Real world impact of 5d3/1dx af system with slower apertures than f4?

[Marsu42]

What's the real world impact of using aperture lenses of slower than f4 for portrait shooting or edge af on the 5d3/1dx? This really bugs me is that I have this 70-300L/4-5.6 around, and it seems even my 60d's edge af is "better" for this lens because all the outer points on the 5d3 will be only horizontal if I zoom in more than 100mm and the aperture goes beyond f4...

[neuroanatomist]

What really bugs me is that I have this 70-300L/4-5.6 around, and it seems even my 60d's edge af is "better" for this lens because all the outer points on the 5d3 will be only horizontal if I zoom in more than 100mm and the aperture goes beyond f4...

First off, let me clarify - the 70-300mm f/4-5.6L IS lens is a 'Group E' lens.  It doesn't matter what the zoom setting is, you will always have 21 cross-type points in the center of the frame, and 40 horizontal line-type points at the sides, even when you're at 70mm and f/4, because that's how the firmware is written.

Realistically, I think the 70-300mm L will do fine with the horizontal line points - the DoF with an f/4-5.6 lens is pretty deep, the more accurate sensors are primarily beneficial with the thinner DoF of faster lenses.

[Marsu42]

Realistically, I think the 70-300mm L will do fine with the horizontal line points - the DoF with an f/4-5.6 lens is pretty deep, the more accurate sensors are primarily beneficial with the thinner DoF of faster lenses.

The dof might be not razor thin at tele range, but that's no substitute for pattern detection as you know. But you don't have the 70-300L, so I guess you cannot tell how good it really is.

Thanks for telling me about the groups, just looked it up in the 5d3 manual an this isn't specified in any other 1dx/5d3 af reviews I looked at. Even the 180mm macro is a "Group G". Is this Canon marketing at work, or is this due to some real hardware considerations? 

* The 24-70 is Group B, while the older 28-70 is Group A???

* Both the oldest 100mm non-L *non-USM* macro and the new 100L macro are  listed as Group C, while the 100mm non-L *USM* is only "Group E" without cross-point sensors to the side. And why isn't the 100mm L macro "Group A" like the 70-200/2.8?

* What about 3rd party lenses like the Tamron 24-70: What group rating do they get?

* What about the Kenko extenders, do they impact the group rating like the Canon extenders do?

[neuroanatomist]

Thanks for telling me about the groups, just looked it up in the 5d3 manual an this isn't specified in any other 1dx/5d3 af reviews I looked at. Even the 180mm macro is a "Group G". Is this Canon marketing at work, or is this due to some real hardware considerations? 

* The 24-70 is Group B, while the older 28-70 is Group A???

* Both the oldest 100mm non-L *non-USM* macro and the new 100L macro are  listed as Group C, while the 100mm non-L *USM* is only "Group E" without cross-point sensors to the side. And why isn't the 100mm L macro "Group A" like the 70-200/2.8?

* What about 3rd party lenses like the Tamron 24-70: What group rating do they get?

* What about the Kenko extenders, do they impact the group rating like the Canon extenders do?

Macro lenses are generally bumped down a group or two because at macro distances the apparent aperture is narrower.

I'm sure there's a technical reason the f/2.8 lenses in Group B are there and not in Group A, either strong vignetting or something else affecting the off-center f/2.8 diagonal crosses - keep in mind that those outer f/2.8 crosses are right at the extreme edges of the AF sensor.  Clearly, they designed the new 24-70 II to work around that limitation.

3rd party lenses spoof the camera by using a Canon lens code, so I expect the lens group will depend on what lens code Tamron et al. choose.  That issue has caused problems for Tamron in the past, and some of their lenses that use older Canon lens codes result in the off-center AF points on the 40D/50D/60D/7D operating as horizontal lines instead of the crosses (more here).

For 3rd party extenders, it depends on whether they're 'reporting' or not.  If they correctly report the aperture to the camera (newer Kenko ones do, for example), then that will change the lens group.  The older TCs, ones that allow an f/5.6 lens with TC to attempt AF on a non-1-series body, will not change the lens group (but the AF performance may suffer).

[awinphoto]

Real world is that the lens should be fine... it looks better on the 60D because it's a crop camera and getting the sweet spot of the lens...  On FF you get the extreme edges of the lens, and that will vary from lens type to lens type...  In this situation it kinda shows the weakness of the zooms vs the strength of the L prime teles, but by all means the 70-300L is still admirably good for a zoom.  On the 5d3 and 1dx, teleconverters maybe an issue, but otherwise it should be a pretty decent lens.  Other lenses may or may not work as well as your 70-300L, but that's another topic.