Interesting blog post regarding lighting and the influence on autofocus

[pato]

Hi all

I found a very interesting blog post regarding autofocus (AFMA) and lightning conditions:
http://www.reikan.co.uk/focalweb/index.php/2014/01/colour-affect-autofocus/
Here one picture which shows the autofocus result with a Canon EF 85mm f/1.2L

Also, it's just a few results and not yet a full range of many tests. But still, it shows a big variance depending on the used lightning when using autofocus, or doing AFMA.

What do you think of this?

pato

 

[J.R.]

Thanks for sharing.

The optics / physics throw up a number of questions but insofar as I am concerned, the key takeaway from the discussion - obtaining and maintaining AFMA for different lighting conditions for 8-10 lenses - goes a bit too far. YMMV

 

[Rudeofus]

The dependency as shown by the graph is not only on the color of the illuminating light, but also on the color of the object, which means a green leaf will require different AFMA compared to e. g. a red flower.

Nobody ever seems to have noticed, that red and green are 15+ AFMA points apart on the 85L. This chart makes me wonder whether AFMA is really as important as it is frequently claimed to be, especially with respect to real world shooting.

 

[Loswr]

I'm really not sure what to make of this. For example, comparing the three lenses tested in 'daylight' illumination, for the 40/2.8 the red and blue 'best focus' are adjacent and the green is an outlier, for the 24-70/2.8L II the blue and green are adjacent and the red is an outlier, and for the 85L all three are well-spaced. The 24-70 has the largest absolute spread, however, the 85L has the most axial CA of the three lenses, by far.

One thing that I'd be very interested in seeing, given FoCal's recommendations for lighting conditions during testing: is there a difference in FoCal's reported AFMA values for testing in 'daylight' (e.g. full sun, 5500K) vs. tungsten (2900-3200K) lighting.

Worth noting that Rich states, "...there’s not enough data to draw any serious conclusions."