Originally posted by Lowell Goudge ...
For example, the aperture is actually continuously variable, and not changing in steps. the steps reported only correspond to steps in the 1/2 EV indexing used by Pentax in their mount. I would be willing to bet, that the exposure actually changes within the constant F stop range and is really only accurate at the middle focal length within the range. In this way the 1/2 stop steps become +/- 1/4 stop deviations from correct (i.e. perfect) exposure.
...
I tested my 55-300mm with a K20D set to 1/3 EV steps, the camera displays
f-number incremental steps of f/4.0 to f/4.5 to f/5.0 to f/5.6. When set to 1/2 steps the camera displays f/4.0 to f/4.5 to f/5.6.
Note that f/4.5 can't be both -1/3 EV and -1/2 EV changes, so the
f-numbers aren't displayed very accurately. But that's another issue.
When you zoom the lens, it jumps to from f/4.0 to f/4.5 at 120mm and stays there thru 190mm using the 1/3 EV system. Based on your hypothesis, the
f-number would be most accurate at about 120 + (190-120)/2 =
155mm for f/4.5.
Just for reference, if you calculate
f-number based on aperture size:
-1/3 EV change should be f/4.38 rounded to f/4.4
-1/2 EV change should be f/4.62 rounded to f/4.6
-2/3 EV change should be f/4.90 rounded to f/4.9
BB