(Non central) Focus errors of Sigma Art 18-35mm/1.8 on Nikon DX cams (esp D500)

[MFloyd]

Andy, congratulations, for the very in depth analysis of the problem. I have Sigma on the radar screen for a few years now.  Though, I never made the step, and sticked with Nikkor lenses. The AF problem described is certainly not an encouragement. Looking further for your additional findings. 👍🏻

[Erik Lund]

Richard,
.....2) The USB dock can let the lens store the AF settings only for one camera. If you have more than one camera you want to use the lens with, it will be hard. You need to make a decision which one should be the one with good focus (with current USB dock: in the central field at least). The other cameras might work, but this is not sure. Just imagine the different locations of AF fields cameras with different AF modules have. The lens has no idea, where the AF locations are. How to correct?
....

As far as I know:

There is no intelligent communication from the camera to the lens.

The AF focus behaviour of the particular lens is communicated to the camera one way, from the CPU to the camera.

[MFloyd]

Slightly off topic: somebody knows what the Nikon Distorsion Control Data comprises as lens data ? I guess (much) more than only distorsion related data ...

[Erik Lund]

Only vignetting correction. If you think of PP?


The rest is done in camera firmware.

[MFloyd]

Thank you Erik. In that case, I will rephrase my question differently. The body receives information from the lens: (1) what kind of information the body receives ?; (2) what information is eventually stored already in the body related to this particular lens ?; and (3) what sort of treatment the body CPU does with the lens data ? Thanks ☺️

[Erik Lund]

The camera does not correct the distortion,,,

[MFloyd]

Thank you Erik. In that case, I will rephrase my question differently. The body receives information from the lens: (1) what kind of information the body receives ?; (2) what information is eventually stored already in the body related to this particular lens ?; and (3) what sort of treatment the body CPU does with the lens data ? Thanks ☺️

Can I turn over my questions to the forum, because I didn't find anything on the web ? Thanks 🙏

[Erik Lund]

Your questions are very broad and thereby related to many processes inside and outside the camera as well Here some of the Lens related data;

The name of the lens is stored with the image, it is used in PP for standard or custom corrections settings like distortion and vignetting along with camera name for further corrections of sensor related colour issues etc. can be used by Nikon software and third party Adobe Camera Raw etc.

The focal length, exit pupil min, max and shooting aperture values used for metering, Matrix metering and comparison to database of images in camera. stored with image

AF-D and AFS lenses, the approximate distance is used for fill flash and stored with image

AF, AF-D and AFS focal length is stored with image

AFAIK there is no treatment, manipulation, changes or calculations done to the lens data by the camera

CPU in camera corrects for sensor data, colour and the electronics etc.

You can see even more in the Exif file,,,

Since your statemet first was related to lens distortion. The mirror-less Leica SL with it's electronic viewfinder, probably the best in the class, show a distortion corrected image in the viewfinder on the fly, quite impressive to see - But disappointing that the new zoom lens has a lot of distortion,,,

[MFloyd]

Thank you 😊 Erik. Indeed my questions trigger a very broad variety of matters.

[kosiniak]

The same issue with Sigma 35mm f/1.4 Art  on D750 (full-frame). When corrected by in-camera autofocus fine-tuning for the central sensor, the left/right most sensor AF was way off.
In D750 the extreme side AF sensors aren't really close to the edge of the frame, therefore I use them frequently. Thus I had to return the lens.