roll, load it in total darkness.

Focus shift and the red dot

Infrared focuses slightly behind visible light. On a wide lens at f/11 you will never see it. On a 50mm wide open, or anything longer, you will.

Many older manual lenses carry a small red dot, line or letter R beside the main focus index. Focus by eye with the filter off. Read the distance on the scale. Then turn the ring so that distance sits against the red mark instead of the normal one. Screw the filter back on without nudging the ring. Modern autofocus lenses often skip the mark; stopping down to f/8 or f/11 buys enough depth of field to cover the error in most landscape work.

Tripods, long exposures and reciprocity

Rate the film at ISO 6 and Sunny 16 gives you about 1/6 second at f/16 in open sun. Stop down to f/22 for depth of field and it is 1/3. In shade, or late in the day, you will be into several seconds. A tripod and a cable release stop being optional.

Once exposures pass about a second, film starts to lose speed, and that loss comes on top of the filter correction. This is reciprocity failure, and it stacks: first work out the time with the IR rating, then correct that time for reciprocity. Doing it in the other order gets you a different, wrong answer. Our reciprocity failure guide covers how the curve behaves, and the reciprocity calculator will do the arithmetic for a given stock.

Wind matters more than usual too. The trees that give you the white foliage are the trees that move. Several seconds of breeze turns them into soft white smudges, which you may like, or not.

Using Zone Light Meter for infrared

The app has an infrared tool under the Tools tab. You choose your filter (R72, 89B or 87) and it applies a filter factor for it to the visible-light reading. A scene bias slider lets you push the result toward the kind of subject in front of you, since a field of sunlit grass and a stone wall do not reflect the same amount of IR. It also reminds you to shift focus to the IR index.

Treat its number the way you would treat any IR starting point. It tells you where to put the middle of your bracket. The film will tell you if it was right.