2
\$\begingroup\$

What makes better sense for situations like: low light, sports/action, wild life?

  • More cross-type sensor on AF (like Canon does)
  • Or more sensitive cross-type sensors (like Nikon does)
\$\endgroup\$
5
  • 4
    \$\begingroup\$ My gut feel is that more points are better for tracking purposes and more sensitive points are better for (extreme) low light. I can't substantiate this so I won't add this as an answer. \$\endgroup\$ Commented Apr 9, 2014 at 7:22
  • 1
    \$\begingroup\$ What do you mean by "better"? More accurate? Quicker acquisition? Better at tracking? The answers to those three question may well be different, and (as already commented) may actually have not very much to do with the AF sensors themselves. \$\endgroup\$
    – Philip Kendall
    Commented Apr 9, 2014 at 8:36
  • \$\begingroup\$ See photo.stackexchange.com/a/37381/15871 \$\endgroup\$
    – Michael C
    Commented Apr 9, 2014 at 12:23
  • 1
    \$\begingroup\$ Just remember that due to the physics involved more sensitive (in terms of what is the minimum size of a lens' maximum aperture with which the point will function) also means less accurate due to the narrower baseline between the light from one side of the lens to the other. \$\endgroup\$
    – Michael C
    Commented Apr 13, 2014 at 0:44
  • \$\begingroup\$ Michael, one question (pardon me if it sounds stupid), with a body supporting sensitivity up to f/8, if I use a f/2.8 lens, do you think the AF will be less accurate? \$\endgroup\$
    – Niranjan
    Commented Apr 14, 2014 at 18:31

3 Answers 3

4
\$\begingroup\$

While is doesn't test the premise you lay out, there are some interesting conclusions in this experiments:

http://nikonrumors.com/2011/03/28/auto-focus-accuracy-a-scientific-cross-brand-analysis-guest-post.aspx/

Such as: Camera brand does not matter. Individual body does matter. Max Aperture matters; around F2.8 is best. both F1.x and slow lenses perform worse. Also brand of lens - the original brand lenses, plus Sigma and Zeiss perform best. 3rd party lenses like Tamron and Tokina does not perform as well.

Maybe if you compare the specs on the body models and max aperture results in this experiment you can derive your answer.

\$\endgroup\$
1
  • \$\begingroup\$ Interesting article, it's nice to find something which confirms your intiuition. Of course this doesn't address speed, tracking and use outside of a controlled environment. \$\endgroup\$ Commented Apr 9, 2014 at 7:24
2
\$\begingroup\$

Limiting strictly at your question:

At the time of writing, Canon 6D has the most sensitive cross-type sensor in the middle rated at -3 EV. Nikon 4Ds is only -2EV.

Most cross-type AF sensors are in Canon 1D X / 5D Mk 3.

If these are the only factors for you to rate the AF performance (generally they are not - see the others' answers), then you have:

For sports/fast moving actions you need many AF points - Canon 1D X

For low light (and rather stationary subjects) you need Canon 6D

\$\endgroup\$
1
\$\begingroup\$

It depends entirely on context. If you need to track something around the scene, the more cross type AF points you have, the better as they will provide more reliable tracking, but that said, having the most sensitive points you can have is also desirable, particularly for low light.

Both give advantages in their own way and both are most useful in different situations. I guess I would say you want as many cross type AF points as you need at the highest sensitivity you can get. If your content is static, then one single high sensitivity AF point is ideal, if things will be moving rapidly, a large number of AF points are needed and so the quantity would then have preference unless it is such low light that the AF points don't function (in which case the preference would go back to sensitivity over quantity.)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.