Quote Originally Posted by pinholic View Post
Hi, Thank you for this information. The chart in Pinhole Assist are based on Arista official data which is scarse : (1 second -> 2x, 10 seconds -> 8x, 100 seconds ->16x).
The charts within PinholeAssist for Arista were made with this data, using linear interpolation. I should have used interpolation on the derivative
For t=2s, the 1st method gives 10 seconds, and the 2nd method 5 second.
For t=4s, the 1st method gives 28 seconds, and the 2nd method 16 second.

This is better, but still higher than your figures.
I will release an updated (with 2nd method) data file soon. I would like to distribute an alternate correction curve based on your data, if you authorize me to do so.
You can use my numbers but they might not suit every one. In recent years I have changed my approach to reciprocity failure and sometimes I don't correct for it at all. Why?

A light scene and a dark scene look different. The dark scene shows some details in the highlights but the shadows and lower tones very quickly slide down to black. What I then see is reminiscent of what film will show if reciprocity failure is not corrected. To correct or not to correct (partially or totally) is an aesthetic decision based on whether a dark scene should be represented by a dark photograph.

If I am faced with a light scene but my optical system (pinhole, small aperture, heavy filter) mandates a long exposure then I will apply a stronger correction and use reduced development and low contrast photographic paper to bridge the brightness/darkness gap.