Other sections are excellent and especially the contrast with direct and transversal light, so also commented that video application, in this work these details are valued more as there is no issue of flares and contrasts.
Nikkor 24/1,4 G ED
If a user wants a clearly AF lens 24 / 1,4 G in mont Nikon, this Nikon is the best. I understand that that poses a Zeiss 25 / 2 values in a special way of construction, mechanical focus, contrast in advserse conditions (sun) and no electronic/mechanical problems in the future (mont. Nikon esp.)
Regards wayne seltzer wrote: and differences in bokeh, [
greenfield wrote:
... then building these lenses with a MFD of 0.25m is even more silly, is it?
Looking forward to your thorough and profound tests at any non-silly distances.
The testing distance is not silly, but shooting a plane target at this distance is silly, unless the lens is meant for reproduction work of paintings and flat art work. In real life, you would be shooting people and 3D objects at this distance and the field curvature will have an incredible effect of bringing into focus areas that would be ridiculously impossible to get in sharp focus with a corrected lens. A lot of the 3D effects are caused by this phenomenon. This is one of the strengths of the Zeiss lens design that not many people seem to understand when they look at online tests.
The corner crops from Thorsten on page 8 show lateral CA in the magintude of 1 pixel wide. (estimated from a 400% view). This is about where I would expect the 2.8/21 to be too.
If we compare the samples we got here with photozone measurements, we get that PZ numbers are 1/3 of what we observe in the corners, for both lenses.
Interesting is the the MTF for the old 4/24mm lens from the 50's. They mention that lens had too small of an image circle and thus fell off too much in the corners and the tangential lines fell off too much due to CA. Well, with new 25/2 they fixed the CA and pushed further out in the frame the start point for the decline to the corners but don't mention why they didn't just increase the image circle so that the full frame would be covered before falling too much.
philip_pj wrote:
'Testing these lenses at under 1.25 meters is just silly.'
Please tell me they don't do that, and you made a mistake...that's ridiculous.
Per waynes pos. 50 x fl. I guess there are folks around here who like giving their portraits huge noses.
I have been advised that my 25 is waiting for me to return. I intend to conduct comparisons between it, the 21 and the 35 f:1.4 in terms of colours, detail, contrast, spatial definition and other trivial matters that only affect 100% pf each of my pictures.
From the context of that statement, it looks like Zeiss was comparing the 2/25 to the Contax-mount 2.8/21. I mention this because I have a suspicion the old 2.8/21 is better than the current version in terms of lateral chromatic aberration. If that's true, the 2/25 might be worse than the old one but still better than the new one (for lateral chromatic aberration).
wayne seltzer wrote:
Well, with new 25/2 they fixed the CA and pushed further out in the frame the start point for the decline to the corners but don't mention why they didn't just increase the image circle so that the full frame would be covered before falling too much.
The coverage is fine, since it fills the full-frame sensor without inappropriate vignetting. Are you arguing against the vignetting or against the extreme-corner softness at large apertures?
The latter, to the extent that it exists (I'm looking forward to more samples from G. Pierre and others) is a classic engineering trade-off. You can't get perfect corners at f/2 for free! You'd need to make the lens bigger, more expensive, softer in the centre, etc. I generally feel Zeiss chooses excellent compromises in their lenses, which is why I like using them.
It could be that the ZE has given up the exceptional CA correction of the old C/Y version. The measurements at PZ, ~0.8 pixels, is good for a wide angle, but not exceptionally good.
wayne seltzer wrote:
Interesting is the the MTF for the old 4/24mm lens from the 50's. They mention that lens had too small of an image circle and thus fell off too much in the corners and the tangential lines fell off too much due to CA. Well, with new 25/2 they fixed the CA and pushed further out in the frame the start point for the decline to the corners but don't mention why they didn't just increase the image circle so that the full frame would be covered before falling too much.
Wayne -- This is what I inelegantly tried to say in my earlier post -- the image circle is simply too small, leading to those bad corners. I know it's no small thing to build a lens with a large image circle, but if Canon can do it to the extreme with the TS, surely Zeiss could figure out how to add a tiny bit more diameter to give sharpness across the frame. The 25 looks good, right up to the border and corner, then all the center sharpness makes it stand out even more for those wishing to photograph with details on the edges.
If I read those correctly, the c/y and ZE are both 0-.3 pixel in center and c/y is .5 at the borders and ZE is .7.
Pretty close, can you even see that difference in a blind comparison?
And the CA you noticed on the woman on the beach is in the extreme corner not the border which I doubt is where the lens testers like photozone are measuring CA. Right?
I guess I would put extreme corner CA performance even behind extreme corner sharpness.
Yes, it would be interesting to find out from Dr. Nasse what were the exact design tradeoffs in why they did not design the 25/2 to have a slightly bigger image circle to avoid the extreme corner problem I see at far distances. I don't expect T/S size image circles but I wonder how much bigger they would have had to increased the lens to fix this? Is it that they would have to have a larger front element like the 21 and if they did would they not have reduced a lot of the vignetting wide open too? I guess I am worried that lens designers will "cut corners" in the future to keep lens sizes on the small side due to the recent invasion of mirror-less cameras and lenses which have all the people here jettisoning their slr's like crazy.
wayne seltzer wrote:
Sure, but how can I send you the large RAW files?
The church shots were all done on a tripod with MLU and using loupe and liveview whereas the shots of the pier were handheld just so you know. Also, the church shots had the sun pouring into the frame from the left side of the church so that might of increased the amount of CA on the left side vs. the right side.
If there is an illumination difference between left and right it doesn't show in the overview shot of the church.
Maybe you can use something like dropbox.com for the raw files? Otherwise it would be useful just to crop the angels' wings and post them here.
I bet it will be just as popular as Bokeh thread here 
, but the sad thing is that sooner or later I'm sure we will really get such a thread.