Please define focus shift

[Olaf_ZG]

Advertisement (gone after registration)

2 minutes ago, bdolzani said:

Or, 'when the lens is smarter than you'

That’s most probably not that hard, even my phone is smarter then I am…

[Advertisement]

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

[UliWer]

18 minutes ago, Olaf_ZG said:

Very simple: it’s the reason why my image isn’t sharp, it surely couldn’t be me, so it must be focus shift 😎

It's still your fault: you didn't use the aperture with exact focus. 

[Olaf_ZG]

5 minutes ago, UliWer said:

It's still your fault: you didn't use the aperture with exact focus. 

Please speak for yourself, I don’t make such mistakes, it must be focus shift.

(how possible I can explain otherwise to my surroundings I am not able to use my o so precious gear?)

[Dennis]

One member explained it very well here. Graphics are smart

https://www.47-degree.com/focus-shift/what-is-focus-shift

 

[jdlaing]

7 hours ago, bdolzani said:

If I may, to sort out my confusion. People say there's shift 'changing from one aperture to another'. This may be semantic, but what if you go directly to, for instance, 2.8? Will the focus be incorrect? If it is, then I feel you should say that 'at 2.8, or 4, the lens does not focus correctly according to the RF patch.' It shouldn't have anything to do with changing apertures. 

I don't own any summilux's, but I'm curious if I do get one. 

Further then, to solve this issue, I assume you need to learn how to focus at that aperture correctly with that lens, via live view (easiest), and adjust the patch slightly ahead or behind alignment?

thank you-

brian 

 

 

 

Focus shift Is an anomaly in an autofocus lens. The camera closes the aperture to autofocus and then when he camera returns the aperture to what you have it set, the image is out of focus. 
 

The term has been used with rangefinder lenses but that error in the lens is either front or back focus at a given aperture. A rangefinder lens isn’t stopped down by the camera and then reopened. That is done by the operator. If you were to focus at a given aperture and then open up the aperture there is chance of helicoid movement and distance to subject variation. If using straight rangefinder operation stopping down to focus is worthless as you are not looking through the lens. 

[jakontil]

8 hours ago, Al Brown said:

Focus shift happens when an object is in focus at maximum aperture with fast lenses and photographed with the lens stopped further down - the optimum plane of sharpnes/focus "moves", shifts. This explanation is about as simple as it gets.
The RF patch focusing represents the widest open (narrowest) and most exact focus plain. When stopped down, the focus shift moves the point of focus away from the point of focus selected by RF patch. You would need to use live view stopped down to eliminate the problem.
It also happens on TTL focusing with SLRs and DSLRs as we (usually) see and focus through the lens at its widest aperture.

 

Al so that means, when i stop down, at that very moment do i have to re focus of that very same object?

[UliWer]

Advertisement (gone after registration)

6 hours ago, jdlaing said:

Focus shift Is an anomaly in an autofocus lens.

This is not true. Focus shift is one of many optical faults inherent in any lens. It is caused by the fact that light rays which go through a spherical object - a lens - „travel“ for a different distance at the lenses center than at the outer regions of the lens. This fault of spherical aberration can be corrected by lens design, which takes care that all rays from all sections of the lens meet - more or less - at one definite focus point. If you block off the outer regions of the lens by an aperture there are only the rays from the lenses center who will define the focus point. It differs from the focus point which will be achieved if you take the rays from all parts of the lens. Usually all these effect are very small, and you will not notice them on the result.  But lenses which need strong correction of spherical aberration - especiallly highly opened lenses - may cause noticable focus shift. The 1:1/50mm Noctilux or the pre-FLE 35mm Summilux asph. were „famous“ for this. Of cause optical design with lots of effort can correct this fault as well.

Or you use a pinhole camera - it does not have any faults caused by the lenses spherical form. 

P.S.: The link given by Dennis in #30 describes it much more vivid.

Edited September 28, 2022 by UliWer

[pegelli]

6 hours ago, jdlaing said:

Focus shift Is an anomaly in an autofocus lens. The camera closes the aperture to autofocus and then when he camera returns the aperture to what you have it set, the image is out of focus. 
 

The term has been used with rangefinder lenses but that error in the lens is either front or back focus at a given aperture. A rangefinder lens isn’t stopped down by the camera and then reopened. That is done by the operator. If you were to focus at a given aperture and then open up the aperture there is chance of helicoid movement and distance to subject variation. If using straight rangefinder operation stopping down to focus is worthless as you are not looking through the lens. 

I would say it differently, focus shift is front or back focus (vs. the wide open position of a TTL autofocus lens or the split image rangefinder calibrated setting) caused by changing the aperture. So it's an anomaly that is present in many (mostly bright/large aperture) lenses irrespective if it's an autofocus or rangefinder lens.

[pgk]

7 hours ago, jdlaing said:

Focus shift Is an anomaly in an autofocus lens. The camera closes the aperture to autofocus and then when he camera returns the aperture to what you have it set, the image is out of focus. 

In theory (and perhaps some do) the software in camera and lens could adjust for optimal focus even if focus shift is present in an autofocus lens.

However in a manually coupled rangefinder lens there is no viable remedy other than, perhaps, experience.

[pgk]

8 minutes ago, pegelli said:

So it's an anomaly that is present in many (mostly bright/large aperture) lenses irrespective if it's an autofocus or rangefinder lens.

Not so sure on this. Lens designers can minimise some aberrations at the expense of others which can then be corrected electronically, so autofocus lenses may have less constraints and be better corrected in some ways than manual focus, mechanical lenses, which have to rely solely on optical correcction alone.

[evikne]

I made this little comparison today. It's taken with my 50mm Summilux pre-ASPH at MFD and the camera put on a tripod. The focus ring is not touched between the photos, the only difference is the aperture and ISO (to achieve a similar exposure).

As expected, the focal plane moves a bit backwards for each shot. But what I think is the most interesting part, is that the text at the top left (inset and shown at 100 % at the bottom), still increases in sharpness for each shot, because of the increased DOF. The text represents what you would have seen in a real picture. So this means that stopping down is no disadvantage even though this lens has some focus shift.

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!

[lct]

25 minutes ago, evikne said:

As expected, the focal plane moves a bit backwards for each shot.

Too bad i have not my Summilux 50 v2 or v3 with me but i don't recall of significant focus shift issues on them, contrary to my Summilux 35 v2. Do you mean such focus shift issues are being reported by reviewers or other users? Just curious.

[evikne]

45 minutes ago, lct said:

Too bad i have not my Summilux 50 v2 or v3 with me but i don't recall of significant focus shift issues on them, contrary to my Summilux 35 v2. Do you mean such focus shift issues are being reported by reviewers or other users? Just curious.

At least it was expected by me, because I already knew about this, from earlier tests. However, if others have lenses with this behavior, they may not even know about it, because it’s not always very noticeable in real life (and it depends on how you shoot).

Edited September 28, 2022 by evikne

[weatherproof]

8 hours ago, jakontil said:

Al so that means, when i stop down, at that very moment do i have to re focus of that very same object?

Not Al, but it's extremely difficult to refocus with a lense that exhibits noticeable focus shift on a traditional rangefinder camera. The viewfinder patch is calibrated to what it assumes the lens has in focus. Sure, if you know the lense backfocuses at 1.4, you might be able to give it a small twist to offset the backfocus, but it would a guess (that could get better with experience).

This is not an issue with cameras that actually look through the lense (live view and SLRs) because your eyes can see whether the image is in focus or not.

[lct]

33 minutes ago, evikne said:

At least it was expected by me, because I already knew about this, from earlier tests. However, if others have lenses with this behavior, they may not even know about it, because it’s not always very noticeable in real life (and it depends on how you shoot).

May i ask if you did your earlier tests in real life situations? Just curious again sorry but I'm wary of rulers for lens tests although i find them useful for some comparisons. 

[pegelli]

3 hours ago, pgk said:

Not so sure on this. Lens designers can minimise some aberrations at the expense of others which can then be corrected electronically, so autofocus lenses may have less constraints and be better corrected in some ways than manual focus, mechanical lenses, which have to rely solely on optical correcction alone.

I'm not sure either, maybe I should have said "some" instead of "many". Is your statement that autofocus lenses might be "more corrected" for for focus shift at the expense of correction of other abberations a theory or is there a source with real world data that supports it?

[lct]

4 minutes ago, pegelli said:

I'm not sure either, maybe I should have said "some" instead of "many". Is your statement that autofocus lenses might be "more corrected" for for focus shift at the expense of correction of other abberations a theory or is there a source with real world data that supports it?

Good question. I would have thought the opposite but i may be totally wrong.

[wizard]

vor 17 Stunden schrieb Al Brown:

It depends where the lens is calibrated ONLY for RF lenses with coupling. Not for any of the other lenses.

Me thinks it also applies to every other lens, if and when that lens is focussed at full aperture and the image is then taken using a stopped down aperture. Using SLR cameras or cameras with an EVF, the problem remains, but could be solved by rechecking focus at the chosen stopped down aperture, which is obviously not possible with rangefinder cameras, except when using liveview.

[Advertisement]

Hallo,

Look here: Leica M

[evikne]

5 minutes ago, lct said:

May i ask if you did your earlier tests in real life situations? Just curious again sorry but I'm wary of rulers for lens tests although i find them useful for some comparisons. 

Tests in real life situations, isn't that a contradiction? In real life I haven't noticed any focus shift, except if I shoot at f/2.8 at very close range. The other tests have been made on a tripod, like this one, but at a more natural shooting distance, about 2 meters. 

One surprise with this new test, however, was that the previous test revealed that the lens was calibrated to f/2, not f/1.4. I may have focused a little inaccurately this time due to the wrong glasses, but the principle of focus shift is the same anyway. I didn't mention it because it has no relevance to anyone but myself.

[UliWer]

37 minutes ago, lct said:

May i ask if you did your earlier tests in real life situations? Just curious again sorry but I'm wary of rulers for lens tests although i find them useful for some comparisons. 

If a photo is not sharp, it must be focus shift. If a ruler proves that a lens has focus shift, it must be the ruler.