Film Exposure Calculator (Reciprocity, Filters & Bellows)
Correct a meter reading for reciprocity failure, filter factors and bellows extension. Uses Harman's published factor for each ILFORD and Kentmere film.
Film Exposure Calculator
Background.
Film does not obey the exposure triangle at long shutter times, and this calculator is about the part of the exposure that stops behaving. Below roughly one lux-second, the reciprocity law that says "half the light for twice the time gives the same negative" simply fails: the emulsion forms development centres less efficiently in dim light, so the same nominal exposure lands thinner on the film the longer you spread it out. HARMAN technology, who make ILFORD film, describe it precisely — "if the same total exposure is given to photographic material over a longer period of time then the density of the image generated is lower (effective speed is reduced)" — and, since December 2023, publish a single factor for each of their films so you can compute the correction rather than squint at a graph.
That correction is a power law, not a stop count, which is why a general exposure calculator cannot do it. Harman's equation is Tc = Tm^P, where Tm is the metered time in seconds, Tc is the time you actually give, and P is a film-specific exponent. Their own example: HP5 Plus, metered at 10 seconds, needs 10^1.31 = 20.4 seconds. Notice what the exponent does — at 10 seconds it is a one-stop correction, but at 100 seconds the same film needs 10^2.62 ≈ 417 seconds, a correction of more than two stops. The error compounds, and it compounds fastest exactly where you are least able to bracket, because each attempt costs several minutes.
This page also carries the two other corrections that live only in a film workflow. A filter absorbs light and the exposure has to be extended to compensate: a filter marked 2× is one stop, 4× two, 8× three, and a neutral density marked by optical density costs that density divided by 0.301, so a 0.9 ND is exactly three stops. And at close focus the lens moves away from the film, which dims the image: for a symmetric lens at magnification m, the effective aperture is N(1 + m), so the exposure factor is (1 + m)² — a quarter-stop at m = 0.1, exactly two stops at life size, four stops at 3:1. Every large-format photographer and every macro worker on a bellows needs that number, and no meter applies it for you.
The order the three are applied in matters, and this page gets it right. Filter and bellows first, because they change how long the shutter must actually stay open; reciprocity last, applied to that extended time, because reciprocity failure is a property of the real duration, not of what the meter said. Apply reciprocity before the filter and you will underexpose a filtered long exposure badly. There is also a threshold that is easy to get wrong in the other direction: Harman state that "exposure times of one second or less will not require any compensation", and blindly applying Tm^P below one second would make the exposure shorter, which is not just wrong but wrong in the direction that ruins the frame.
One honest scope note. Harman publishes factors for ILFORD and Kentmere films and nobody else's. Kodak, Fujifilm, Foma, Adox and Cinestill present their reciprocity data in incompatible forms — some as graphs, some as stop tables, some not at all. Rather than invent exponents for them, this page offers a Custom field so you can enter the factor from your own film's data sheet, and it says clearly that colour films additionally suffer reciprocity colour shifts that no single exponent can describe.
What is film exposure calculator?
Reciprocity is the assumption that photographic response depends only on the total light delivered — the product of illuminance and time — and not on how that product is split between them. It is named after the Bunsen–Roscoe law and it holds well for silver halide emulsions over the range of exposures ordinary photography uses. Outside that range it fails at both ends. At very short durations (electronic flash, high-speed work) there is high-intensity reciprocity failure; at long durations and low light levels there is low-intensity reciprocity failure, which is the one this calculator handles and the one that matters for night, astro, pinhole, deep-ND and studio still-life work. HARMAN's technical note defines it as the case where "the same total exposure is given to photographic material over a longer period of time" and "the density of the image generated is lower (effective speed is reduced)", caused by "a reduced efficiency in forming stable development centres with lower levels of light". Their model is deliberately simple and empirical: rather than publishing a curve for each emulsion, they fit a single exponent P to a range of exposure times, so the corrected time is the metered time raised to that power. The exponents are close together — from 1.25 for Ortho Plus to 1.43 for SFX — but the difference matters at length: at a metered 60 seconds, Ortho Plus wants 148 seconds and SFX wants 264. Digital sensors do not have this problem in the same form. A CMOS pixel accumulates charge linearly with time, and ISO 12232, the standard governing digital ISO speed, contains no reciprocity-failure provision at all; what limits long digital exposures is dark current and read noise, which is a noise problem rather than a sensitivity problem. That difference is the entire reason this page exists separately from an exposure-value calculator.
How to use this calculator.
- Pick the direction. Forward — "meter reading → shutter time" — is the usual case: you metered, and you want to know what to set. Backward is for when the exposure time itself is the creative goal, and you need to know what meter reading to engineer with aperture, ISO and neutral density.
- Choose your film. The eleven ILFORD and Kentmere entries carry Harman's own published factor. For any other film, choose Custom and enter the factor from that film's data sheet — do not guess it, and be aware that colour films shift colour as well as losing speed.
- Enter the metered time in seconds, taken with no filter and at normal focus. Convert fractions first: 1/125 s is 0.008, 1/60 s is 0.017, 1/30 s is 0.033.
- Enter the filter cost in stops. If you metered through the lens with the filter already fitted, enter 0 — the meter has already accounted for it. If you used a hand-held meter, add the factors: a 4× orange filter and a 3-stop ND together are 5 stops.
- Enter the subject magnification if you are anywhere near close focus. On a view camera it is (bellows draw ÷ focal length) − 1; on a macro lens the scale is usually engraved on the barrel. Below about m = 0.1 the correction is under a third of a stop and can normally be ignored.
- Read the shutter time, then read the breakdown. The three correction figures tell you where the exposure went — and if reciprocity is doing more than about two stops of the work, consider a faster film or a wider aperture instead, because the uncertainty grows with the correction.
- Round to something you can actually count. Harman's own advice on their 20.4-second example is to "round this result off to give an exposure of 20 seconds". Precision beyond about 5 % is not meaningful at these durations.
- Bracket if the shot matters. Harman warn that at very low light levels "some other variables come in to play such as the accuracy of the light measurement", so "some trial and error may be required". Also expect contrast to rise on long exposures, which may call for reduced development.
The formula.
Three corrections, applied in a specific order.
**1. Filter.** Filters are rated by an exposure factor: 2× means the exposure must be doubled, which is one stop. Neutral density filters are often marked by optical density D instead, meaning transmission of 10^−D; converting, the cost in stops is D ÷ log₁₀2 = D × 3.3219, so a 0.9 density is exactly 3 stops and a 3.0 density is 10 stops. This page takes the correction in stops and shows the conversions in the field hint, rather than shipping a table of filter constants it would have to guess at.
**2. Bellows extension.** At magnification m, a lens of symmetric pupil sits at f(1 + m) from the film rather than f, so its effective aperture is N_eff = N(1 + m). Image illuminance goes as 1/N², so
bellows factor = (N_eff/N)² = (1 + m)² bellows stops = 2 log₂(1 + m)
At m = 0 the factor is 1 and costs nothing. At 1:1 it is 4×, exactly two stops — the number every macro photographer knows. On a view camera you rarely know m directly but you can measure the bellows draw v, and m = (v ÷ f) − 1, so the factor is equivalently (v/f)².
Those two combine into a single multiplier: Tm = Tm₀ × 2^(F + B).
**3. Reciprocity.** HARMAN technology's technical note gives the model directly: "This uses the equation Tc = Tm^P. Where Tm is the metered (indicated) time and Tc is the corrected time. P is a factor calculated following a range of exposure times." Their published factors are SFX 1.43, Pan F+ 1.33, Delta 100 1.26, Delta 400 1.41, Delta 3200 1.33, FP4+ 1.26, HP5+ 1.31, XP2 1.31, Ortho+ 1.25, Kentmere 100 1.26 and Kentmere 400 1.30. Their worked example is HP5 Plus at a metered 10 seconds: Tc = 10^1.31 = 20.4 seconds.
The threshold is part of the model, not a convenience: "Exposure times of one second or less will not require any compensation." This is load-bearing, because for Tm below 1 the expression Tm^P returns a value smaller than Tm — applying it blindly would shorten the exposure, which is exactly backwards. Below and at one second, this calculator applies no reciprocity correction at all.
The order matters. Reciprocity is a property of how long the shutter is genuinely open, so it must be applied to the time after the filter and bellows corrections, not before. A 4-second reading with 3 stops of filter is a 32-second exposure before reciprocity, and it is 32 that goes into the power law, not 4. Getting this backwards on FP4 Plus costs about a stop and a half.
The backward mode inverts the same chain exactly: Tm = Tc^(1/P) when Tc exceeds one second, then Tm₀ = Tm ÷ 2^(F + B).
What the model does not include: high-intensity reciprocity failure at flash durations, which is a separate phenomenon; the contrast increase Harman warn about on long exposures, which is a development decision; colour crossover on colour film, which needs three exponents and colour-compensating filtration rather than one number; and the meter's own accuracy at very low light, which Harman explicitly flag as a source of residual error.
A worked example.
A still life on a monorail: ILFORD FP4 Plus, the subject framed at half life size, a red filter costing three stops on the lens, and a hand-held meter reading 4 seconds. Start with the bellows. At m = 0.5 the exposure factor is (1 + 0.5)² = 2.25×, which is log₂ 2.25 = 1.1699250014 stops. Add the filter's 3 stops and the total pre-reciprocity correction is 4.1699250014 stops — a multiplier of 2³ × 2.25 = 18. So the metered 4 seconds becomes 4 × 18 = 72 seconds of genuine shutter time before the film's own behaviour is considered at all. Now reciprocity. FP4 Plus has a published factor of 1.26, and 72 is well past Harman's one-second threshold, so Tc = 72^1.26 = 218.8965644859 seconds, which the calculator prints as 3 min 39 s. That final step added log₂(218.897 ÷ 72) = 1.6041805004 stops on its own. The whole journey from the meter to the cable release is log₂(218.897 ÷ 4) = 5.7741055018 stops. Two things are worth taking from that. First, the reciprocity correction is not a rounding error: 1.6 stops is the difference between a printable negative and a thin, muddy one, and it is the piece that no meter, no camera and no digital exposure calculator will tell you about. Second, order of operations is not academic. Applying reciprocity to the 4-second meter reading first would give 4^1.26 = 5.66 seconds, then multiplying by 18 would give 102 seconds — less than half the correct 219, an underexposure of more than a stop. Harman's Tm is the time the shutter is actually open, and that is 72 seconds, not 4. As a sanity anchor, switch the film to HP5 Plus, clear the filter and bellows fields and enter a metered 10 seconds: the calculator returns 20.4 seconds, which is exactly the example Harman work in their own technical sheet.
Frequently asked questions.
Why does film need this correction when my digital camera does not?
Which films are covered, and what about Kodak or Foma?
Does the one-second threshold really mean short exposures need nothing at all?
In what order should I apply filter, bellows and reciprocity corrections?
How do I find the magnification for the bellows correction?
Should I trust the calculated time exactly, or bracket?
References& sources.
- [1]HARMAN technology Limited (2023). Technical Information — Film Reciprocity Failure Compensation, v2 (document dated Dec 2023). The source of the model and of every published factor on this page. States: 'This uses the equation Tc = Tm^P. Where Tm is the metered (indicated) time and Tc is the corrected time. P is a factor calculated following a range of exposure times'; 'Exposure times of one second or less will not require any compensation'; the worked example 'using HP5+ at a metered time of 10 seconds gives us: Tc = 10 exp 1.31 = 20.4 seconds'; and the factor table SFX 1.43, Pan F+ 1.33, D100 1.26, D400 1.41, D3200 1.33, FP4+ 1.26, HP5+ 1.31, XP2 1.31, Ortho+ 1.25, K100 1.26, K400 1.30.
- [2]International Organization for Standardization (1993). ISO 6:1993, Photography — Black-and-white pictorial still camera negative film/process systems — Determination of ISO speed. Published, 2nd edition, last reviewed and confirmed in 2023. Scope: 'Specified is a sensitometric method for determining and expressing the speed of photographic negative materials.' This is the standard behind the ISO speed printed on a black-and-white film box, and therefore behind what a meter reading means before any correction on this page is applied.
- [3]International Organization for Standardization (2019). ISO 12232:2019, Photography — Digital still cameras — Determination of exposure index, ISO speed ratings, standard output sensitivity, and recommended exposure index. Published, 3rd edition, confirmed 2024. Cited for the contrast drawn on this page: the digital ISO speed standard contains no reciprocity-failure provision, because CMOS sensors do not exhibit low-intensity reciprocity failure in the form silver halide emulsions do.
- [4]Ray, S. F. (2002). Applied Photographic Optics: Lenses and Optical Systems for Photography, Film, Video, Electronic and Digital Imaging, 3rd ed. Focal Press. Standard reference for the effective-aperture relation N_eff = N(1 + m) for a symmetric-pupil lens, from which this page's (1 + m)² bellows-extension factor follows directly, and for the pupil-magnification correction that applies when the lens is not symmetric.
In this category
Embed
Quanta Pro
Paid features are coming later.
- All 590 calculators remain free
- No billing is enabled