j.liam wrote:
The transition from focus to blur is quite striking, I will admit. I don't think I have seen a 3-D rendering quite like it from a Zeiss lens before. Tuth be told, the CY35/1.4 was not a lens I paid much heed to primarily because as a Nikon shooter, it can't be converted with a Leitax mount.

Do you all simply avoid shooting planar subjects except at f-stops smaller than 5.6 or so? Have you been stymied by the apparent lack of sharpness at even f/2 or does the manipulation of contrast PP easily offset it?

Well, yes. I wouldn't shoot planar subjects at f/2 with my 35/2 ZE either. I pretty much have two modes - wide open or stopped down (f/5.6-f/8, f/16 if I absolutely must).

The old 35/1.4 is solid stopped down, and actually a bit sharper in the center than the 35/2 ZE/ZF but has weaker edges/corners. It does however have a somewhat different color rendering - much warmer than the ZE.

I've never used it much stopped down (it's at f/1.4 where it is truly special), but here are two shots, probably shot around f/5.6:

The background really does melt away into a gentle, pleasing, soft colored blur. It will be interesting to see whether it performs similarly in field shooting. Does anyone know whether this new version and the old CY have the same or similar optical formulae?

No, the new one has a completely different design. It came as sort of a surprise as the old one was a relatively heavy but compact lens. I suppose they wanted to get a combination of interesting rendering wide open and high performance stopped down. The old design isn't quite up to modern ZE/ZF standards stopped down.

The sheer size of the new lens was supposed to reflect the added glass needed for a highly corrected optic so the weak performance thus far on the testing is a bit startling. With such a seemingly poor performance from f1.4-2, and about equivalency to the Z*35/2 @ f/2.8 and smaller, it will have to prove itself quite remarkable in other ways to warrant the elevated price and heft. I am a bit puzzled, surprised but also disappointed at first glance.

j.liam wrote:
The sheer size of the new lens was supposed to reflect the added glass needed for a highly corrected optic so the weak performance thus far on the testing is a bit startling. With such a seemingly poor performance from f1.4-2, and about equivalency to the Z*35/2 @ f/2.8 and smaller, it will have to prove itself quite remarkable in other ways to warrant the elevated price and heft. I am a bit puzzled, surprised but also disappointed at first glance.

i would never expect an f/1.4 lens to outperform a premium f/2 lens at f/5.6 or smaller. it should be easier and cheaper from a design perspective to get the same quality (and better contrast) stopped down from a slower lens. the main reason f/1.4 lenses often outperform slower lenses of the same focal length is that the slower lenses were designed to be budget lenses. this is clearly not the case for the zeiss 35/2. i'm sure if zeiss were to design an f/1.4 and f/2 lens for the same price point and focal length the f/2 lens would always be better. the big price difference between the two lenses demonstrates how difficult it is to get an f/1.4 lens to perform as well as an f/2 lens.

incidentally, how does the new nikkor compare to zeiss 35/2 stopped down?

It would stand to reason that the faster lens would show its mettle at the wider end because that's where the 'money is'. Or rather went toward the design. I would fully expect it to be weaker stopped down, of course. This does not appear to be the case here so some other characteristic is the design focus to the detriment of other lens metrics. Zeiss PR focuses in its press leases on the 'extraordinary bokeh'. Is this enough to justify the expense? Or is it really meant for videographers rather than still shooters. I don't shoot video so can't answer that question.

I refer again to the Lloyd Chambers review and comparison against the Z*35; although he had issues with the AF, in the final analysis he did mention that the Nikkor may be the finest 35 he'd shot with up to then.

j.liam wrote:
I refer again to the Lloyd Chambers review and comparison against the Z*35; although he had issues with the AF, he did mention that the Nikkor may be the finest 35 he'd shot with up to then.

i don't have a subscription, so the referral doesn't mean much to me. also, from the other quotes i've seen lloyd seems to have said that about a lot of 35s over the years. from the images i've seen the nikkor seems optimized for sharpness and lack of SA wide open. i'm guessing it competes with the Z* 35/2 stopped down for sharpness but has lower contrast?

They run head to head in his review even stopped down and he mentions that there's no clear winner to be had.

j.liam wrote:
They run head to head in his review even stopped down and he mentions that there's no clear winner to be had.

sounds like a winner, i think i'd rather have the c/y 35/1.4 and the zf 35/2 for the same price though (or more likely the c/y 35/1.4 and $800). it all depends on what you want the lens to do.

As a Canon 35L user, I read the comparison by LC with great interest. It is difficult to discuss his findings here, as only those with a subscribtion are able to see what I talk about, but this cannont stop us from discussing it either.

The micro contrast at 1.4 as seen in LC's center crops is strikingly higher with the 35L than with the ZE. We are not talking about subtle differences here. This relationship is the opposite of what the MTF charts would suggest.

Canon at 10 and 30 lp/mm wide open (black):

The local/micro contrast is primarily the 10 lp/mm line and the Canon chart shows a higher performance. However as you pointed out, the Canon charts are theoretical while the Zeiss MTFs are measured. In my experience the Canon provided MTF curves have only a vague relation to reality.

The 40 lp/mm is relevant for the fine detail in the image and shots of a test chart isn't very helpful in that respect. You have to look at real world images that actually contain a lot of fine detail.

denoir wrote:
The local/micro contrast is primarily the 10 lp/mm line and the Canon chart shows a higher performance.

I read at 10 lp/mm: Canon 1.4 ~0,75, and ZE 1.4 ~0.85 in the centre. In what way do you get the Canon to show a higher performance? The blue lines are at f/8.

From my limited knowledge of SA where the observed side effect of focus shift is evident, I would not be surprised if it affects contrast at very low frequencies, even down to just a few lp/mm.

alundeb wrote:
In what way do you get the Canon to show a higher performance?

By comparing the Canon at f/8 to the Zeiss at f/1.4 (I forgot that the blue lines are f/8)

denoir wrote:
...
The 40 lp/mm is relevant for the fine detail in the image and shots of a test chart isn't very helpful in that respect. You have to look at real world images that actually contain a lot of fine detail.

However Zeiss should also use test charts to produce those MTF and not nature as source AFAIK.

By the ways Nikon theorical MTF estimate 35 AFS : 65% contrast center @1.4 @ 30LP while Zeiss measured 42%, center, f/1.4, 40LP

j.liam wrote:
The sheer size of the new lens was supposed to reflect the added glass needed for a highly corrected optic

Really? I haven't come across this claim before. Do you have a reference?

My personal theory is that in ye olden days Zeiss had much tighter control over the manufacturing process, but today they are made by Cosina, which is not in the same caliber as Zeiss, or Leica for that matter. Thus I expect that the redesign was primarily to yield a design which is more robust in the face of lower tolerances, and that Zeiss strove to maintain the look of the old lens, with better coatings, not to make any particular "improvements", which could just as easily have destroyed what makes this lens so loved.

trusty wrote:
However Zeiss should also use test charts to produce those MTF and not nature as source AFAIK.



Zeiss uses a K8 MTF machine to produce the MTF charts. It uses a scanning slit system (i.e laser interferometry) and does not rely on test charts.

Zeiss doesn't use simple test charts for measuring MTF. They make a machine that measures MTF and this is what they use and the MTF reported by Zeiss (unless they say otherwise are measured at infinity).

carstenw wrote:
My personal theory is that in ye olden days Zeiss had much tighter control over the manufacturing process, but today they are made by Cosina, which is not in the same caliber as Zeiss, or Leica for that matter. Thus I expect that the redesign was primarily to yield a design which is more robust in the face of lower tolerances, and that Zeiss strove to maintain the look of the old lens, with better coatings, not to make any particular "improvements", which could just as easily have destroyed what makes this lens so loved.

I'm not so convinced by that. Zero of my ten Zeiss primes or any of my Canon cameras and lenses have been in for repairs while my M9 and my 75 Summicron have been back to Solms twice now. I have not heard of a Leica owner who doesn't send back the gear for repairs on a regular basis. Leica has the worst quality control I've ever seen in camera equipment. The basic construction and material of the lenses are first class but the assembly is not.

My guess is that Zeiss was as bad when the lenses were hand built. The modern Cosina production lines are far more automated which probably leads to better error margins and overall more consistent and accurate assembly.

Leica uses tighter tolerances in their designs than most manufacturers these days, which is part of the difficulty. They use both machines and manual labour for assembly, and make many of the machines themselves, as do/did Zeiss. I have read somewhere that their tolerances are up to 10 times tighter than usual for lens production, but I don't have a source for that. Leica, possibly more than any other lens manufacturer today, still values compact size, and much of their tolerances are taken up with this, but the levels of CA compared to the main competitors are also often much lower. The largest part of the high prices on a Leica lens goes towards supporting this labour-intensive process, not towards extra quality control, and yes, Leica lenses do seem to need adjustment (too?) frequently. However, Leica really stands behind their product, so apart from time constraints, no one should worry about the performance of their lenses. Leica will set it right again, if it needs it.

I think one must differentiate between tolerances in optical performance and tolerance in mechanical performance. Optically, yes, I agree with you but that has more to do with the use of expensive exotic glass than with the actual lens assembly. The manual assembly process on the other hand seems to have a very high mechanical failure tolerance given the high rate at which lenses are sent back for repairs.

I remember reading somewhere that the mechanical quality of the assembly of Leica glass dropped significantly after the 1970's. This would not surprise me. It looks to me like they are trying for high volume manual assembly now, and that it isn't working too well neither in terms of volume nor quality.