Originally posted by khenna From the In-depth review of the 16-85 lens:
Exposure is somewhat uneven across the aperture range. Wide-open the lens underexposes. Stopping down half a stop increases exposure to being correctly exposed. Exposure then drops back into underexposure again until F16, where exposure increases some.
Overall it seems like under exposed. But I mean really who should expect an electronic sensor to expose exactly the way our brain would interpret the original scene on side or even more complex how it should be exposed from an artistic point of view?
The photographer should not accept the default exposure all the time and should instead choose the one that is really a match from an artistic point of view. With modern sensors the key thing is to avoid over exposure while a bit of under exposure is nothing to wory about. A bit of post well done post processing create the mood, style and help subtain the point/subject of the photo.
Technically, It look like the 16-85 would just benefit of +0.3/0.5 exposure by default on camera but this might be a lot linked to the metering mode and exact scene.
But really while the test has value, I would say to really state there an issue, the testing should be done on a few different scenes with different lighting/colors/tones and with different exposure mode. To really see what's going on. To see if the exposure bias doesn't come from camera, one should also try with another lense (like the 18-135) to see if that's not just the camera fault here.