RF800mm/f11 and size of AF-area on EOS R3/R5/R6/R7/R10

Sittatunga wrote:

RDM5546 wrote:

Stig Nygaard wrote:

But how does one explain the 80% x 80% coverage on the EOS R3. Some other factor than sensor size must be involved too (in general)?

It is not the sensor size per se but the individual sensor design that matter. The R3 uses a stacked CMOS image sensor design using a radically different technology than the R5 or R6.

I think it's a geometric thing, a combined effect of the angles of the light from the opposite side of the lens, the relative sizes of the microlenses and the photocells and the distance of the microlenses in front of the cells. I'd expect the patches over which the DPAF works properly would be very roughly similar absolute sizes for the front illuminated sensors (which makes it a bigger proportion of the cropped sensors) and noticeably larger for the back illuminated stacked sensor in the Eos R3.

Thanks RDM5546 & Sittatunga. So sounds like technical reason could be related to reason you can get magenta vignette when using extreme wide lenses like f.ex. the Venus Laowa 9mm f/5.6 FF? I understand that vignetting depends on thickness of the layer above the sensor's cells, and that stacked sensors in general perform much better than non-stacked (I have the Laowa 9mm and I have played a bit with it on the R6. The results have strong colored vignetting, though I'm able to correct it pretty well in post-processing ).

Well, another reason to hope for a stacked R7 some day. But for now happy to know AF area using the RF800/11 at least will be bigger on (my hopefully soon to get) R7 than it is on the R6.

/Stig (Copenhagen, Denmark)
https://www.rockland.dk/
https://flickr.com/photos/stignygaard/
Flickr Fixr: https://www.flickr.com/groups/flickrhacks/discuss/72157655601688753/
xIFr, the Firefox exif-viewer: https://addons.mozilla.org/firefox/addon/xifr/