focus shift - please explain it

[Sparrow]

Originally Posted by ampguy

Anyone who has used a stop down (metering, DOF preview, etc.) SLR, and observed critical focus, will have seen it.

So those who haven't seen it, have either a) bad eyesight, b) wide open SLR or RF use only, or c) have seen it, but haven't been bothered by slightly out of focus subjects.

I just tried that, and I don't agree .. one focuses at full aperture, then when stopped down the DOF covers any error ... oh, and I don't have bad eyesight, a gear fixation or any prints I'm ashamed of ...

[ampguy]

Not questioning if DOF covers it for your photographic needs, but are you able to see the shift of exact focus from wide open to several stops down, with your particular setup?

Not everyone wants to take photos like yours, some want to know their exact single point of focus, and base their front and rear DOF based on that exact point, which changes with aperture, and which can be seen with say a Nikon F3 with center line image focusing screen, and bright enough light to stop down a fast lens and still focus.

Originally Posted by Sparrow

I just tried that, and I don't agree .. one focuses at full aperture, then when stopped down the DOF covers any error ... oh, and I don't have bad eyesight, a gear fixation or any prints I'm ashamed of ...

[JHutchins]

Two lenses I use frequently, the 75 Summilux and the 50/1.5 Jena Sonnar are known for focus shift and, in fact, display it. I notice it principally when at the near end of the focus range. There, the solution is easy -- just focus, as you would normally, then lean in a tiny bit before shooting.

[rxmd]

Originally Posted by ampguy

Not questioning if DOF covers it for your photographic needs, but are you able to see the shift of exact focus from wide open to several stops down, with your particular setup?

Not everyone wants to take photos like yours, some want to know their exact single point of focus, and base their front and rear DOF based on that exact point, which changes with aperture, and which can be seen with say a Nikon F3 with center line image focusing screen, and bright enough light to stop down a fast lens and still focus.

As soon as you stop down "several stops", the exact point of focus becomes pretty much irrelevant and, in fact, hard to see in a print, let alone on a focusing screen. Moreover since whatever divergence is there is eaten by DOF at that point.

It's when you stop down less than "several stops" (one or two at most) that you may be able to discern it, or, in fact, probably not. In the end it doesn't really matter in practice with SLRs.

[Sparrow]

Originally Posted by ampguy

Not questioning if DOF covers it for your photographic needs, but are you able to see the shift of exact focus from wide open to several stops down, with your particular setup?

Not everyone wants to take photos like yours, some want to know their exact single point of focus, and base their front and rear DOF based on that exact point, which changes with aperture, and which can be seen with say a Nikon F3 with center line image focusing screen, and bright enough light to stop down a fast lens and still focus.

... well I tried a 50 105 and 135 on a F3 with the H type screen but failed to replicate your observation, the small variation is covered by the DOF and at full aperture where DOF is at its minimum there is obviously no change at all with a SLR

[bananasplit]

Focus shift you said?

A few weeks ago I sent my M9-P to Leica because all my pictures taken at full aperture with 50/2, 90/2 APO, 75/1.4... all Leica lenses were out of focus. This was at close distance as well as medium or long distance.

Leica told me that I was having a focus shift issue????

What the ... are we talking about? Same lenses with film Leica bodies no issue then a $8000 camera supposedly top of the art and we have a focus shift.

I was told by Leica that this was absolutely normal. You see, the digital sensor is absolutely flat while the film in a M7 or any M film body is not. Well I am learning something.

Wait, I was advised to stop down to get a sharp picture, to play with the camera in order to find the right point of focus (may be they should remove the rangefinder and cut the cost). Yeah right. Get an M9-P and a Noctilux for something near $20.000,00 and you have to stop down to have a focused picture. I never though of that. So stupid I was using a Noctilux at full aperture. Who am I to think I can get a sharp picture. I just have to stop down, right?

Then why should I buy a Noctilux and not a Summicron or even a Voigtlander 50mm. You get it I love to spend money ...

I really apologize if my words may hurt some of you but this is driving me crazy and please correct me if I am wrong. I may have missed something.

Right now I am thinking of going back to film, sell the M9-P or why not buy a Nikon D 800, shame on me!

[Richard G]

Bananasplit that does sound like rubbish from Leica. Which Leica division was it? I use my C Sonnar 50 without trouble. Can you post some test shots? I've regularly used my 135 at max aperture admittedly only f4 and here is a version 5 Summicron at f2 min distance focussed on the Peroni tab on the glass.


L1000812.jpg by Richard GM2, on Flickr

[ampguy]

Setup a photo where you're photographing the top of a fence from min distance out to 50 yards +.

Put a subject at about 15 yards next to the fence and focus on the eyes.

Take 2 photos at the same exposure values. First focused wide open, Second focused stopped down.

Compare the front parts of the prints (fence area near lens) and see if you can spot a difference.

Originally Posted by rxmd

As soon as you stop down "several stops", the exact point of focus becomes pretty much irrelevant and, in fact, hard to see in a print, let alone on a focusing screen. Moreover since whatever divergence is there is eaten by DOF at that point.

It's when you stop down less than "several stops" (one or two at most) that you may be able to discern it, or, in fact, probably not. In the end it doesn't really matter in practice with SLRs.

[john_s]

Originally Posted by bananasplit

.............Get an M9-P and a Noctilux for something near $20.000,00 and you have to stop down to have a focused picture. I never though of that. So stupid I was using a Noctilux at full aperture. Who am I to think I can get a sharp picture. I just have to stop down, right?............

The f1 Noctilux I had focused amazingly perfectly at full aperture (Techpan fine film, tripod etc) and your Leica lenses should focus perfectly at full aperture also. In the case of the f1 Noctilux, even with film, focus shift occurred with smaller apertures, to a degree that increasing depth of field did not compensate. Since I bought it for full aperture use this was acceptable.

The worst focus shift I experienced was with the Zeiss 50mm/f1.5 which was not optimized for full aperture because the focus shift in this lens is huge. Optimizing it for full aperture would have made sense to me, but I disowned it.

As far as I know, Leica lenses are supposed to work optimally at full aperture. As you said, why else would you bother with them?

[BobYIL]

Originally Posted by bananasplit

A few weeks ago I sent my M9-P to Leica because all my pictures taken at full aperture with 50/2, 90/2 APO, 75/1.4... all Leica lenses were out of focus. This was at close distance as well as medium or long distance.

Leica told me that I was having a focus shift issue????



I was told by Leica that this was absolutely normal. You see, the digital sensor is absolutely flat while the film in a M7 or any M film body is not. Well I am learning something.

Wait, I was advised to stop down to get a sharp picture, to play with the camera in order to find the right point of focus (may be they should remove the rangefinder and cut the cost).

May I inquire where this "Leica" is located? With your wording such an absurd commentary could be expected from a plumber but definitely not from any Leica service.

A rangefinder mechanism needing a readjustment is one thing, to blame such lenses like the 50/2, 90/2 APO and 75/1.4 for focus shift or being out of adjustment is totally another thing! And this "absolutely normal"??? "the digital sensor is absolutely flat while the film in a M7 or any M film body is not." "I was advised to stop down to get a sharp picture, to play with the camera in order to find the right point of focus"

The camera in concern is an >$7K one.. the lenses in concern are all top $$$$ ones.. and these were absolutely normal?? And they did not "correct" them but told you to "stop down". Holy Grail!!

I'd be reporting these immediately to Solms.. This is the most ridiculous commentary to be heard from an authorized service, be it Leica or any other brand.... or it was a bad joke...

[ColSebastianMoran]

Here's a test sample that I found interesting. Watch how the plane of best focus shifts away from the camera as you go from f/1.4 to f/2.8 on a Zeiss 50 f/1.4 ZF. (Mouse over the f-numbers at "Bokeh Fringing" near the bottom of this Photozone.de page to see the effect.)

Also note the different false coloration of the OOF foreground vs. the OOF background.

In this example and all others I can find, the increased DOF makes up for the shift, so the original focus point is still sharp. But... 1) The optimal focus plane does shift back, and 2) I can imagine a lens for which the increase in DOF does not cover the focus shift. That would be a bad design.

[JHutchins]

Originally Posted by ColSebastianMoran

Here's a test sample that I found interesting. Watch how the plane of best focus shifts away from the camera as you go from f/1.4 to f/2.8 on a Zeiss 50 f/1.4 ZF. (Mouse over the f-numbers at "Bokeh Fringing" near the bottom of this Photozone.de page to see the effect.)

Also note the different false coloration of the OOF foreground vs. the OOF background.

In this example and all others I can find, the increased DOF makes up for the shift, so the original focus point is still sharp. But... 1) The optimal focus plane does shift back, and 2) I can imagine a lens for which the increase in DOF does not cover the focus shift. That would be a bad design.

Bad design? It'd be a design. The focus shift would be a disadvantage (one that could be mitigated by using a camera with a live viewfinder) but it would bring with it its own advantages. Every lens design is a tradeoff and looking at any lens over one dimension (like susceptibility to focus shift/correction for spherical aberration) is a mistake becuase poor performance on any given dimension is usually a result of a tradeoff.

The Planar is actually very well corrected for spherical aberration and thus not particularly prone to focus shift. In the Makro-Planar it essentially doesn't exist (unsurprising since it's designed to be used at the near end of its focus range). Sonnar designs, by contrast, are highly prone to focus shift, but contrastier & more flare resistant (even without coatings) and smaller and, besides, have a really lovely look.

[thirtyfivefifty]

Originally Posted by f16sunshine

Focus shift is caused by the internet.

Yep. With that said, time to move on now. Nothing else to see here.

[rxmd]

Originally Posted by PKR

When using an 8X10 for table top work, compositions were always made with the lens wide open. Once a basic placement of the subject was finished - the lens (210mm f 5.6 WA Schneider) was stopped down (22-45) and a "hot light" was used to refocus the lens, at a small aperture, for accurate FP and DoF (the camera Standards were readjusted for best DoF -Scheimpflug).

In most instances the "hot light" was a 500W Quartz. If the subject matter was sensitive to heat from the "hot light" (food photography) a series of very expensive 8x10 Polaroids would be used to confirm critical focus and DoF.

Focus shift is a real problem.

https://www.schneideroptics.com/ecom...y.aspx?CID=169

The point of that exercise was to verify DOF and whether you had enough of it, which is critically important. Focus shift comes a distinct second in relevance, being pretty much unmeasurable at f/45, table top or not.

The Schneider page and brochure you linked to don't even mention focus shift.

[rxmd]

Originally Posted by ampguy

Setup a photo where you're photographing the top of a fence from min distance out to 50 yards +.

Put a subject at about 15 yards next to the fence and focus on the eyes.

Take 2 photos at the same exposure values. First focused wide open, Second focused stopped down.

Compare the front parts of the prints (fence area near lens) and see if you can spot a difference.

What lens do you propose using?
At which apertures?
What behaviour do you expect except "there is more in focus"?

[Scheelings]

Originally Posted by Freakscene

The difficulty with focus shift in a lens like the 50/1.5 Sonnar is that if the lens is adjusted for accurate focus at f1.5, as you stop down the depth-of-field increases, but it shifts backwards sufficiently that the point you focus on is in front of the plane of acceptable sharpness - and therefore unsharp. If you focus on an eye you end up with a sharp ear. The increase in depth-of-field in the 50/1.5 Sonnar does not overcome the focus shift until ~f4.5 as this test shows quite clearly: http://photo.net/leica-rangefinders-forum/00aKSW

Focus shift is not imaginary, or an invention, but it's also not fatal. Understanding it can facilitate you being able to better use of lenses that display it, and make photos that look like you want them to, which is the point after all.

Marty

Ok, but if I focus on something at 1.5, take a shot, then step down to say 2.8, does refocussing correct it - or is this lens always out of focus - even if it appears in focus in the viewfinder - from F2~4?

[JHutchins]

Originally Posted by Scheelings

Ok, but if I focus on something at 1.5, take a shot, then step down to say 2.8, does refocussing correct it - or is this lens always out of focus - even if it appears in focus in the viewfinder - from F2~4?

What camera are you using? If you're using something that allows you to look through the lens at the selected aperture (e.g. an electronic viewfinder interchangeable lens camera, or an SLR with the depth of field preview engaged) then something that appears to have the center of focus on a particular point at f2.8 actually has the center of focus on that point at f2.8 (barring user error).

If you're using a rangefinder or an SLR that holds the aperture open while you focus and stops down only for the exposure, then something that appears to be have the center of focus on a particular point at f2.8 may actually have teh center of focus on a different point.

[Scheelings]

Originally Posted by JHutchins

What camera are you using? If you're using something that allows you to look through the lens at the selected aperture (e.g. an electronic viewfinder interchangeable lens camera, or an SLR with the depth of field preview engaged) then something that appears to have the center of focus on a particular point at f2.8 actually has the center of focus on that point at f2.8 (barring user error).

If you're using a rangefinder or an SLR that holds the aperture open while you focus and stops down only for the exposure, then something that appears to be have the center of focus on a particular point at f2.8 may actually have teh center of focus on a different point.

I'm using a Leica M8. So Actually I was weighing up whether to get the 1.5 Sonnar or the 2.0 Planar. Half a stop doesn't sound like much to give up if more of my shots are going to be as tack sharp as I would expect from Zeiss....

Maybe I need to research some planar vs sonnar threads.

Is there much difference with Bokeh?

[JHutchins]

Originally Posted by Scheelings

I'm using a Leica M8. So Actually I was weighing up whether to get the 1.5 Sonnar or the 2.0 Planar. Half a stop doesn't sound like much to give up if more of my shots are going to be as tack sharp as I would expect from Zeiss....

Maybe I need to research some planar vs sonnar threads.

Is there much difference with Bokeh?

So with a Sonnar on an M8 the center of sharp focus can and will be different from what your rangefinder tells you it is. This can be dealt with -- it's just a matter of having an idea of how your lens behaves at various apertures and distances.

But these are very different lenses and if you're considering the Sonnar just for the extra half stop then probably you'd be happier with the Planar. It is a sharper lens, it vignettes less, it is more neutral. The Sonnar imparts a color of its own, yes, the bokeh is significantly different from the Planar, also contrast, field illumination &c. Apologies for the obvious advice, but if you're on the fence you really should do a tag search on flickr for both lenses because the best reason to get a Sonnar is that you like the Sonnar look and to understand it you should see it.

[135format]

What JHutchins said and to add that the Zeiss "C" designation lenses are a throw back to old and loved lens designs with warts and all. They are meant to be like that for good reason. If you want clinical looking images then the Planar is probably for you.