ferider
Veteran
I am still wondering how much of the problems mentioned in the LL review and observed by some C-Sonnar users is due to sample variation. We have now seen C-Sonnar photos with acceptable close-up performance (from Joe, among others). Strangely the "good" close-up photos have all been made with a Zeiss Ikon body. Yes, focus shift might be there, but it shouldn't be as prominent. Also the Noctilux is known to have focus shift and nobody complains about its wide-open, close-up performance. My classic fast 50 Sonnars (Nikkor and Canon) are softer wide open than at 2.8, but still acceptably sharp wide open, in particular in the image center.
Since the Zeiss 85/2 is not available yet, the C-Sonnar is the ZM lens with most shallow depth of field. All resolution tests with wide angles that I have seen were done at large distance. I keep wondering if there might be a very slight difference in registration distance, film pressure or RF calibration when comparing ZI to Leica bodies.
Has anybody used fast non-ZI lenses, say a 50/1.4 Summilux, 90/2 or 90/2.8 lens wide open and close up on a ZI body ?
Thanks,
Roland.
Since the Zeiss 85/2 is not available yet, the C-Sonnar is the ZM lens with most shallow depth of field. All resolution tests with wide angles that I have seen were done at large distance. I keep wondering if there might be a very slight difference in registration distance, film pressure or RF calibration when comparing ZI to Leica bodies.
Has anybody used fast non-ZI lenses, say a 50/1.4 Summilux, 90/2 or 90/2.8 lens wide open and close up on a ZI body ?
Thanks,
Roland.
Last edited:
Huck Finn
Well-known
ferider said:
Focus shift doesn't mean that you won't get good photos. It means that you'll get "softer" photos - one of the characteristics of the Sonnar wide open & close up. And it means that the point of focus may be at a different point than the photgrapher chose when he focused. You might see a picture you like, but only the person taking the picture knows what he chose as his point of focus & whether the result was different.
ferider
Veteran
I understand, Huck. My point is that even with the focus shift, you might get pictures like the one below (Canon 50/1.5, wide open, no shift precautions, focused on the eyes). Certainly not 100 l/mm, but "acceptably sharp". Good enough for a web post, for instance. There are many similar pictures made with the Noctilux. And what is observed by some C-Sonnar users (not good enough for a web post) seems to be way worse than sharpness lost via shift. In other words, we might be talking about two different issues.
And, I am in complete agreement with you wrt your comments on Zeiss marketing.
Roland.
And, I am in complete agreement with you wrt your comments on Zeiss marketing.
Roland.
Last edited:
fgb2
Established
This is actually an easy experiment to do. So I thought I would prove to myself that my ZM C-Sonnar (which I like) did not have this problem.
I put the lens on an R-D1 on a tripod angled down toward a spread-out newspaper. Camera was 9 inches above the table, focusing distance 0.9 m set on the lens. I used the Leica 1.25x magnifier held up to the viewfinder to position the paper so I knew just where the rangefinder plane of focus was. Then I took a sequence of images at apertures 1.5, 2, 2.8, 4, 5.6 without touching anything but the aperture ring.
Yes, the lens focuses about 2 cm closer to the camera at aperture 1.5 than at 2.8. You can see the answers right on the camera’s LCD. There seems to be little movement in the plane of focus after that, but it’s hard to estimate since so much of the paper is in focus at the smaller apertures.
A Zeiss pre-war Contax mount 50mm Sonnar at 0.9 m had about the same amount of focus shift. Incidentally, the plane of focus for the lens differed from the rangefinder plane by about 3 inches with this adapter, made from a Contax helicoid. I think Roland made a similar observation on one of these adapters.
Canon 50/1.2, at 3.5 feet, about the same focus shift between apertures 1.4 and 2.8. I think I remember this is a double Gauss design and not a Sonnar.
Noctilux 50/1.0 at 1.0 m, similar focus shift from 1.4 to 2.8. Puts confirms this for this lens in his Lens Compendium.
Summilux 75/1.4 at 0.9 m, aperture 1.4 to 2.8, about the same shift as for the other lenses.
I don’t have the lenses you would think would be corrected for this (modern 50 Summilux ASPH, Summicron, ZM Planar, etc) and would urge people to try this with their own lenses. It takes about 3 minutes.
For my own R-D1 + ZM C-Sonnar combination the rangefinder and lens agreed on the same plane of focus at aperture 2.8 or 4. I wouldn’t trust my uncalibrated refurb rangefinder as a gold standard here. But comments on Zeiss’ bulletin seem to indicate Zeiss has calibrated the lens for correct focus at smaller apertures. Puts says the Noctilux cam is calibrated for the wider apertures, which would be my choice. But it’s clear one has to make a choice, because by aperture f/2 the line of newsprint that was in focus at f/1.5 is no longer sharp for me with this lens – which I still like a lot.
I put the lens on an R-D1 on a tripod angled down toward a spread-out newspaper. Camera was 9 inches above the table, focusing distance 0.9 m set on the lens. I used the Leica 1.25x magnifier held up to the viewfinder to position the paper so I knew just where the rangefinder plane of focus was. Then I took a sequence of images at apertures 1.5, 2, 2.8, 4, 5.6 without touching anything but the aperture ring.
Yes, the lens focuses about 2 cm closer to the camera at aperture 1.5 than at 2.8. You can see the answers right on the camera’s LCD. There seems to be little movement in the plane of focus after that, but it’s hard to estimate since so much of the paper is in focus at the smaller apertures.
A Zeiss pre-war Contax mount 50mm Sonnar at 0.9 m had about the same amount of focus shift. Incidentally, the plane of focus for the lens differed from the rangefinder plane by about 3 inches with this adapter, made from a Contax helicoid. I think Roland made a similar observation on one of these adapters.
Canon 50/1.2, at 3.5 feet, about the same focus shift between apertures 1.4 and 2.8. I think I remember this is a double Gauss design and not a Sonnar.
Noctilux 50/1.0 at 1.0 m, similar focus shift from 1.4 to 2.8. Puts confirms this for this lens in his Lens Compendium.
Summilux 75/1.4 at 0.9 m, aperture 1.4 to 2.8, about the same shift as for the other lenses.
I don’t have the lenses you would think would be corrected for this (modern 50 Summilux ASPH, Summicron, ZM Planar, etc) and would urge people to try this with their own lenses. It takes about 3 minutes.
For my own R-D1 + ZM C-Sonnar combination the rangefinder and lens agreed on the same plane of focus at aperture 2.8 or 4. I wouldn’t trust my uncalibrated refurb rangefinder as a gold standard here. But comments on Zeiss’ bulletin seem to indicate Zeiss has calibrated the lens for correct focus at smaller apertures. Puts says the Noctilux cam is calibrated for the wider apertures, which would be my choice. But it’s clear one has to make a choice, because by aperture f/2 the line of newsprint that was in focus at f/1.5 is no longer sharp for me with this lens – which I still like a lot.
Huck Finn
Well-known
ferider said:
Point well taken, Roland. If there are issues with sample variation, I hope they're few & far between.
Huck
Huck Finn
Well-known
fgb2 said:
Thanks for the test. Great job!
It's interesting that you still like the lens. How does this information influence your use of the C-Sonnar?
Interesting stuff, fgb2! In practice, then, it sounds like a good idea to just lean forward a couple cm before tripping the shutter after focusing at very wide apertures. And that by f/2 or f/2.4 this would no longer be necessary. I'll have to do some tests with mine, wide open, as my few (non-test) shots so far have not displayed the problem.
jano
Evil Bokeh
Doug said:
That might explain why my close ups are sharp.. I waver forward and back, and typically snap at the forward part
Trius
Waiting on Maitani
Jano: Combining your post (I waver forward and back, and typically snap at the forward part) with your sig line is totally, freakin' hilarious.
Mazurka
Well-known
Huck Finn said:
Huck, "alleged" does not mean "false" or "fabricated." It's just another word for "not verified." I cannot find this information on zeiss.com, zeiss.de or zeissikon.com and would like to know the actual source, not FF or the forum where it was quoted .
I agree that it doesn't make sense to set the lens to focus correctly at f/2.8 and not a bigger aperture where the focus shift is less easily masked by DOF, hence my dissatisfaction with the so-called explanation. It's sad if Zeiss was being disingenous like Leica with their M8.
And for that matter, "C" still means "compact" on both zeiss.com and zeiss.de
Last edited:
jano
Evil Bokeh
Trius said:
Oh my, that would explain a lot!!
Huck Finn
Well-known
Mazurka said:
Here's the source:
www.rrisonmind.net/entry/The-answer-from-Carl-Zeiss-from-Germany
It appears to be the same kind of prepared PR release from the Marketing Dept. in response to a customer inquiry. Didn't mean to offend. I stressed that it was not alleged because I thought maybe you had not seen my earlier post in which I thought that I had verified the LL information.
The fact that you, I, & mailman can't find it on the Zeiss websites bothers me. If this kind of publicity info is available on request, it shoud be available before the sale - not after the fact when a user is trying to figure out what's wrong with his lens.
I agree that it doesn't make sense to set the lens to focus correctly at f/2.8 and not a bigger aperture where the focus shift is less easily masked by DOF, hence my dissatisfaction with the so-called explanation. It's sad if Zeiss was being disingenous like Leica with their M8.
And for that matter, "C" still means "compact" on both zeiss.com and zeiss.de
I don't shoot with lenses this fast . . . althought I've been tempted.
So, I don't really know first hand about the issues with this kind of focus shift. Puts & others say that the Noctilux is unusable at wide apertures below f/1. "Unusable" can mean a lot of things to different people. As we've seen here, some people have been dissatisfied with the Sonnar at f/1.5 & others say "What's the problem?" However, there seems to be some kind of issue such that when you have this kind of focus shift, you can't get all of the wide apertures in perfect focus. It appears that Zeiss has chosen f/2.8 as their standard setting although their communication with Sabael indicates that they will adjust to the users' request. My understanding of focus shift had been that it's more common that people realize, but that manufacturers set accurate focus for the widest aperture & that depth of field covers the issue as the lens is stopped down. I guess this doesn't work completely in extreme cases.
So, Zeiss strikes me not as disingenous - just not forthright about the issue. Maybe that's what you meant. Are you referring to Leica's after-the-fact fix-it for a problem they should have known about all along? And probably did? The only difference I see is that the Leica issue appears to be a defect or a failure in development. It seems that they rushed the camera to market before they had all the bugs worked out because they had to have a digital product or stop making cameras. The Sonnar issue appears to be a lens characteristic of an old design which people have forgotten about over the past 50 years. If Zeiss wanted to reissue a classical design with both its charm & its warts, they should not have been so coy in "revealing" characteristics which may be problematic for the user. Or maybe it's "reminding." Buyer, beware. Laissez faire.
cheers,
Huck
Tom A
RFF Sponsor
I have been using the 50/1,5 Sonnar C since last summer. I took it as my primary 50 for a 3 week trip to Europe and used it in a variety of situations (from 1/8 sec at 1,5 to fully stopped down). There is a slight shift when used wide-open but no more than I get with my 50/1,4 Asph. It has a better contrast 'taming" quality than the 50/1,4 Asph which is inherently very contrasty. Both of these lenses are most likely the best you can find and the choice comes down to price and how you like your out of focus areas to appear. The C denotation stands for "compact" and I find the Sonnar very comfortable to hold whilst the 50/1,4 Asph is a bit of a handful, though the focussing tab is better than the Sonnars. What is remarkable with the ZI line of lenses is that they render the image the same way, all the way from the 21 to the 50's. Looking at negatives it is only the angle of view that changes, the shadow.highlight rendition is the same on all of them. The Leica lenses differ dramatically from each other, sometimes to the point that you can pick which lens was used from how the negatives look. I only shoot black/white so my opinions are only valid for that.
The next issue of the LHSA Viewfinder (4/2006) will have some stuff shot with the 50/1,5 as well as with the 40/1,4 SC and the 21/4 VC.
The next issue of the LHSA Viewfinder (4/2006) will have some stuff shot with the 50/1,5 as well as with the 40/1,4 SC and the 21/4 VC.
Huck Finn
Well-known
Tom A said:
Gee whiz, Tom, do you have to inject some sanity into this discussion? We were having so much fun going nuts over this tempest in a teapot.
Your thoughts & insights are much appreciated. Even with all of the criticisms that have been posted inthis thread, I've had my sights set on one as soon as my bank account expands. You've just wet my appetite more. When I explain to my wife, I'll blame you.
I read your review last January & enjoyed it very much. I wonder if you have any thoughts at this point on the repeated postings of rangefinder misalignment. That's the one criticism of the camera that seems to have been sustained. I haven't experienced it, but it has gotten a lot of attention here.
By the way, check the link in my post above or go back to my post at the bottom of page 6 for the release from Zeiss entitled: "Information About Special Features for Dealers and Users." They are now telling customers that the "C" refers to Classic - as in classical lens design. Not that it matters all that much.
Cheers,
Huck
Mazurka
Well-known
Huck Finn said:
Thanks Huck. Can't agree more. Unfortunately we stilll don't know from whom or which department of Zeiss the piece of info supposedly came. Perhaps someone who reads Korean could help.
I really hope Zeiss would explain why they've chosen f/2.8 instead of a larger aperture. Remember, the closest focusing distance is already rather long at 90cm (same as the ASPH Nokton), compared with the pre-ASPH Lux at 70cm.
Well they mean the same thing to me and according to onelook.com
Like the Sonnar being set to focus correctly at f/2.8, the M8's ultra-thin hot mirror is a conscious design decision. The consequences for both could not have escaped the designers and manufacturerers.
However bad a solution the IR-block filter may be, at least it's a work-around. But how is the user supposed to deal with the Sonnar's focus shift at close range and wide aperture? Use the much cheaper Nokton or an SLR?
Well, at least the Sonnar costs a lot less than an M8, if nothing else.
Last edited:
Dan States
Established
My Noctilux F1.0 focused perfectly at all apertures and the focus shift was only visible under the gaze of a grain focusing magnifier on the enlarger and only then when the shot was made at around F3.5. It is not visible in slide projection or in actual prints at standard viewing differences. I have scads of prints shot at all apertures and they are sharp as hell, if a little "warm" in color.
Older Sonnar lenses did focus shift a bit, but it was not of practical difference. Mine all focus perfectly at full aperture. The Sonnar F2 suffered more shift than the F1.5 for some reason, which could account for it's mixed reputation.
It's interesting to see that the old Contax cameras had much more generous depth of field scales for their 50's than did Leica...Probably due in part to the effect of the movement of plane of focus as you stop down. I had always wondered why my Sonnars seemed to have greater depth of focus at smaller apertures than my Summitars and Summars and this may be a reason.
The hyperfocal distance on a Contax IIA at F5.6 runs from 15' to infinity. The same aperture on the Elmar shows a depth of field running from about 23' to infinity. Either Leica used a smaller circle of confusion for their calculations or Zeiss figured the nature of focus shift into the design of their entire system.
Older Sonnar lenses did focus shift a bit, but it was not of practical difference. Mine all focus perfectly at full aperture. The Sonnar F2 suffered more shift than the F1.5 for some reason, which could account for it's mixed reputation.
It's interesting to see that the old Contax cameras had much more generous depth of field scales for their 50's than did Leica...Probably due in part to the effect of the movement of plane of focus as you stop down. I had always wondered why my Sonnars seemed to have greater depth of focus at smaller apertures than my Summitars and Summars and this may be a reason.
The hyperfocal distance on a Contax IIA at F5.6 runs from 15' to infinity. The same aperture on the Elmar shows a depth of field running from about 23' to infinity. Either Leica used a smaller circle of confusion for their calculations or Zeiss figured the nature of focus shift into the design of their entire system.
Dan States
Established
By the way, if Zeiss feels that the performance of the Sonnar at full aperture is acceptable they need to restate the MTF data on their website. Clearly they refocused the lens to correct for this shift when the created the MTF chart, because it indicates a lens with actually quite GOOD performance at both 1.5 and 4.
If we are to accept lenses that can not focus accurately at full aperture they should post MTF data that represents the real world performance so users can make educated choices.
Dan
If we are to accept lenses that can not focus accurately at full aperture they should post MTF data that represents the real world performance so users can make educated choices.
Dan
HAnkg
Well-known
For the 85/2 Sonnar ZM which is likely to cost a lot more then the 50/1.5 Zeiss writes that it will have floating elements and incorporate an ultra-precise, nonlinear rangefinder coupling mechanism. The cam can't know what aperture you are using but perhaps it can minimize the effect when focusing at close range.
rxmd
May contain traces of nut
Why not?HAnkg said:
Philipp
ferider
Veteran
I really find this shift issue over-emphasized.
Here is another practical effect to put it in perspective:
Take a 50mm lens. Your focus target is at the 1/3rd intersections (like the eye of the girl in my picture above). You first use the RF patch to focus then move the point of focus off center, maybe by a quarter of the field of view (very typical, IMO). On a film diagonal of 43mm this is roughly 10mm. Let's assume the subject is 1.5m away, then this corresponds to a focus error of about 3cm. More than the focus shift of the lens, certainly. If the lens shifts in the right direction, it might even help
Unless you use a tripod and focus dead-center, focus inaccuracies happen easily (it gets worse if you include back-lash, RF inaccuracies, etc). But who does this except for lens testers ?
Roland.
Here is another practical effect to put it in perspective:
Take a 50mm lens. Your focus target is at the 1/3rd intersections (like the eye of the girl in my picture above). You first use the RF patch to focus then move the point of focus off center, maybe by a quarter of the field of view (very typical, IMO). On a film diagonal of 43mm this is roughly 10mm. Let's assume the subject is 1.5m away, then this corresponds to a focus error of about 3cm. More than the focus shift of the lens, certainly. If the lens shifts in the right direction, it might even help
Unless you use a tripod and focus dead-center, focus inaccuracies happen easily (it gets worse if you include back-lash, RF inaccuracies, etc). But who does this except for lens testers ?
Roland.
Last edited: