topic Re: AF Point Sensitivity with Narrow Aperture Lenses in EF & RF Lenses

AF Point Sensitivity with Narrow Aperture Lenses

Waddizzle — Wed, 30 Nov 2016 15:14:56 GMT

I guess that I've been thinking too much again.

I've run up against one of those, "Now how is that supposed to work," issues. Except for the high end models, most of Canon's DSLRs cannot focus at f/8. The cameras seem to top out at f/5.6.

But, some lenses, like a Sigma 150-600mm f/5-6.3 DG OS HSM Contemporary, can stop down to f/6.3 at the long end of the zoom range. So, how is the AF systemsupposed to work properly when the lens stops down that far?

I've used the "big Siggy" on camera bodies that focus to f/5.6 and others that can focus at f/8. The f/8 camera bodies seem to focus more accurately at maximum zoom than the f/5.6 bodies. Granted, the focus systems are not the same, which probably explains most of the differences.

But, how are the f/5.6 bodies supposed to auto focus when the lens itself stops down to f/6.3? It seems to work, but why?

Re: AF Point Sensitivity with Narrow Aperture Lenses

kvbarkley — Wed, 30 Nov 2016 15:17:46 GMT

I have heard that the Sigma/Tamrons simply tell the system that they are at F5.6 during autofocus.

Re: AF Point Sensitivity with Narrow Aperture Lenses

Waddizzle — Wed, 30 Nov 2016 15:23:01 GMT

@kvbarkley wrote:
I have heard that the Sigma/Tamrons simply tell the system that they are at F5.6 during autofocus.

Oh, so they cheat. Well, that's a sensible explanation.

Do you think that might explain why images shot with f/5.6 bodies seem softer than those with f/8 bodies?

Re: AF Point Sensitivity with Narrow Aperture Lenses

kvbarkley — Wed, 30 Nov 2016 15:59:14 GMT

I would think it the other way 'round. The problem with focussing with small apertures is that there is a large "dead zone" because of the depth of field, which I guess gives the control algorithm fits.

Re: AF Point Sensitivity with Narrow Aperture Lenses

Waddizzle — Wed, 30 Nov 2016 16:16:17 GMT

@kvbarkley wrote:
I would think it the other way 'round. The problem with focussing with small apertures is that there is a large "dead zone" because of the depth of field, which I guess gives the control algorithm fits.

Yeah, that makes sense, too. That might explain a lot of things. If the camera thinks the aperture is f/5.6 when it atually is not, I can see how that might throw things off just a little. The hyperfocal distances would differ significantly.

Re: AF Point Sensitivity with Narrow Aperture Lenses

jrhoffman75 — Wed, 30 Nov 2016 16:47:05 GMT

http://www.the-digital-picture.com/Photography-Tips/canon-eos-dslr-autofocus-explained.aspx

See "Lines, crosses and double crosses" section.

Re: AF Point Sensitivity with Narrow Aperture Lenses

Waddizzle — Wed, 30 Nov 2016 18:10:44 GMT

@jrhoffman75 wrote:
http://www.the-digital-picture.com/Photography-Tips/canon-eos-dslr-autofocus-explained.aspx

See "Lines, crosses and double crosses" section.

Most of the stuff in the entire article I am already familiar with, which is what prompted the question in the first place. The explanations of ALL of the different AF sensors went into more detail than before. I think this excerpt....

"Note that these thresholds are not absolute - a lens with a narrower aperture than the threshold might still work, but at reduced effectiveness, accuracy, and speed. Thus, Canon limits the functionality to the rated aperture for a given AF sensor. However, some third party lenses (e.g. Tamron and Sigma zooms with a max aperture of f/6.3 at the long end) effectively trick the AF system into thinking there's an f/5.6 lens attached."

... seems to confim my observations that f/5.6 AF point bodies are giving softer images than f/8 AF point bodies. It may also explain why the lenses occasionaly seem to hunt more at the long end, too.

I have previously heard about how some 3rd party lenses with narrow apertures fool the camera body into thinking that it using a wider aperture lens. The only problem with that suggestion is that the aperture is still somehow reported correctly in EXIF data.

Re: AF Point Sensitivity with Narrow Aperture Lenses

diverhank — Wed, 30 Nov 2016 21:35:17 GMT

@Waddizzle wrote:

But, how are the f/5.6 bodies supposed to auto focus when the lens itself stops down to f/6.3? It seems to work, but why?

I think it's because Canon made an arbitrary decision not to allow focus beyond f/5.6 for their lower end cameras to prevent people complaining about the AF systems. There is no physical limitation say at f/8, just that it doesn't always focus well. For this reason...the Sigmas and Tamrons can report to the camera AF something else like f/5.6 (and report to exif the correct value of f/6.3) and the AF system is fat dumb and happy autofocusing away...

This is the same reason you can tape off the 3 pins on the teleconverter and fool the cameras into letting you focusing at f/8. I used to do this a lot with my old 50D.

Re: AF Point Sensitivity with Narrow Aperture Lenses

kvbarkley — Wed, 30 Nov 2016 21:43:40 GMT

Arbitrary might be a bit strong. I would hope it would have to do with loop gains and stability and margins to future compatibility.

Re: AF Point Sensitivity with Narrow Aperture Lenses

ebiggs1 — Thu, 01 Dec 2016 14:23:02 GMT

It has to do with accuracy and cross-type focusing. In order to make sure you get accurate and reasonably quick AF, Canon limits it to f5.6. Except on the higher end cameras where the extra grunt can be added.

Sigma and Tamron lenses report f5 to the camera so they will work. The actual aperture is recorded when you shoot.

If you want to test this you can tape three of the pins on a tele converter and try it. This will fool the camera and it will think you have what ever lens you have attached. So if it is a f5.6 lens the camera will still think it is a f5.6 lens. The lens will AF albeit slowly and maybe not to accurately even though the converter changed the f5.6 to f8.

Re: AF Point Sensitivity with Narrow Aperture Lenses

Waddizzle — Thu, 01 Dec 2016 16:13:38 GMT

@ebiggs1 wrote:
It has to do with accuracy and cross-type focusing. In order to make sure you get accurate and reasonably quick AF, Canon limits it to f5.6. Except on the higher end cameras where the extra grunt can be added.
Sigma and Tamron lenses report f5 to the camera so they will work. The actual aperture is recorded when you shoot.

If you want to test this you can tape three of the pins on a tele converter and try it. This will fool the camera and it will think you have what ever lens you have attached. So if it is a f5.6 lens the camera will still think it is a f5.6 lens. The lens will AF albeit slowly and maybe not to accurately even though the converter changed the f5.6 to f8.

Thanks, I've been curious as to why I seemingly get softer images with an f/5.6 body, compared to an f/8 body, at the long end of the 150-600. I've seen it enough to know it's not my imagination.