I might be able to get some tests in Friday in full daylight if it is not rainy here. I have a decent selection of lenses to choose from, including the 40mm f2.8, so I can use that if you want. BTW, the 40 f2.8 and my 43 LTD both need max adjustment on the K5 for good focus in daylight. The 43 at max adjustment is still not quite perfectly adjusted and is FF, so this lens has no room for more adjustment on the K5 for low light. The rest of my lenses are closer to zero, but all needed some adjustment.
I have a Sekonic handheld meter that I can use for an incident reading of the light falling on the subject.
Ray, that sounds good. We need at least two data points as to illuminant colour temperature with the same tests to prove what it is. I see no problems continuing to use your 16-50 f/2.8 lens for testing (I assume that's what you used), since you know it and since it does not require so much AF micro adjust normally.
Best regards, GordonBGood
Daylight test:
K20D, FA 50 f1.4, set at f2.8, ISO 100
Light level reading: *
Sekonic L-358 incident reading at target position - 1/200 f11 ISO 100
Color temp readings - *
ACR Raw color dropper tool measured on the white card of a 4 card WhiBal cardset.
Cardset located at the target location and illuminated by direct sun over my left shoulder located on average 93 million miles from the target
(3 images shot):
1. 5000K
2. 4950K
3. 4700K (shot through the .9, .6, .6 ND filter stack)
More fun with interesting photographic subjects:
Both images exposed at 1/3200 f2.8 ISO 100 and developed in ACR with default settings except color balance set to 5000.
Note that the first image was focused with no filters over the lens and is pretty well centered focus-wise.
The second image was refocused through the ND filters (.9 + .6 +.6 stacked or 7 stops less light if my calculations are correct) but exposure was left the same.
Note the shift to FF at when the focus was locked in at the lower light level entering the lens (second image).
For anyone wondering about the focus lock/exposure method I used, it was as follows:
2 second delay on.
Focus normally (center point) and shoot an image.
De-focus manually and re-focus with half-press of the shutter (no filters over the lens)
Place filters over the lens and half-press to re-focus, hold half-press.
Remove filters and take image by completing full press.
The reason for this rather than trying to focus AND capture the image through the filters is that I wanted to equalize all other parameters in the exposure as much as possible. Since the camera meters wide open, placing the filters over the lens and then trying to meter changes the entire exposure calculation.
Note that I could remove the filters and half-press and the lens would shift focus. Putting them back and re-focusing caused the lens to shift back again.
This is just like the behavior seen in the video of the K5 using the led light on the target.
It has clouded up here a bit so I probably cannot get the K5 out and duplicate this test, but frankly, I do not need to. In every case where the light was reduced by placing ND filters over the lens, and at all color temperatures, the focus shifts to FF with both cameras.
I do not recall seeing lots of problems in my images with the K20 due to this, but it could be that my lens calibrations were done in light that was not too bright and not too dim, giving me a compromise calibration. I also almost always shoot somewhere near f5.6 indoors.
It could also be that my tests are not perfect or well controlled or have a less than 100% correlation to normal shooting conditions.
Ray
. If the lens was a f/2.8 aperture lens and the lighting was of a daylight nature, than it shows that the K20D doesn't have a low light issue for this colour of illuminant . I suspect that it will have a front focus issue for tungsten lighting no matter the lighting level just as my K200D has, and just as recently demonstrated for the K-r by alberto_b.
