Focus shift question

[myoder]

Advertisement (gone after registration)

As someone getting ready to return to Leica photography, I want to make sure I understand the focus shift issue. My understanding is that lenses with focus shift issues only shift focus if you change aperture "after" focusing, correct? I'm surprised this affects so many shooters, since when I consider how I shoot, focus is the last thing I typically do. If I'm in a situation where I am active as a photographer, in the majority of situations, my aperture is already selected manually, or locked in an aperture priority mode. Anybody care to illuminate me if I am way off base here? Basically, my point is, depending on your shooting style, focus shift might not be much of an issue in lens choice.

[Advertisement]

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

[Hariseldon]

The camera does not know what aperture the lens is set to. The rangefinder is calibrated to focus correrectly any lens.

 

Most lenses are calibrated to offer correct focus wide open, if lenses subject to focus shift are stopped down a stop or two then there may be focus error , typically focussing behind where you expect. The more you stop down then the greater the depth of field tends to cloak the effect. Not all lenses show this effect, the closer the subject the more likely is the effect to be seen.

 

From my own experience lenses that reportedly exhibit focus shift did not show any such effect in my tests to check the effect, in practice even with an affected lens you may not see it in normal shooting. I wouldn't worry too much and whether you focus then stop down or the other way round will have no effect.

 

You are more likely to have a problem with a poorly adjusted camera or lens or operator error.

[mjh]

My understanding is that lenses with focus shift issues only shift focus if you change aperture "after" focusing, correct?

The rangefinder mechanism used for focusing doesn’t know or care about the aperture. If a lens does exhibit focus shift, it doesn’t matter when you stop down; all that matters is that you do.

[pico]

Has anyone had the opportunity to use a lens that someone claims to have focus issues? I wonder how much of it might be operator error. I don't have any lenses with such problems, but of course I have made focusing errors. I'm sure we all have.

[brianv]

Most fast lenses demonstrate focus shift. Stopping down from F1.5 to F4 on a Summarit will shift the focus about 1" at 36". On a film camera, some might say it is "soft". On a digital camera, with "pixel peeping" you can see the shift of the point of best focus.

 

Aspherical lenses offer the lens designer a chance to reduce focus shift, it is a function of spherical aberration.

[lars_bergquist]

Spherical aberration means that rays passing through peripheral parts of the lens – far from the optical axis – come to a different focus than rays passing closer to the central axis. With large-diameter, i.e. fast lenses where this phenomenon is noticeable, lenses are normally adjusted so that when the lens is wide open, the resulting bundle of rays hits the film/sensor where the bundle is narrowest and therefore most sharp. When the lens is stopped down, the peripheral rays are progressively cut off, and the sharp 'waist' of the bundle moves to the rear. Consequently, the plane of best focus on the subject side moves forward. The lens back-focuses. This happens always, at all times, at all distances, and whenever you stop the lens down.

 

I bought a 25mm Summilux ASPH in 1998. It gave no problems on film, because the depth of the emulsion layer hid most of the focus shift. On the M8, it was quite irritating. I simply could not trust the lens at apertures from 2 to 4. At 5.6, increasing depth of field did hid the fact that the focus had shifted, but when the subject had considerable depth, it was clear that there was disproportonately little depth of field this side of the point where I had focused the rangefinder, and lots of it behind it!

 

I replaced the lens with the current version with the floating element. This has reduced the focus shift to about half, which is completely acceptable, and I love the lens. The picture was made with the Summilux-M 1:1.4/35mm ASPH FLE at about 2.8, and would have been highly improbable with the previous version.

 

The aspherical old man

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!

[TomB_tx]

Advertisement (gone after registration)

Has anyone had the opportunity to use a lens that someone claims to have focus issues? I wonder how much of it might be operator error. I don't have any lenses with such problems, but of course I have made focusing errors. I'm sure we all have.

I have a Voigtlander Heliar 50 f2.0 that had pretty bad focus shift as it is stopped down. I figure it's due to trying to sqeeze good perfformance at f2 out of so simple an optical design.

After testing focus at different stops I see it does focus farther away as it is stopped down. It is possible to compensate for f-stop if you have tested enough to know how much to far focus (by double image in tha RFpatch) at common stops.

I also found my 69 v2 50 Summicron has some shift. I could tell it front focused a bit at f2, then finally checked and found it is in perfect focus at f4. It is slight enough I might not have noticed if I hadn't tested carefully for it. This version Summicron went down to 6 elements, so again this might be part of teh compromise for fewer elements.

[Guest #12]

...When the lens is stopped down, the peripheral rays are progressively cut off, and the sharp 'waist' of the bundle moves to the rear. Consequently, the plane of best focus on the subject side moves forward. The lens back-focuses. This happens always, at all times, at all distances, and whenever you stop the lens down.

...

 

Do you have a source for this? ... I don't think things are this simple.

[jaapv]

Basically Lars is describing the effect of stopping down on spherical aberration, which seems to me to be correct.

[mjh]

I don't think things are this simple.

But it is. See Spherical aberration.

 

This was a hot topic here after the M8 came out; I wrote an article about this issue for LFI back then.

[lars_bergquist]

Do you have a source for this? ... I don't think things are this simple.

 

They are quite complicated in practice, but that is the theory, and the simplest explanation that is true. Erwin Puts discusses this phenomenon in his Leica Compendium. The Wikipedia too has a concise explanation of 'spherical aberration'.

 

The old man again

[Hannes Lummes]

Nokton 50/1.1 is pretty bad at close distances. So bad that I actually never even cared to learn to shoot it well stopped down when close to subjects. I shoot film, and it is very obvivous how the focus shifts about 5 cm at 1meter distance from f/1.1 to f/4.

[Paulus]

Has anyone had the opportunity to use a lens that someone claims to have focus issues? I wonder how much of it might be operator error. I don't have any lenses with such problems, but of course I have made focusing errors. I'm sure we all have.

 

I had a 90mm elmarit-M. At ∞ it din't focus right. I have made my own tests and later Transcontinenta Netherlands tested it together with me. It was a real issue. Lens and Camera ( M8 ) were sent to Leica. They 6-bit the lens and adjusted the camera with all my lenses ( for free!).

The problem stayed with the 90mm. Transcontinenta have traded the lens at my request for a Zeiss lens ( 21mm ) .

[Guest #12]

But it is. See Spherical aberration.

 

This was a hot topic here after the M8 came out; I wrote an article about this issue for LFI back then.

 

The example in the article is a single element, and you can't extrapolate this to all lenses.

 

It's my understanding that, sure, the focus of some lenses wanders progressively backwards as the lens is stopped down. But it's also typical for a lens to wander backwards, and then progressively wander forwards.

 

So what sources did you use for your article?

[mjh]

The example in the article is a single element, and you can't extrapolate this to all lenses.

It can be generalized to all lenses with under-corrected spherical aberration; these are the lenses exhibiting focus shift. This is standard optics stuff, nothing fancy or new. Once you’ve understood what is happening, what additional sources (there were several sources mentioned here) do you need?

[Double Negative]

Check out my review of the Zeiss C Sonnar T* 1,5/50 ZM lens and scroll down to the "Focus Shift?" section and read that for more info - which is illustrated down below in the "Sharpness" section.

 

Of course, this applies to any lens that exhibits focus shift.

[Guest #12]

It can be generalized to all lenses with under-corrected spherical aberration; these are the lenses exhibiting focus shift. This is standard optics stuff, nothing fancy or new. Once you’ve understood what is happening, what additional sources (there were several sources mentioned here) do you need?

 

no it doesn't generalize, I just gave the counter-example.

 

These are really the best sources, the Leica Compendium and wikipedia articles?

[mjh]

no it doesn't generalize, I just gave you the counter-example.

Huh? Haven’t seen it.

 

These are really your sources, the Leica Compendium and some wikipedia articles?

Grab an optics textbook, any optics textbook will do. People have known about and understood this phenomenon (German even has a name for it: “Blendendifferenz”) for more than 100 years; it’s common knowledge.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[jaapv]

no it doesn't generalize, I just gave the counter-example.

 

These are really the best sources, the Leica Compendium and wikipedia articles?

 

Try this book:

 

The art and science of optical design by Robert Rennie Shannon

[Guest #12]

I will try once more...the notion that lenses progressively shift focus backwards with smaller and smaller apertures,

 

or even the

 

 

...lenses with under-corrected spherical aberration; these are the lenses exhibiting focus shift...

 

is not really correct. It really depends on the lens. The usual thing is to trace a ray for each of numerous zones and plot the shift vs. aperture, and then you can basically see the residual aberration at each aperture by inspection.