[Edgar_in_Indy]

The monitor has nothing to do with it, because monitors are consistent.

The slight exposure differences between images A and B are and were more noticeable on my 24" monitor than on my laptop or work monitors. I can't explain exactly why that would be, but that was my observation. I have also noticed that viewing angle of the LCD has an effect on how noticeable the slight differences are.

[Jewelltrail]

Edgar in Indy: The slight exposure differences between images A and B are and were more noticeable on my 24" monitor than on my laptop or work monitors. I can't explain exactly why that would be, but that was my observation. I have also noticed that viewing angle of the LCD has an effect on how noticeable the slight differences are.

Aha--now I see what you mean with repsect to monitors.

Yes, I agree that between monitors you are going to see a difference--we are on the same page here. What I mean is, a half stop difference between one shot, compared with a half stop difference on another shot, will be consistent on the same monitor. My point being that using the same monitor to compare various light differences will be accurate, consistent, across the board of aperture ranges.

But, clearly, viewing the same shot on different monitors is going to yield different results--absolutely. That is why I still prefer and use the older CRT monitors for picture viewing: the colors are more accurate, and screen brightness is the same, no matter the vertcal plane from which you view the monitor.

Thanks for your in depth explanation.


99% of my shooting is in the manual mode as well--it is very rare I leave it, only for something special.

[Edgar_in_Indy]

Okay, as requested I am posting samples from the two lenses at f/2.8, f8, and f16. Unfortunately they are not at 50mm as requested.

I already returned the 17-50mm, so I cannot take any new pictures. For the tests I conducted when I still had both lenses, I chose 35mm, thinking that I wanted to avoid the extreme focal lengths of either lens. Shooting at 50mm would be at the very extreme end of the 17-50mm and somewhere in the comfortable mid-range of the 28-75mm. I don't know whether this would have any bearing on the tests, but some lenses are known to perform worse at their extremes.

Both lenses have 35mm markings, which I tried to use, but looking at the EXIF data shows that the 28-75mm was at 38mm, while the 17-50mm was at 34.4mm. I don't think this should effect the results.

Lighting was provided by two Paul C. Buff White Lightning Ultra600 monolights with softboxes. The lights were positioned to the left and right of the still life, and their position and output settings remained constant the entire time. They were triggered using a hotshoe-mounted radio transmitter/trigger set.

The shoot was conducted in a basement with no windows and no natural light.

The camera was mounted stationary on a tripod, and ISO was set at 200 throughout the shoot and white exposure was set for flash lighting. I was careful to not move the camera when changing lenses. I used a wireless remote and 3 second mirror lockup for each shot.

In order to maintain proper exposure as I changed aperture settings, I would turn the shutter speed down by one click every time I closed the aperture a click. So, for example, at f/2.8 shutter speed was 1/13, and at f/3.2 it was 1/10. Shutter speed settings were identical between lenses for each aperture setting.

Please note that the focus point ended up being slightly different between the two sets of images, so these images should not be used to compare sharpness of the two lenses.

Here is a picture of the testing setup:

And here is a link to a directory where you can download the original RAW files:

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...mm_vs_28-75mm/

And in case you prefer, here are direct links to each of the files:

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...-50mm_f2_8.PEF

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...-75mm_f2_8.PEF

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...17-50mm_f8.PEF

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...28-75mm_f8.PEF

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...7-50mm_f16.PEF

https://oncourse.iu.edu/access/content/user/ealucas/Filemanager_Public_Files...8-75mm_f16.PEF

[Edgar_in_Indy]

And one more note, in order to figure out the effective max aperture of the 17-50mm lens, I took another series of pictures where I stepped down the aperture, but did not adjust shutter speed to compensate. I then compared the two sets of images and saw that the exposure of the 17-50mm images at f/2.8 fell somewhere between the f/3.2 and f/3.5 images from the 28-75mm lens. I used the same method to determine that the difference was a full stop by f/10.

[Entropy]

A comment on the original images:

You indicate different focal lengths in use, but the subject appears to be about the same size in each image, which means that the camera-to-subject distance changed. This really renders any comparison invalid.

[Edgar_in_Indy]

Actually, the longer focal length images are slightly zoomed in. Look how much empty space there is under the SD card at the bottom of the frame in the 34.4mm images versus the 38mm images.

But for the purposes of this test, which is testing exposure at different apertures, I don't see how it would make any difference. The camera is in the same location, angle remained the same, and lighting and exposure settings were maintained for each lens. If there's something I'm missing about how such a small focal length difference would render these exposure tests invalid, could you please explain?

Feel free to ignore my results and perform your own tests. I would love to see whether others with the 17-50mm get the same results.

[paperbag846]

I wouldn't mind access to some web-sized images instead of .pefs...

The different aperture settings are to determine whether the exposure error is due to the camera incorrectly setting the aperture with the aperture lever. If this is so, the difference in exposure should be increasingly large at the smaller apertures.

Thanks for the work .

[Edgar_in_Indy]

I wouldn't mind access to some web-sized images instead of .pefs...

I'm at work now until late tonight, but I may post some reasonably size jpegs when I get home. I wanted to get the raw files out there so that there was no question about whether images had been manipulated.

[Lowell Goudge]

I have not read all the posts but I can report that the 28-75 ( at least my copy) suffered from a gradual exposure drift perfect wide open to one stop over at F32

I test in Av mode at each aperture against a block wall

You should really test all your lenses this way