Benefits of large number of autofocus points?

[William W]

. . . But when shooting wide angle 14mm 102 degrees you can go up for 51 degrees. Think of a group of trees in a lane and you want the first tree in focus.

If you do the geometry, I think you'll find that using a very wide angle lens, for objects like trees, less relevant.

I think if you do the geometry, the most relevant is very tight framing of very close Objects or Subjects and when using very large aperture lenses, wide open. . . so for a 14mm lens that would mean a "small" tree (or let's say for an easier comparison a face of a person really really close, maybe about 500~600mm SD).

For a typical small tree (let's say 6ft or 2m high) and at about 6ft or 2000mm and let's say you want it way off to the side of the frame and you focus and recompose by TWISTING the camera - you'll get something like this, which was shot with a 14mm lens on a 5DMkII . . .



I think if you draw the geometry for that shot, and if the shot were to use F/2.8 (which I think is the fastest 14mm lens available) at a SD = 2000mm the DoF is about 5800mm, and well within the "twist" of the plane of sharp focus if you were to use that technique for Recomposing.

The above is all supposition, and I haven't done the Maths/Geometry of these shots - but my supposition is based upon making many shots like this and using Focus and Recompose Techniques for many years.

WW

[george013]

If you focused with the center focus point you twisted the subject up till about 1/4. A 14mm lens has a AOV of 102 degrees so your twist will be something of 25 degrees.
The same drawing but with the relevant distance d.


I didn't manage to calculate that distance for a twist of 25 degrees. Theoretical the size of the subject is not important, it's only the plane of focus. The f-number and distance, the dof, gives that plane an extra dimension.
As said before: recomposing might bring your subject out of focus.

I experienced that myself. Since then I try to avoid that, special with long twists.

There's another reason I don't recompose. Sometime I want to know why my pictures are out of focus. In CaptureNx I can see the focus point that was used. If I recomposed the framing, it would show me a wrong place.

Another problem I experienced with focal points is the difference between cross and normal. Special with short focal length. The focal points far away of the center had trouble focusing on vertical lines. Or was it horizontal lines, I'm not sure at this moment. Nikon service center explained to me that was due to the normal type of focuspoints, the middle where cross type. I only experienced it with the Tamron 15-30 at 15mm.

George

[rpcrowe]

Rather than reinventing the wheel by going into a long discussion regarding auto focus points which would only be viable for Canon users and for 7D Mark ii users specifically. I am going to post this link to Grant Atkinson's series on setting up the 7D Mark ii for action shooting.

http://www.grantatkinson.com/blog/ca...s-for-ai-servo

I do not use manual focus in shooting action. I skipped an entire generation of Canon cameras (EOS Film SLR generation) because I thought that I could handle focus better with a manual focus camera than with auto focus. I was wrong! AND THE CANON A-1 I WAS USING HAD FAR BETTER MANUAL FOCUS CAPABILITY THAN TODAY'S DSLR CAMERAS - AT LEAST CANON DSLR CAMERAS)

IMO, the 7D2 has the very best auto focus system of any camera I have ever used - although I have heard that the Canon 80 also has excellent auto focus...

I will use the upper cluster of focus points for sports because I want heads and faces to be in focus above anything else...



I used the center cluster of points shooting the Red Bull Air Races and the camera/lens did an excellent job of autofocus against a very busy background...

[Stagecoach]

A zero to much: 27mm. Before 13mm and behind 14mm. For a FF.
George

I'm not sure what you are referring to by "A Zero too much:" George?

If it's my DoF figure of 270mm, 3 different DoF calculators gave me that figure for a FF at those parameters. What am I doing wrong?



And the same DoF calculators give me the same DoF as Bill's figure in his example when his parameters are used

But when shooting wide angle 14mm 102 degrees you can go up for 51 degrees. Think of a group of trees in a lane and you want the first tree in focus.

George

Again with real world shooting of wide angles and taking Bill's tree/flagpole in his typical example above, using his parameters there would have been a 194mm change in camera to subject focusing distance if he focused using the centre point and re-composed to put the pole where he has. And at those parameters there's a 5800mm DoF.

[tao2]

I'm not sure what you are referring to by "A Zero too much:" George?

If it's my DoF figure of 270mm, 3 different DoF calculators gave me that figure for a FF at those parameters. What am I doing wrong?



And the same DoF calculators give me the same DoF as Bill's figure in his example when his parameters are used

Nothing wrong with yer calculations, Grahame. Doing it old style(in my head) but writing down the totals, ah,m within a couple of mms. of yer calculations. Ah think George has simply missed out a decimal point (or a comma - as used in mainland Europe) eg. 2.7 or 2,7 CMs. which, of course gives 270 MMs. Ah stand tae be corrected though.


PS

Rather than reinventing the wheel by going into a long discussion regarding auto focus points which would only be viable for Canon users and for 7D Mark ii users specifically

Ahem (cough) Richard...what about Sony users? whose recent cameras have hundreds more AF points than many Canikon cameras combined...

[george013]

I'm not sure what you are referring to by "A Zero too much:" George?

If it's my DoF figure of 270mm, 3 different DoF calculators gave me that figure for a FF at those parameters. What am I doing wrong?



And the same DoF calculators give me the same DoF as Bill's figure in his example when his parameters are used

I must apologize. I can't reproduce it anymore. . Just enough zero's.

George

[Dave Humphries]

I must apologize. I can't reproduce it anymore.

I suspect the confusion might have been that when George wrote in post #19, he was reading or calculating in cm (elsewhere), but thinking it was mm. Very easily done.

e.g. 0.27 m = 27 cm = 270 mm.