Originally Posted by
Doremus Scudder
Actually, the test should be done with a grey card illuminated by nice, neutral 5600K daylight. Note that the color of the object you read with and without the filter will skew the results. Read the blue sky with and without the filter and you'll get a rather low "factor" (the meter's just seeing blue the whole time - the factor may be close to zero). Try that same test with a red-painted house and the "factor" will be much higher.
The laboratory method of finding a filter factor (described in the my copy of "Wratten Filters" pub. by Kodak Ltd, London, 1953, pp. 9-10) was to expose the film through the filter with a calibrated light sources using an intensity-scale densitometer. Curves are plotted and speed point found. The factor is based on the exposure through the filter that results in the same density as the exposure made without a filter.
The method given for finding one's own filter factor is as follows:
"Select a scene containing a neutral-grey area and make one exposure without the filter. Then, with the filter in position, make a series of exposures of the same scene. Start the series with a lens opening one-half stop larger than the one used without the filter. Progress by half-stops through a range of 2-3 stops. Match the density of the neutral area in the unfiltered negative with the density of the same area in one of the filtered negatives. The factor for that particular filter can then be calculated from the difference in exposure between the two negatives.
It is important that, when matching the density of the neutral area, the gamma values of the two negatives must be roughly equal ; a 25 per cent, increase in development is recommended for negatives made through a dark blue filter to compensate for the loss of contrast.
For very critical work, tests should be carried out using shutter speeds similar to those which are to be used in practice."
My later Kodak Photographic Filters Handbook, pub. Eastman Kodak Co. 1990, has a more abbreviated section on filter factors. The method for determining filter factors given there is:
"Choose a subject having a neutral-gray area, or place a KODAK Gray Card or a photographic gray scale in the scene to be photographed. (In photomicrography, use the illuminated field without a slide or with a clear area of slide.) Make one exposure without a filter. Then, with the filter in place, make a series of exposures, increasing by half-stop increments, through 2 to 4 stops greater exposure, depending on the filter. Match the density of the unfiltered shot with the comparable density on one of the filter series; do this either visually or with a densitometer. Determine the filter factor from the difference in exposure between the two exposures producing equal densities."
Note that there could still be discrepancies due to the difference in spectral response between meter and film.
All that said, instead of determining and applying factors, you can just meter your subject through the filter and generally get good results. Keep notes and add "fudge factors" for specific film/meter/filter combinations (development adjustments might need to be made as well).
Personally, I find spot-metering through the filter after having tested for the above to be more accurate than applying factors.
Interestingly, neither of the two references gives factors for the larger majority of available filters; only for CC filters, the color-conversion filters and the light-balancing filters. There is a ton of transmittance data and graphs, however, from which one could extrapolate starting points for factors.
Best,
Doremus
Bookmarks