RF Focus/Recompose Technique

[johnwolf]

Advertisement (gone after registration)

I'd appreciate any thoughts on how you M users handle focus/recompose at apertures with very slim DOF.

 

I'm used to a DSLR with a manual focus screen, in which I can focus well on any part of the screen and need not recompose.

 

Thanks.

 

John

[ndjambrose]

The trick is to maintain the focal plane during recomposition, so moving the camera horizontally or vertically but not forward or backwards. In other words, avoid rotating from the hip or shoulders.

Edited December 16, 2010 by ndjambrose

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

Not quite, this holds true only if you move the camera in a parallel line to the plane of focus. In reality you will turn the camera and your head when recomposing; if you draw the effect you will see that the plane of focus turns into a cylinder of focus (constructed out of tangential planes of focus) with the camera in the center as you rotate. That means you need to lean slightly but progressively forward when rotating the camera.

Edited December 16, 2010 by jaapv

[ndjambrose]

...this holds true only if you move the camera in a parallel line to the plane of focus

 

Moving the camera parallel to the plane of focus is what I'm suggesting.

 

You don't have to rotate or move your head. You may do, but physiology also permits keeping it still.

[johnwolf]

Thanks, Jaap and Neil. Now that you point these techniques out they seem pretty obvious. I've been simply rotating my shoulders, which is obviously a problem. Moving along the plane of focus makes more intuitive sense to me, but I'll try both methods. Appreciate your help.

 

John

[jaapv]

...this holds true only if you move the camera in a parallel line to the plane of focus

 

Moving the camera parallel to the plane of focus is what I'm suggesting.

 

You don't have to rotate or move your head. You may do, but physiology also permits keeping it still.

Yes, I quite understood that, Neil. Of course if you can do it this way it is best as you won't have to calculate the distance shift on the fly. But the amount of recomposing you can do this way is rather limited and you may be forced to resort to rotating.

[tgray]

Advertisement (gone after registration)

The trick is to maintain the focal plane during recomposition, so moving the camera horizontally or vertically but not forward or backwards. In other words, avoid rotating from the hip or shoulders.

 

The other point to remember it really doesn't matter that much for many situations. The long minimum focal distance of .7m means even with a fast lens like a 50/1.4, you have more DOF then you would with many lenses on an SLR focused at .4m.

 

Also, if you aren't careful, during your recompose, it's quite easy to change subject-camera distances by 1 cm or so and never realize it, either with subject movement or photographer movement. So sometimes it's better just to focus and recompose without worrying about the focusing error.

 

It certainly can matter where the utmost accuracy in focusing is required, but in many practical situations, I've found it's not a concern.

[Guest mc_k]

...That means you need to lean slightly but progressively forward when rotating the camera.

 

You need to lean backward, not forward ...

 

E.g. you focus on the model's eye, tilt the camera down to the waist, and at this point you need to pull back a little.

[Guest mc_k]

theoretically

[tobey bilek]

An slr is the proper tool to use for this work. Leica is not for everything.

[daveleo]

I thought that the only lenses that had true flat planes of focus where "flat field" lenses designed for copy work. All other lenses do not have flat planes in focus, they actually have a spherical surface in focus (center of the sphere of course is the lens).

 

I am certain that this becomes less important as you focus out far, but if someone could clarify this technicality, I would appreciate it. I think, because of this, it is correct to rotate to recompose, not translate, but I am not confident of this answer.

[giordano]

Dave,

 

You're right that most lenses don't have a perfectly flat field and that (in the absence of floating elements) curvature of field tends to get worse at closer distances.

 

Your suggestion would mean that maximum sharpness would be on the surface of an imaginary sphere of radius equal to the distance focused on. Actual curvature of field on lenses for general photography is much less than this. But it would have the great advantage of making brick walls and the like meaningless as a test of lens performance.:D

[pop]

This has been discussed before. The last time I joined a discussion about that topic I did a very ugly but hopefully useful --> sketch depicting the situation.

I thought that the only lenses that had true flat planes of focus where "flat field" lenses designed for copy work. All other lenses do not have flat planes in focus, they actually have a spherical surface in focus (center of the sphere of course is the lens).

Curved surfaces of focus would be an entirely different kettle of fish. I would rather think that for all practical purposes the plane of focus provided by a photographic lens by Leica is exactly plane.

[daveleo]

well I did some reading (what a concept ! ) . . . my education on flat field lenses was decades out of date.

 

today's compound lenses more or less eliminate the "spherical aberation" and focus on virtually flat planes (not spherical surfaces, as did some older, simpler lenses.)

 

so, all the remarks above about rotating = error are correct.

 

Philipp . . . neat sketch . . . crazy how that other thread got so intense ! !

[pop]

crazy how that other thread got so intense ! !

 

They lost their focus along the way.

[daveleo]

They lost their focus along the way.

 

and thus their comments were filled with aberrations

[jaapv]

This has been discussed before. The last time I joined a discussion about that topic I did a very ugly but hopefully useful --> sketch depicting the situation.

 

 

 

I can do ugly sketches too

If you focus on a plane of focus at distance "a" and rotate, you add distance "b" to the original focussing distance "a1" to reach the original plane of focus on which you presumable want to focus,so you must lean forward tocompensate.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[pop]

It seems we're discussing two different use cases.

 

The one in mind when I did Ugly Sketch One was the following:

(1) Photographer focuses on eye of Utterly Beautiful Model (UBM), obviously in the center of the frame.

(2) Photographer recomposes such that face of UBM now is at the left hand edge of the now Hopelessly Blurred Image (HBI). Recomposition is done by rotating the camera around the center of the lens.

 

In this setting, the UBM is too close to the image plane. The photographer has to lean over backwards to accomplish his desire.

[jaapv]

Of course. It depends what you want to do,as you rotate your plane of focus. We are both right. And there is one spot that doesn't need leaning.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Guest mc_k]

I can do ugly sketches too

If you focus on a plane of focus at distance "a" and rotate, you add distance "b" to the original focussing distance "a1" to reach the original plane of focus on which you presumable want to focus,so you must lean forward tocompensate.

 

No, only Pop is correct; you have drawn the error "b" in the wrong place. You have focused on a point (the intersection of "plane of focus" and line "a"), call it point X. Then you rotated until X is on the right side of the frame. You are now interested in the distance between X and the new plane of focus.

 

You have posted this before and there was a discussion with the same result, btw.