Thanks guys. I get the feeling the 50MP gets a bit overshadowed by the 100MP on the forum, but they are both great IMHO and very different focal lengths.
Rodluvan, lovely shot! Nice composition, lines and textures.
joekraft wrote:
Drat. This looks better on my calibrated monitor; there is a better tonal range.
Hmm, but why? Are you using a browser which doesn't support colour profiles? Or maybe you exported to sRGB and this image really needs a larger gamut (I doubt this though, in B&W)? Or maybe the photo lost its colour profile tag somehow, and is interpreted in sRGB although it is not?
Still clinging to the medium format but I have been following the progression of the Zx lens line since their introduction. Financial needs have made the stepdown to 35mm dslr imminent but I'm not ready to buy in just yet. After skulking around here for a long time, I am beginning to get a feel for how some of these lenses render, even as seen in the thread on my laptop, and am constantly being blown away by many, many pictures here from many IMHO top shooters (and I'm still only on page 535).
Getting to the point of this post, I'd like to ask a somewhat theoretical question on something that perplexes me and that I hope a few of the more knowledgeable and experienced Zx users may discuss:
Zeiss' MTF graphs for the 50 Makro Planar show the 40 lp/mm curves clearly higher, straighter with tan = sag out to 15mm than the 50 Planar. However, in tests by Photozone (not my bible, by any means, as their lens evaluations don't seem to account for focus shift, field curvature or the remarkable photos presented by the "flawed" lenses on this forum), the 50P clearly has significantly higher centre resolution than both the MPs, shown below in line widths per picture height, all at f/5.6, yet its centre contrast mtf value for 40 lp/mm is a mere 60%:
50P centre 3960, border 3320, extreme 3357
50MP centre 3631, border 3229, extreme 3191
100MP centre 3700, border 3357, extreme 3373
At the wide end the 21 and the new 25 Distagons centre contrast values at 40lp/mm are also both about 80% but differ significantly both from each other and the 50P in resolution:
21/2.8 centre 3513, border 3147, extreme 3006
25/2.0 centre 3728, border 3313, extreme 2906
I used to think that mtf and resolution were in somewhat directly related, meaning that a lens with >80% mtf @ 40 lp/mm would be very sharp at that image height, and one with 60-70% would still be good but with significantly less fine detail. Now I am seeing that things ain't so simple. The 100MP (consensus status - "top of the bunch") begins its relatively modest mtf curve at 70% gradually receding towards 15mm. The 50MP's mtf is higher (around 80% at 10mm) out to 15mm. The 50P has a more fluctuating curve with greater seperation of T and S and 80% peaks at 10 and 18mm from centre, but it has the lowest centre contrast value and the greatest resolution. Instead the relationship almost seems inverse.Two other lenses I have (had) whose images I really like are the Hasselblad 120 4 Planar and Leica R 60 2.8 Elmarit, both Makros and both having mtf curves with highish centre values but falling off a cliff with image height.
I should just shoot more and think less, you're right, but I'm at a loss how all these variations are explained. My overall surmise seems to be that the mtf curves are a more comprehensive "preview" as to a lens' performance - higher, flatter, straighter, S=T along a greater image height all being better and presenting more of an image "rightness", ie closer to actually being there in the flesh, and that tests of resolution differences at the levels quoted above really mean bugger all. Any light?
Thank you. I hope this is the right place to ask.
The 50P stopped down, the 21, the 50MP and the 100MP are all really sharp lenses. I would not worry about these numbers, they could easily differ on a different test with different copies. All these lenses, perhaps excepting the 21, benefit a bit from stopping down for ultimate sharpness, but as a whole, they are of very high quality and sharpness.
I should just shoot more and think less, you're right, but I'm at a loss how all these variations are explained. My overall surmise seems to be that the mtf curves are a more comprehensive "preview" as to a lens' performance - higher, flatter, straighter, S=T along a greater image height all being better and presenting more of an image "rightness", ie closer to actually being there in the flesh, and that tests of resolution differences at the levels quoted above really mean bugger all. Any light?
Thank you. I hope this is the right place to ask.
A few considerations.
Photozone does account for focus shift and field curvature, but in a way that does not correspond to the real world of photography.
Lens performance depends on image magnification. PZ uses a magnification of 0.030 whereas Zeiss uses 0 (and a few additional values for the Makro Planars).
PZ measures system MTF, not lens MTF.
Most importantly, the focus criterion of Zeiss is to optimize MTF at full aperture, in the image center, and at 20 lp/mm. A lens like the P50/1.4 has some focus shift and field curvature, such that the optimal focus shifts with a decreasing aperture and the lens seemingly performs better off axis than on axis. Zeiss does not compensate for the focus shift with the f/5.6 curves; they measure the MTF at f/5.6 but still use the f/1.4 focus. If they would focus at f/5.6, you would see very different curves with a center value of at least 80% and off-axis values that do not exceed the center value.
Bob -- I'm a big fan of a lot of your work, especially the city shots and architecture. You have more great shots in this last batch, but a few remind me why I got rid of the ZE 21.
I call it the "poorly cooked roast effect": charred on the outside, and raw in the middle.
I bought the lens in the hopes that it would provide an excellent addition to my architecture and twilight cityscape arsenal. But I found the extreme light falloff/vignetting to be too problematic. I had a similar (worse) effect with a Leica 28/2.8 3-cam (first version). To get a proper exposure in the center (below f/5.6), the borders would be terribly under-exoposed. Proper border exposure made for a super-hot center, similar to your man under the scaffolding. A mid-range balance, left both the center and borders in poor condition.
My priorities are even illumination across the frame, whenever it is practical to expect from a lens. I can always add vignetting in PP, but removing it leaves severe artifacts in the corners.
It was unavoidable to notice bad vignetting on smooth subjects when shooting the 21, especially when I could shoot the same subject with the 24 TS-E II, where falloff is negligible. I accept some falloff with the 16-35II because of its versatility in zooming, but it seemed no worse, and maybe better, than the 21. Squeezed between these two Canon lenses, the Zeiss had to go.
Some subjects or image borders look nice with vignetting, drawing attention to center. Unfortunately, for me, that wasn't the way I wanted to use the 21. So now it has a (presumably happy) new home, and I'm without any Zx lenses at the moment.
I still love seeing the good uses and color rendition of the Zx images on this thread! Thanks.
Toothwalker wrote:
Most importantly, the focus criterion of Zeiss is to optimize MTF at full aperture, in the image center, and at 20 lp/mm. A lens like the P50/1.4 has some focus shift and field curvature, such that the optimal focus shifts with a decreasing aperture and the lens seemingly performs better off axis than on axis. Zeiss does not compensate for the focus shift with the f/5.6 curves; they measure the MTF at f/5.6 but still use the f/1.4 focus. If they would focus at f/5.6, you would see very different curves with a center value of at least 80% and off-axis values that do not exceed the center value.
Then the 50P is even better than the charts indicate, if well focused to account for shift etc. Zeiss does highlight its sharpness in their lens description of it...highly specific information you have at hand, TW, and thanks to you and Carsten for your replies.
Rodluvan: Well done ! Nice composition and sharp
rji2goleez: Like image # 2 ... nice colors and sharp and fine details
Lieutenant Z: Nice colors and sharp and fine details
jojomon11: Nice bokeh on the first pic
arionelli wrote:
Then the 50P is even better than the charts indicate, if well focused to account for shift etc. Zeiss does highlight its sharpness in their lens description of it...highly specific information you have at hand, TW, and thanks to you and Carsten for your replies.
The 50P is better than the charts indicate for the center, not necessarily for the borders. The figures at the bottom of page 27 show a similar case.
te4o wrote:
I haven't been around for a long time: stunning imagery on the previous 50 pages (after long hours of review)
Here is the reason why I was missing (it's a bit hidden)
