What is reciprocity failure in long-exposure film photography?

At long exposure times, film's effective sensitivity drops below what a straightforward metered reading predicts, so the actual exposure needed runs longer than the math alone suggests. Using a commonly published approximation (corrected time equals metered time raised to the power of 1.3), a metered 30-second exposure needs roughly 83 seconds of actual exposure — well over 50 seconds of extra time the simple reciprocity assumption misses entirely.

Why Normal Exposure Math Assumes Something That Breaks Down

Ordinary exposure calculation relies on reciprocity: halving the light intensity and doubling the exposure time produce an identical total exposure, regardless of which one you change. Film holds to that assumption closely across normal handheld shutter speeds, but at extended exposure times — typically starting somewhere around one second and worsening from there — the film's chemistry responds less efficiently to the same total light than reciprocity predicts, and the correction needed grows nonlinearly, not by a fixed percentage.

Why the Correction Isn't the Same for Every Film

The exponent used above (1.3) is a commonly published starting-point approximation, not a universal constant — real reciprocity behavior genuinely varies by film stock, and manufacturers typically publish their own specific correction charts or formulas for long exposures on their own film. A photographer shooting long exposures regularly on a specific stock is generally better served by that stock's own published data than a generic exponent.

Digital sensors don't exhibit this effect in the same way, which is why reciprocity correction is a film-specific concern this site's other long-exposure guidance (ND filter calculations, for example) deliberately doesn't need to account for.