AlexDROP wrote:
A stone wall and painted door test shot ---> Samuli, are you sure that the wall in the middle of the frame is properly in focus? It seems that you nailed focus on doors that are located farhter ahead about 07-1.0 m.
Could you add a crop of the center part of wall (by height) left or right to the door and a door crop just to confirm focus accuracy?
F#ck... of course I focused to doors, they are the main interest point of the image. Like said was not trying to shoot test images. I'll restart the Mac Pro to process one more photo.
Toothwalker wrote:
Well, I am not much of a lens tester. I am more interested in optical phenomena than in lens tests or comparisons, but occasionally I join the test crowds. We'll see.
Today I went out with a bagful of Distagons and tackled a museum wall. The framing is about the same with all lenses, but slightly more tilted as the focal length decreases because the patch of grass goes downhill towards the building. This subject is not well suited to study lateral chromatic aberration because there is not much color variation or high contrasts, but still there are clear differences between these lenses. Focus is on the arch above the door, using 10x live view at the maximum aperture. You best download the images and compare them in your favorite viewer. The image magnification is 0.003, representative of infinity performance. Don't look at the grass in the foreground, which is out of focus.
Samuli, I think that shot should be the model for sh*tty corners!
Toothwalker, I think your ZE 25/2 wall shot at f8 (top corners) shows the same lousy corners I saw with my copy and I have seen in other's shots. Whether that bothers you or not is a personal thing but if you compare a crop of the corner with one from the center you see a world of difference.
Toothwalker wrote:
Today I went out with a bagful of Distagons and tackled a museum wall.
Thanks for test images and the long period of time you spent. Quite educational. But I found some images vary in WB. Is it due to different shooting times during sunset or D21 has strong blue cast by itself compared to pinkish of ZE25/35? How long did you take test shots with the whole set of lenses?
wayne seltzer wrote:
Toothwalker, I think your ZE 25/2 wall shot at f8 (top corners) shows the same lousy corners I saw with my copy and I have seen in other's shots. Whether that bothers you or not is a personal thing but if you compare a crop of the corner with one from the center you see a world of difference.
The corners are not ideal. The design compromise is clearly one of high performance over a large part of the frame at the expense of the corners. The F/2 performance is very good, much better than that of the old D28/2. In fact, the D25/2 outperforms the Y/C D28/2 in every way at F/2 and F/4. At F/8 however the 28/2 has the edge in the corners. I would of course have preferred a lens with a uniformly high performance over the entire frame, but that is not possible for the same price. It doesn't bother me much, since these corners are only a few percent of the total area and relatively unimportant in most compositions. The lateral chromatic aberration of the D28/2, D35/2, and D35/2.8 (not evident in these shots) bothers me much more.
AlexDROP wrote:
Thanks for test images and the long period of time you spent. Quite educational. But I found some images vary in WB. Is it due to different shooting times during sunset or D21 has strong blue cast by itself compared to pinkish of ZE25/35? How long did you take test shots with the whole set of lenses?
The whole set took 15 minutes. We are looking at the shadow side of a building on a blue sky day and the color balance differences between lenses have a much larger effect than changes in ambient light. My D21/2.8 is just cold, something that I already noticed in the film days.
Toothwalker wrote:
The corners are not ideal. The design compromise is clearly one of high performance over a large part of the frame at the expense of the corners. The F/2 performance is very good, much better than that of the old D28/2. In fact, the D25/2 outperforms the Y/C D28/2 in every way at F/2 and F/4. At F/8 however the 28/2 has the edge in the corners. I would of course have preferred a lens with a uniformly high performance over the entire frame, but that is not possible for the same price. It doesn't bother me much, since these corners are only a few percent of the total area and relatively unimportant in most compositions. The lateral chromatic aberration of the D28/2, D35/2, and D35/2.8 (not evident in these shots) bothers me much more.
Ok, I understand you like the 25/2's excellent CA correction.
I forgot to add the corners are worse at wider apertures than the f8 ones.
Also, it will be even more noticeable on higher MP cameras like D800.
In my earlier posted midday landscape comparison shots I liked the warmer ZE 21 shots over the 25/2 which I found to have a more bluish cast on the grey grasses.
Maybe you could shoot a couple of color chart comparison shots with same light?
BTW, thanks for all your effort for the above test shots.
Toothwalker wrote:
My D21/2.8 is just cold, something that I already noticed in the film days.
Actually I hoped the cast was due to different lightning conditions. I'm on market for ZE21 now so I\m suprised how cold D21 is. A new Z version of the lens is as cold as Contax ver, isn't it? If you know...
wayne seltzer wrote:
Toothwalker, I think your ZE 25/2 wall shot at f8 (top corners) shows the same lousy corners I saw with my copy and I have seen in other's shots. Whether that bothers you or not is a personal thing but if you compare a crop of the corner with one from the center you see a world of difference.
To put thinks into perspective; I have seen photos (e.g. 50MP), in which soft corners are clearly seen even on web resolution. 2/25 isn't nearly as bad as 50MP. After two minute PhotoShop work here us combilation of Toothwalker's f/8 image in various print and monitor size.
Based on my printing experience (experienced for hobbyist, but not professional understanding of printing) I would say this lens after print sharpening prints fine for A3 (for work in which you look with loupe or from very close) and A2 if you just watch it with eye (e.g. for gallery).
Personally I'm fine with the sharpness defects this lens has, I rarely print larger than A2 (well I don't remember ever having any single frame image printed larger, some stitched panoramas have been printed much bigger). For printing such large I'm more worried to find something worth printing (A2 photos are rather expensive). In computer I have largest "better than sRGB, well calibrating"-monitor, and it has only 2560x1600 resolution - I would prefer larger but none available at the moment. And I don't do pixel peeping so should be sufficient.
AlexDROP wrote:
Actually I hoped the cast was due to different lightning conditions. I'm on market for ZE21 now so I\m suprised how cold D21 is. A new Z version of the lens is as cold as Contax ver, isn't it? If you know... 21ZE and 100ZE images are colder than rest of the ZE, specially "simple" lenses Planar 50 & 85. You will notice it ONLY if you shoot exactly same subject with exactly same settings. Otherwise you can equalize the coldness with using 50-75K warmer WB in RAW processing. (Disclaimer: based on experience, not measurements)
I just bought a NEC PA271W, and although I am generally happy with the colour/gamut, the operational characteristics are less satisfying (hum all the time although very low, flickering through multiple white balance/gamut settings on startup, fiddly to get it to go to sleep properly, some USB wakeup issues with my mouse, I think) and I miss the sheer size and resolution of my old Cinema HD 30", so I am already slowly beginning to plan my next monitor... The resolution issue also becomes more important now that I am planning to get the D800E.
carstenw wrote:
Samuli, which monitor do you have?
I just bought a NEC PA271W, and although I am generally happy with the colour/gamut, the operational characteristics are less satisfying (hum all the time although very low, flickering through multiple white balance/gamut settings on startup, fiddly to get it to go to sleep properly, some USB wakeup issues with my mouse, I think) and I miss the sheer size and resolution of my old Cinema HD 30", so I am already slowly beginning to plan my next monitor... The resolution issue also becomes more important now that I am planning to get the D800E.
I have had already many years HP LP3065, 2560x1600. Calibrates well, really nice profile with shallow curve changes.
Only issue is the lack of menu. It causes that I need to calibrate again every time I click + or - button accidentally, since there is no way to confirm the settings due to lack of menu. HP explained lack of menu because there was not powerful enough "GPU" to drive 2560x1600 (It was the case with all 30" monitors in beginning, now they have it). Lack of GPU also means that the ONLY resolution monitor accepts is 2560x1600 - no issue with Mac, but Windows installation/safety mode cannot be done with this monitor, had to buy 24" HP for PC to install Windows.
Crosspost also to here - bokeh at medium distance wide open; well could be better...it's worse than C/Y 2.8/28 (only talking about additional contrast in bokeh close to corners and shorted image borders - thou not bad as ZE 2/28 on f/2 or f/2.8). Next time I go to forest, I'll check how f/2.8-3.2 work.
Samuli Vahonen wrote:
To put thinks into perspective; I have seen photos (e.g. 50MP), in which soft corners are clearly seen even on web resolution. 2/25 isn't nearly as bad as 50MP. After two minute PhotoShop work here us combilation of Toothwalker's f/8 image in various print and monitor size.
Based on my printing experience (experienced for hobbyist, but not professional understanding of printing) I would say this lens after print sharpening prints fine for A3 (for work in which you look with loupe or from very close) and A2 if you just watch it with eye (e.g. for gallery).
Personally I'm fine with the sharpness defects this lens has, I rarely print larger than A2 (well I don't remember ever having any single frame image printed larger, some stitched panoramas have been printed much bigger). For printing such large I'm more worried to find something worth printing (A2 photos are rather expensive). In computer I have largest "better than sRGB, well calibrating"-monitor, and it has only 2560x1600 resolution - I would prefer larger but none available at the moment. And I don't do pixel peeping so should be sufficient....Show more →
I disagree with you, as I have done building wall shots awhile ago with 50MP at all apertures when Boris brought up the extreme corner problem on that lens and we both saw it go away when stopped down to f5.6. It was never this bad on a planar wall shot at f8.
