There may not be that much light difference but what about DOF, have you checked that?

[AndrewG NY]

There may not be that much light difference but what about DOF, have you checked that?

How would one verify that with any degree of accuracy?

[AndrewG NY]

bump... any one else with k20d and 50 1.4?

I gave a quick try, meter, ISO 400, Av, aiming even surface (ceiling).

Shot four shots, f/1.4, 1.6, 1.8, and 2.0.

Reviewing on K20D's LCD, I could see significant falloff on the f1.4 shot, somewhat less at f1.6, mostly gone at f2.0. So switching between the shots, there is a visible lightening of the frame between shots.

Reviewing the luminance histogram on the LCD, you can see that the histogram changes shape noticeably, mostly due to the changes in falloff characteristics between the different shots...but it does appear that the rightmost peak is a little bit (really only a couple of pixels) further left on the f/1.4 shot--slightly closer to the center-line marker on the histogram--almost on it--than it is on the f/1.6 shot--so it would appear that f/1.4 underexposes *slightly*.

As I said before, this doesn't surprise me much and I doubt there's anything wrong with your particular lens unless yours is much worse than mine. I've noticed other lenses that don't necessarily expose 100% accurately across the range either. I remember that some of the old Popular Photography reviews actually used to test exposure accuracy with lenses. FA50/1.4 tested at f/1.45 according to them.

Hope this allays your concerns.

-Andrew

[adhall82]

I gave a quick try, meter, ISO 400, Av, aiming even surface (ceiling).

Shot four shots, f/1.4, 1.6, 1.8, and 2.0.

Reviewing on K20D's LCD, I could see significant falloff on the f1.4 shot, somewhat less at f1.6, mostly gone at f2.0. So switching between the shots, there is a visible lightening of the frame between shots.

Reviewing the luminance histogram on the LCD, you can see that the histogram changes shape noticeably, mostly due to the changes in falloff characteristics between the different shots...but it does appear that the rightmost peak is a little bit (really only a couple of pixels) further left on the f/1.4 shot--slightly closer to the center-line marker on the histogram--almost on it--than it is on the f/1.6 shot--so it would appear that f/1.4 underexposes *slightly*.

As I said before, this doesn't surprise me much and I doubt there's anything wrong with your particular lens unless yours is much worse than mine. I've noticed other lenses that don't necessarily expose 100% accurately across the range either. I remember that some of the old Popular Photography reviews actually used to test exposure accuracy with lenses. FA50/1.4 tested at f/1.45 according to them.

Hope this allays your concerns.

-Andrew

Thanks Andrew. So after review of MANY shots to test this issue, it seems that it could possibly be the light fall off/vignetting you are speaking of, but indeed more extreme on my copy; to the point where it appears I underexposed a shot by a stop, and then lightly dodged a small circle in the center of the frame. Oh well, it really is a great lens other than this, it's just a bummer I can't seem to take full advantage of it's intended speed. Maybe I'll come across a better copy down the road. I don't think I feel like sending it back again...

Thanks again for all your input and help.

[AndrewG NY]

Thanks Andrew. So after review of MANY shots to test this issue, it seems that it could possibly be the light fall off/vignetting you are speaking of, but indeed more extreme on my copy; to the point where it appears I underexposed a shot by a stop, and then lightly dodged a small circle in the center of the frame. Oh well, it really is a great lens other than this, it's just a bummer I can't seem to take full advantage of it's intended speed. Maybe I'll come across a better copy down the road. I don't think I feel like sending it back again...

Thanks again for all your input and help.

I understand your disappointment but if you check the reviews at photozone, virtually all the fast 50's (Canon, Nikon, Voightlander, Sony, Zeiss) vignette ~0.7 EV wide open but are substantially better by f/2. It is the nature of fast glass, I'm afraid, even when shot on APS-C. It's hardly the only attribute of these lenses that improves substantially when stopping down a little--resolution & contrast are pretty weak wide open as well.

The problem tends to be even worse on "full frame". (I have no love for the term but everybody knows what it means). For example, the Canon 50/1.8 suffers falloff of -3EV (!!!) on a 5D Mk II.