As a follow-up to
this thread where we discussed using M or K lenses on a K10D and the variable results using the 'Green Button', I performed a quick little test.
I don't own a gray card (although I probably should) so I substituted the cardboard back of a spiral-bound notebook. (
Note: This means that the numbers I'm posting are not true 'gray card' numbers! This is not so much an exercise in evaluating the meter's accuracy but rather its consistency.) These are also shot handheld, center-weight average, manually focused with my eye to the viewfinder under dismal overcast skies. It was about 55° F.
I shot the same scene at ISO 100 with the Super-Multi-Coated Takumar 35/2 and the Super-Multi-Coated Takumar 50/1.4. I shot the binder one last time with the SMC Pentax (i.e. the K version) 200/4 at ISO 200 (the last three at ISO 800 due to slow shutters). With all three lenses I fired a shot at every single aperture click. (11 shots with the 35/2, 13 shots with the 50/1.4, and 11 shots with the 200/4.)
The pictures were opened in Photoshop CS2 using default ACR values (not 'Auto'), a linear curve, and no noise reduction, cropped identically to 800x800 center pixels. What I'm reporting here is the median pixel (brightness) value of the resultant crop.
The graph:
A sample of the scene, this one at f/8 with the S-M-C Takumar 50/1.4:
(On the right is a print of
this picture.)
My conclusion is that the K10 experiences a gradual increase in calculated exposure as you stop down. I found it fairly linear and predictable until around f/16. To my tastes, I think the 'correct' number for my test was about 180 - meaning the number that fairly reproduced the scene. (As jeffkrol points out this number doesn't mean much.)
Last edited by carpents; 04-26-2007 at 08:21 PM.
Reason: Pbase is stupid.