Quote Originally Posted by polyglot View Post
Consider a bright-night shot (with 5-stop range, ought to just fit OK in a print) on Fomapan, 2s indicated exposure. Bright parts (Zone VIII) would have a light level implying a 0.25s exposure for Zone V and therefore no reciprocity failure while dark parts (Zone III) would meter at 8s. At that lower light level, a 3-stop reciprocity correction is required, so what should have been nicely textured shadows at Zone III have now dropped off to completely blank (if you don't do the reciprocity-failure correction). If you apply the 3-stop reciprocity correction for the shadows, they're back at Zone III as desired but your midtones (that only required a 1-stop correction) are now up at Zone VII and your highlights are 3 stops off the scale and completely blown. Net effect: get a huge N+3 expansion just from reciprocity failure in a typical scene, without changing development at all.
Even with exposures at 240 seconds, Bond found no increase in contrast with TMY or Delta 100, no increase in contrast beyond experimental error with Tri-X, and less than a one Zone increase in contrast with HP5+ and TMX. This doesn't include the films with least reciprocity failure like Acros. TMY-2 is decent, but isn't particularly outstanding where reciprocity failure characteristics are concerned.

Have you actually tested Fomapan carefully and gotten the N+3 expansion that you describe? If so, I'd love to see the data. Your logic is correct, but that doesn't appear to be the way current films work in practice, based on the data I've seen from Bond and astrophotographers' testing.

As Bond mentions in his article, most manufacturers' data sheets are woefully out of date and inaccurate where reciprocity failure characteristics are concerned.

FWIW, I actually do sometimes use films with greater reciprocity failure to increase exposure times for things like pinhole work.

Lee