Inescapable logic! A "straight line on log-log (are you sure you don't mean "deci-log"?) paper. Uh .. yeah .. but in real time .. it isn't a straight line ... but a logarithmic curve.
Hopefully I won't have to revert to explaining logarithms here ...
I know that the intensity of the light that reaches the film will determine the time of the exposure. I've known that for some .... Hah!! You thought you'd catch me admitting my age!. Well - a
lot longer than I haven't known that.
It is when the
TIME itself becomes inordinately long that reciprocity comes into play.
I tried that web address - unfortunately, it doesn't work. I have the Unblinking Eye bookmarked somewhere ... I'll give that a try - once I get some sleep.]
ED,
www.unblinkingeye.com should work. The title of the article there is "LIRF is Lurking at Your F-Stop". If you search for "LIRF" on the unblinkineye site, you should find it.
The expression of LIRF correction as a straight line on log paper was determined by curve-fitting actual data, not as a logical theory. It makes no difference whether the log has the base 10 or 2 or 2.71828 the curve shape will be the same. The fact that the slope of the line, for practical use, is 1.62 for all films I had data for means that all you need to know about any given film is the time to be added when the measured time is 1 second, or for that matter, any other time. You can the draw a line through that point with a rise of 1.62 inches, centimeters, miles or kilometers and a unit run in corresponding units.
One thing you will notice about logarithms is that there is no zero. The number whose log is zero is not zero but minus infinity. The logarithm of 0 is 1.0 in any base.
When you use graph paper with a logarithmic scale, you are plotting time to be added against measured time, not their logarithms. It is not easy to get log-log paper these days, and if you cannot get it you can plot log vs log. The line will still be straight. I wrote a program to make my own log-log and semi-log graph papers. I can furnish a CD to set up that program on any version of Windows for the cost of reproduction and shipping.
[My original question wasn't answered ... What correction would/ could be made in terms of EV? An example:
To the Hasselblad lenses ... EV9 is 1 second @ f/22;
and 1/8 second @f/8;
and 1/60 second @ f/2.8. My exposure meter tells me I need EV9. Is there any (at all - no matter how little) reciprocity error / effect between the EV9 given as 1 second @ f/22 and the EV9 as 1/60 second @ f/2.8? The total amount of light that reaches the film will be equal - only the
time will be different.
If so, what correction in terms of EV would be appropriate?