RustyBug Online Upload & Sell: On
|
p.1 #3 · p.1 #3 · Sharpness Test (non-Scientific) Canon Lenses vs. Carl Zeiss Lenses | |
Interesting, but imo there's a fatal flaw in play with the results.
Since the amount of contrast is a component of the perception, the amount of EV illumination should be the same for all. Recording more / less EV from the reflectance values of the ink vs. paper will yield a different return based on the different absorption / reflectance rate per a different time / aperture if they are not equitable.
Or, think of it this way ... turn the lights down low or off and they will all have the same amount of light being reflected (i.e. none) and the perceived contrast will be the same for all. Conversely, blast them all with a 1,000,000 EV's and they will all return so bright that they will look the same regarding contrast perception.
Granted, these are somewhere between 0 -1,000,000 but with different EV's of illumination / return (time based) there will be inherently different contrast levels recorded ... not necessarily because of the lens, but because of the exposure variance. Granted, different lenses have diff transmission rates may generate different outcomes, but since sharpness is perceived contrast, imo all pieces are to be factored in for the outcome ... without a compensation to offset the contrast of the light transmission.
Admittedly, no matter how one tests, someone will always find flaws in the approach, but I might just run them at the same lighting, same aperture, same shutter for a standardized EV exposure ... and let the chips fall where they may.
Your choice of f/4 or maybe 5.6 seems reasonable (if using the slow zoom kit in the test group). Imo, by the time you get to f/8, the dof impact could be providing false "sharpness"
|