Focus shift in the 50mm Sonnar ZM at close range

Summary: I have investigated the focus shift that occurs in the Zeiss C Sonnar 1.5/50 ZM lens at close range (one meter). It seems that that wide open the point of best focus is shifted about six centimetres in front of the intended focus plane. The Leica Summilux ASPH 1.4/50 lens does not exhibit similar behaviour. Despite the focus shift I find the Sonnar a useful lens for certain purposes.

Background: I have the 50mm Summilux ASPH but wanted a normal lens with a more classical rendering, particularly with respect to the out of focus areas (not that the Summilux is bad in this respect). The new Sonnar from Zeiss seemed like an exiting product and after having seen some examples of photos I acquired the lens. I must say that I am already fond of the Sonnar. During the first roll and a half I managed one shot that I really like myself . This picture is more of a family snap , but I really like the way the background is rendered.

A pleasing rendering of out of focus areas often goes hand in hand with strong spherical aberration. Indeed, the lovely Rodenstock Imagon soft focus lenses are designed to take advantage of spherical aberration in a controlled way. Here is an example of a portrait taken with the Imagon. It is recommended that the Imagon should always be focussed stopped down since the amount of spherical aberration allowed will cause the focus to shift. Erwin Puts has also documented focus shift in the Noctilux.

I then came across the following note from Zeiss:

"C-Sonnar T* 1,5/50 ZM

Information about special features for dealers and users

The C-SONNAR T* 1.5/50 ZM is a very special lens; based on a classical lens design concept from the 1930's. The additional letter "C" in the name of the lens expresses this designation.

This lens design helps to achieve pictures with a special artistic touch. This lens 'draws' your subject in a fine, flattering manner and is therefore ideally suited for portraiture. It renders a sharpness that is slightly rounded, being less aggressive than in contemporary lens designs, but at the same time not soft in its rendition.

Many famous portraits of glamorous and prominent people during the 1930's used this technique to great effect. These images are characterized by portraying the person in a shining, nearly celestial way. This effect is very well balanced and not exaggerated; therefore many viewers see it in a subconscious way. The trained observer, however, understands the underlining technique and enjoys the results.

This lens design exhibits some additional effects, which should be understood to achieve the maximum benefit from the C-Sonnar T* 1.5/50 ZM:

Because of the above mentioned classical characteristic of the lens the best focus position in the object space can not be kept exactly constant for all f- stop settings. The passionate photographer might notice a slightly closer best focus in his pictures than expected. When stopping down the lens to f/2.8 or smaller this effect is minimized, so the focus position will be as expected. In order to balance the performance at full speed and other f-stop settings the lens is adjusted with above described characteristic.

The special features of the C-SONNAR T* 1.5/50 ZM are best used in emotional, artistic, narrative images, portraits or atmospheric landscapes. For documentation or technical subjects CARL ZEISS recommends to stop down the lens at least to f/5.6 or to use the PLANAR T* 2/50 ZM lens."

All this made me think that it would be useful to investigate the amount of focus shift so that I would be able to understand the properties of the lens better, especially when used as a portrait lens. Not attempting to be scientific I set up a simple test (described in more detail below). I marked a strip of paper in such a way that when photographed at a 45 degree angle the major tick marks of the scale would be projected as one centimetre apart, and also separated one centimetre in depth. With the camera on a tripod I focussed on the zero mark of the scale (I also checked the distance to be one meter with a measuring rod), and took a series of pictures at various apertures.

Discussion: I have collected the results in this folder. It contains composite images of the scales from the Sonnar (800x1875pix, 400kb jpg), the Summilux ASPH (800x1875pix, 400kb jpg ), and a combined image from the two lenses (1600x1875pix, 850kb jpg ). I have also included a number of pictures that shows a couple of bottles situated 25-35 centimetres behind the point I focussed on.

From Dudak's calculator the theoretical depth of field for these lenses wide open at one meter is about three centimetres, and at f5.6 it is about eleven centimetres. From visual inspection I judge the point of best focus for the Sonnar wide open to be six centimetres in front of the intended focus plane. No (or only a very slight) effect is present in the Summilux. Stopping the Sonnar down moves the point of best focus closer to the zero mark but it is not until f4.0 that it is close, and it is not until f5.6 that we see a depth of field that is roughly symmetric around zero.

The Sonnar is no match for the Summilux at the wider apertures, but at f5.6 there is not much between them. (It might be the case that the Summilux exhibits ugly rendering of the near out of focus area - see the -8 at full aperture.) I would also say that the sharpness of the Sonnar at the point of best focus wide open is less than the sharpness of the same lens at f5.6 and at the point where the theoretical depth of field ends, i.e. the Sonnar is not sharp at all wide open at one meter. (On the other hand - as with other soft focus lenses - the depth of field might appear greater.)

Conclusion: The Sonnar is a useful lens when you want a pleasing rendering of the out of focus areas and stopped down it is sharp, but if you want the best performance wide open you should go with the Summilux ASPH. If you want to use the Sonnar as a soft focus portrait lens at close range you must take the focus shift into account. A useful trick could perhaps be to focus on the ear of the subject!

The test setup: Using a graphics program I put major tick marks on a strip of paper 1.4 centimetres apart. I also included equidistant minor tick marks. The strip of paper was attached to a wooden box that was placed on a table at an angle of 45 degrees relative to the film plane. The projected image of the major tick marks would thus be one centimetre apart and they would also differ by one centimetre in distance from the film plane.

The lenses were mounted on a M4 body equipped with a cable release and placed on a tripod. I put the zero mark of the scale in the centre of the frame and focussed as accurately as I could. The scene was lit by household tungsten lights and the exposure time on Fuji Acros ISO 100 film was one second at f5.6. For the alternative exposures I took care to change aperture and shutter speed without disturbing the focus setting.

The film was professionally developed (in XTOL) and scanned by me using a Minolte DSE 5400 v I at full resolution and Vuescan software. I applied a mild amount of grain reduction and capture sharpening using Neat Image before downsizing the images. I have applied levels and curves adjustment in Photoshop. All settings in Vuescan, Neat Image, and the levels and curves adjustments are identical between the exposures. No additional sharpening was done before saving jpeg:s. I judge the posted images to quite accurately reflect what I see under a ten times loupe.

I would appreciate comments on both method and results.

 

Really excellnt methodology and illustative results. Obviously much care was taken. The pictues do the talking in this particular test by limiting the focus to it's minimum distance. The OOF image of the 50/1.4 aspheric wide open helps to make the Sonnar's case for those favoring extra smooth bokeh. It also demonstates that a factory reset of best focus from f/2.8 to f/1.5 will open up another can of worms for those looking for extreme sharpness at a given focus point when used at smaller openings than f/1.5. I'm also curious as to how this lens performs at longer distances, say 1.5 meters and longer to see if the effects of SA are as pronounced. Any idea from your own experiece at such distances with this lens?

 

Very comprehensive and neatly done test and write up! As clean as your images, Svante!

 

Svante,

Very nice images. Thanks for all the work that went into the test. I too have this lens and it appears that the focus shift is a non-issue if you do not use it wide open at the minimum focusing distance. I am using it on a ZI camera.

Les

 

As Alan may be able to tell you, this has been a topic of interest on the Rangefinder Forum, e.g.,

http://www.rangefinderforum.com/forums/showthread.php?t=35561

http://www.rangefinderforum.com/forums/showthread.php?t=36028

& you may want to weigh in there w/your findings.

 

What is not being discussed is the effect of focus shift at MIDDLE distances, where it
makes the lens useless until stopped down to around 2.4

Focus shift is larger at longer distances. At .9 meters the focus shift is little more
than what you would get from your body wavering...which is why I had decent results
on about 50% of my shots at minimum distance.

The real problem was from 2-15 meters where the shift was so dramatic that the lens
is essentially garbage unless stopped down to f2.8 or smaller. Even Zeiss has said the
shift is as much as a full METER at a distance of 5 meters to subject. My shots
confirmed this exactly.


I can't imagine why I would want a $900 F1.5 lens that I can't focus accurately until
2.8.

The threads on RFF are a great example of folks sticking with their team regardless of
how badly they screwed up...lots of justifying why the lens is actually ok even though
you can't focus it accurately. If Nikon had put out a lens like this they would have
been crucified.

 

Brilliant post! I don't have the latest Sonnar, but have a Opton Sonnar 50mm/f2 on a contax iiia and enjoy the same painterly, oof qualities. I will pay more attention to possible focus shift on that lens as well. Thanks!

 

Very nice post, wish all were as well stated and done. I have this lens and like the softness and bokeh it renders, will try to pick up on what you noticed, thank you.

 

>The real problem was from 2-15 meters where the shift was so dramatic that the lens is essentially garbage unless stopped down to f2.8 or smaller. Even Zeiss has said the shift is as much as a full METER at a distance of 5 meters to subject. My shots confirmed this exactly.

Funny how the 50/1.5 Nokton ASPH and the 40/1.4, each at at 1/3 the Sonnar's cost, never have such problem reported.

It also makes you wonder why the original Contax mount version doesn't have this bad rap.

 

>The threads on RFF are a great example of folks sticking with their team regardless of how badly they screwed up.

Are you saying this isn't something exclusive to Leicaphiles?

 

The focus shift on the Sonnar has to do with uncorrected spherical aberration.

http://www.vanwalree.com/optics/spherical.html

Ron

 

By the way, I love the term "special features" Zeiss uses to describe what is essentially
a problem. Ironically these are the same bs artists who touted the LACK of focus
shifts on the ZM lenses when they were introduced.

The Zeiss literature and website made big points around the control of this
problem...something that had never even come up in Leica lit over the years. If for
that reason only they deserve all they get on this issue.

Leica should re-issue the Summar and tout the "special feature" of massive flare!

 

Anyone care post some images shot wide open with an object 5 and 6 meters that are closely positioned with each other in the frame to demonstate the degree of softness at 5 meters and the simultaneous sharpening at 6 meters. If the blur at 5 meters is noticably significant, it does present a problem with large aperture shooting. DOF at 5 meters with a 50/1.5 wide open (assuming a standard circle of confusion of 0.03 mm) is just +/- 0.5 meters.

 

Thanks for the work. I've experienced this also with my LTM lenses when mounted on my R-D1. Makes sense. I always thought it was something with my adaptor or the lens itself, etc... But this sounds like it could also explain the Summar, Summicron, etc...

 

I sent Alan an e-mail and will do the test he outlined and post the pics.

 

"Even Zeiss has said the shift is as much as a full METER at a distance of 5 meters to subject. My shots confirmed this exactly."

Can you post a link at where exactly Zeiss has stated this? I have never seen a focus shift at that range on any of the older sonnars or copies only close up.

 

Here is one of the shots i posted at RFF when this all erupted. It's pretty downsized
for this forums requirements, but I think you can see what I mean. This is just one of
several hundred shots I made with the Sonnar...actually the second Sonnar I tried. I
used them on two different cameras that focus perfectly with all my other lenses,
including the Noctilux.

 

Since the boy's face is somewhat off center, you obviously moved the camera to compose after you focused on his face. Yes?

 

Not correct. This is a crop of a somewhat larger image. As I said above, I have a boat
load of other images, some taken on a tripod, that look just as bad or worse. I
frankly couldn't believe my own results. Tony Rose rushed a second lens to me for
testing, along with a Planar. The second Sonnar was the same on two different
cameras.

The Planar focused perfectly at all apertures, as did my Summilux and Noctilux
lenses....both of which have more field curvature than the Sonnar.

Frankly I'm pretty miffed because I want a modern Sonnar. I love my old ones but
they are only in Contax mount. A modern flare free version would be my dream lens
if it were no sharper than the original whatsoever.

The kind of shift with the new Sonnar is MUCH stronger than I find in my old F1.5.
While the F2 did have more shift, it was mainly visible at smaller apertures, where
depth of field masks the effect.

Here is another shot, taken at full aperture with an ND filter to keep my shutter speed
at 1/1000th in daylight.

 

>Even Zeiss has said the shift is as much as a full METER at a distance of 5 meters to subject.

Even an LTM Jupiter-3 probably focuses more accurately than that, and they were in production as late as the 1980s.

Post-war Oberkochen Sonnars aren't known to be flarey. Do you use a lens hood with it? Also a Bessa R2C is a better shooter than the Contax IIa/IIIa if you want to use the lens on something else.

 

By the way, Dan, how much do you want for the lens?

 

Thank you for all your encouragement, that is what makes posting here worthwhile.

Ron, I think I have seen the page you linked to at some time. It was good of you to post the link since it is a good explanation of what is going on.

Best regards

Svante

 

Dan, no interest in selling that "garbage" lens?

 

Svante
Thank you for a helpful review of the lens. I do not understand the technicalities behind the differences in the design of the rangefinder focussing systems between the Leica M camera and the Zeiss Ikon rangefinder camera. I understand though that there are differences.

Coming from a practical point of view, lens manufacturers could be expected to design a lens that would favour the use of their own camera body rather than a rival manufacturers camera body. Do you think that there is there any possibility that the focus shifting effect of the C Sonnar F1.5 lens is controlled better by the Zeiss Ikon body when compared to a Leica M body?

Query, did you use a Leica camera body in undertaking the test? Do you know of anyone that has tried the same test on a Zeiss Ikon Body just in case there is a difference? I would like to think it shouldn't make a difference but then again, I don't know.
Thanks Again
Noel

 

Noel, this is not an incompatibility issue between Leica M and Zeiss bodies. The focal length of this lens effectively changes as you stop down. The rangefinder cam of an M mount lens can be correct for only one focal length.

One could theoretically make a very complicated lens mount that shifted the rangefinder cam as you stopped down. But it would cost a blooming fortune.

There are large format lenses that also have focus shift when stopping down.

 

I have this lens and was concerned about focus shift from net reading and some focusing difficulties. I spoke to Zeiss and they advised that the lens is averaged to minimize the problem.
Thereafter I conducted several tripod test at all apertures from 1.5 to f/16 using a statue at close range focusing on the eyes. At all apertures I was unable to find any focus shift even though the images were blown up to 100%. Of course, I would not expect a problem by the time I reached f/4 even if focus shift was a real issue.
In practice I take everything at 1.5 - as the bokeh is great and I like a narrow DOF mostly. If I want a slower shutter speed I use an ND filter rather than stopping down to f/2-f/4. Not perfect - but OK.
I think the focus shift is a non-issue in practice and apprehend that the problem is more likely camera movement and/or inaccurate focusing. My difficulties were that my M9 has a built in diopter measurement of -5 but I need +1.5. Buying a +2 diopter solved the issue.
If anyone is concerned Zeiss will optimize this lens (free of charge in warranty period) to ensure accuracy at 1.5. or or presumably other apertures but this will cause more of a problem where other apertures than that to which the lens has been optimized are used.

 

I have this lens and was concerned about focus shift from net reading and some focusing difficulties. I spoke to Zeiss and they advised that the lens is averaged to minimize the problem.
Thereafter I conducted several tripod test at all apertures from 1.5 to f/16 using a statue at close range focusing on the eyes. At all apertures I was unable to find any focus shift even though the images were blown up to 100%. Of course, I would not expect a problem by the time I reached f/4 even if focus shift was a real issue.
In practice I take everything at 1.5 - as the bokeh is great and I like a narrow DOF mostly. If I want a slower shutter speed I use an ND filter rather than stopping down to f/2-f/4. Not perfect - but OK.
I think the focus shift is a non-issue in practice and apprehend that the problem is more likely camera movement and/or inaccurate focusing. My difficulties were that my M9 has a built in diopter measurement of -5 but I need +1.5. Buying a +2 diopter solved the issue.
If anyone is concerned Zeiss will optimize this lens (free of charge in warranty period) to ensure accuracy at 1.5. or or presumably other apertures but this will cause more of a problem where other apertures than that to which the lens has been optimized are used.