front/back focusing - to whom is it important?

[xpatUSA]

You don't focus on a distance, but on a plane parallel to your sensor. If you recompose, rotating the camera, that plane is moving, too. The distance to your subject might stay the same, but the plane has changed. On short distances that's very noticeable. You will get front focusing

This should only matter if you rotate quite a bit and have a very shallow DOF. The ratio of original to final distance to the point of focus, looking straight ahead without rotation, is just the cosine of the angle of rotation, right? So a 10 degree rotation, for example, would make the distance to the focal point 0.985 times the original distance.

Yummy, Dan, actual numbers to consider!

My DSLR has a 20.7mm x 13.8mm sensor. 24.878mm diagonal (pardon the precision).

Therefore, my 50mm prime would give a half-angle of arctan(12.439mm/50mm) = 13.97 degs, the cosine of which yields a ratio of 0.97 at the very corner of the frame - irrespective of distance to the subject (apart from the amount of lens extension due to focusing). I do like dimensionless numbers

Not saying that George is wrong here, BTW.

A hazy subject for me; aren't lens supposed to counter-act this effect? The term "flat field" springs to mind.

[Mike Buckley]

That's a very silly comment to a very reasonable question.

True. However, the silliness was so easy to overlook because it was so funny.

[george013]

Yummy, Dan, actual numbers to consider!

My DSLR has a 20.7mm x 13.8mm sensor. 24.878mm diagonal (pardon the precision).

Therefore, my 50mm prime would give a half-angle of arctan(12.439mm/50mm) = 13.97 degs, the cosine of which yields a ratio of 0.97 at the very corner of the frame - irrespective of distance to the subject (apart from the amount of lens extension due to focusing). I do like dimensionless numbers

Not saying that George is wrong here, BTW.

A hazy subject for me; aren't lens supposed to counter-act this effect? The term "flat field" springs to mind.

See it a bit different. Focus in the viewfinder on a certain point and than recompose. Lets say you change over half the viewfinder. If you have a lens with a fov of 45 then jou have a correction of 22centigrade. If you have a telelens, then that correction is half of the for of that lens.
Written on a ipad, that is a torture for me.
George

[DanK]

My DSLR has a 20.7mm x 13.8mm sensor. 24.878mm diagonal (pardon the precision).

Therefore, my 50mm prime would give a half-angle of arctan(12.439mm/50mm) = 13.97 degs, the cosine of which yields a ratio of 0.97 at the very corner of the frame - irrespective of distance to the subject (apart from the amount of lens extension due to focusing). I do like dimensionless numbers

Not saying that George is wrong here, BTW.

A hazy subject for me; aren't lens supposed to counter-act this effect? The term "flat field" springs to mind.

No, I don't think so. I think a flat field lens (like many macro lenses) is supposed to correct for this curvature within the FOV. You would still get a slight effect from rotating the camera, which would create a change in distance much larger than that within the FOV (but still very small).

[William W]

^
One point that I take away from Dan's post is that 10 degrees Camera Rotation, for most Shooting Scenarios would be quite a lot.

WW

[george013]

^
One point that I take away from Dan's post is that 10 degrees Camera Rotation, for most Shooting Scenarios would be quite a lot.

WW

It's not only the angel that counts, also the distance. With a telelens you won't see a difference. But a shorter lens the difference will be visible.
Let's say you use 1/3 of the view for recomposing, meaning focus with the center focuspoint on a subject and the recompose so that that subject is to the left.
A 200mm lens has an angle of view of 10 degrees. Calculating with 1/3 correction that will be about 3 degrees. And a long distance.
A 24mm lens has an angle of view of 74 degrees. A same correcting will cover nearly 25 degrees. And a short distance.

George

[William W]

It's not only the angel that counts, also the distance. With a telelens you won't see a difference. But a shorter lens the difference will be visible. . . .

I Agree - but, the bottom line, for most practical purposes: it really about the FRAMING and the APERTURE used - as explained below

*

On the topic of Focus – Lock Focus – Recompose:

Addressing this from a PRACTICAL viewpoint, that is from a viewpoint of "what is relevant to consider when shooting", many Photographers will pivot the camera when executing the Focus and Recompose Technique, but, even though the Plane of Sharp Focus will change when the Camera is pivoted , for mostly all applications of Focus & Recompose, there won't be any problem in the final image.

The other common instance of pivoting the camera is for a Panorama Stitch, and, again there is mostly not an issue, by virtue of the fact the stitch is usually of a 'distant' panorama and using a small aperture.

Certainly I have come across a few examples (of Portraits) when Focus and Recompose might present an issue, these are all very TIGHT SHOTS made at a very LARGE APERTURE, (the FL of the lens doesn't matter, because the DoF basically remains constant dependent upon the FRAMING).

In these rare cases, because the framing is SO TIGHT then the RECOPOSE is a very small movement. In these rare cases, it is easy to adopt a technique of RECOMPOSING by moving PARRALEL to the Subject and NOT pivoting the camera.

This is an example:



EOS 5D + EF 85 F/1.8 @ F/1.8

In this shot The Photographer used the CENTRE POINT AF and the Plane of Sharp Focus was locked on the leading eye (Subject's L. Eye) and then the Frame was Recomposed by moving the camera a small distance parallel to that Plane of Focus and not by twisting or pivoting the camera.

WW

[george013]

I Agree - but, the bottom line, for most practical purposes: it really about the FRAMING and the APERTURE used - as explained below

*

On the topic of Focus – Lock Focus – Recompose:

Addressing this from a PRACTICAL viewpoint, that is from a viewpoint of "what is relevant to consider when shooting", many Photographers will pivot the camera when executing the Focus and Recompose Technique, but, even though the Plane of Sharp Focus will change when the Camera is pivoted , for mostly all applications of Focus & Recompose, there won't be any problem in the final image.

The other common instance of pivoting the camera is for a Panorama Stitch, and, again there is mostly not an issue, by virtue of the fact the stitch is usually of a 'distant' panorama and using a small aperture.

Certainly I have come across a few examples (of Portraits) when Focus and Recompose might present an issue, these are all very TIGHT SHOTS made at a very LARGE APERTURE, (the FL of the lens doesn't matter, because the DoF basically remains constant dependent upon the FRAMING).

In these rare cases, because the framing is SO TIGHT then the RECOPOSE is a very small movement. In these rare cases, it is easy to adopt a technique of RECOMPOSING by moving PARRALEL to the Subject and NOT pivoting the camera.

This is an example:



EOS 5D + EF 85 F/1.8 @ F/1.8

In this shot The Photographer used the CENTRE POINT AF and the Plane of Sharp Focus was locked on the leading eye (Subject's L. Eye) and then the Frame was Recomposed by moving the camera a small distance parallel to that Plane of Focus and not by twisting or pivoting the camera.

WW

That's exactly what I meant.

In that photo the eye has moved about 18% from the centre. Just measured with a measuring tape. A 85mm lens has a angle of view of 25mm. So to my logic that would be some more as 4 degrees if you would shoot that picture with rotating the camera. The photographer didn't trust that I think.

I will look for some practical examples I encounter often.

George

[William W]

That's exactly what I meant. . .

Good, I am glad that I had an example to show your point. An example that when twisting the camera to recompose would probably be very dangerous and stuff up the final image.

***

The photographer didn't trust that [twisting the camera] I think.

No, I didn't.

One of the techniques that I employ in Photography is knowing and applying "safe limits".

I have lots of limits which I have collected and I have tested them all. Also, I am always finding new limits.

Knowing the point of how tight I can Frame a Subject and what Aperture I can use before I CANNOT twist the camera to recompose is just one of those limits which I have collected and tested.

WW