Why is it so difficult to believe that the 25 may be sharper than the 21. It's not like many people own both and have done extensive comparisons. I have no problem believing that, until proof of the contrary
HerbChong wrote:
i knew this would be asked. why try something i am not interested in?... ... i worked out a COC that represents my actual usable diffraction-corrected resolution for objects at infinity and it turns out to more or less matching shooting with the f8 range marks instead of the actual f11 markings. is this rocket science? no.
Of course there are occasions when extreme DOF is necessary. But as I usually find the DOF appropriate for landscapes already at f/5.6, which is about the optimal f/ for the 21mm, and your experience of the lens contradicts the "common knowledge", it feels natural to search for an explanation. Myself, I simply could not resist trying, experimenting. Thatīs how I learn new things. To me itīs equally valuable to understand how others do and think, especially if they donīt do it my way, but do know what they are doing. I may find a better way, or at least add an understanding of the variability of the human mind.
No, HFD is not rocket science, just a bad (usually) assumption and some geometry. You and I know that. But there are many more reading this, perhaps even a HFD-believer that just might see the light.
I think the "common knowledge" is the heritage from Contax days when the 21 was the king and the 25 was below average. The current situation is different but no one seems to bother investigating it. The 25 is still handicaped by the bad reputation of the Contax version, while the new 21 is undoubtedly similar in performance to the Contax with a very incremental improvement, if any at all.
edwardkaraa wrote:
Why is it so difficult to believe that the 25 may be sharper than the 21. It's not like many people own both and have done extensive comparisons. I have no problem believing that, until proof of the contrary
Actually it might be hard to find out as both lenses appear to outresolve the present sensors, except maybe in the +-extreme corners. If they donīt there is something wrong with the specific lens at hand.
Blizzard yesterday, blizzard tomorrow. Everything is either covered with snow or buried in snow. Sixth month with snow, OHH, it hate snow. Otherwise I would get some test shots of actual test targets: The Reality!
i hardly think f11 represents extreme DOF even for a 21 on a full frame body. i find my self wishing to be able to shoot at f22 to capture from infinity to very close even on the 21 but i give up too much sharpness. even when a sharpening tool compensates for diffraction, there is only so much it can do. opening up to f11 and with proper capture sharpening and it's possible to resolve 1 pixel detail over much of the DOF distance range. the 25 is slightly better because i need to sharpen just a fraction less.
Herb...
Z250SA wrote:
Of course there are occasions when extreme DOF is necessary. But as I usually find the DOF appropriate for landscapes already at f/5.6, which is about the optimal f/ for the 21mm, and your experience of the lens contradicts the "common knowledge", it feels natural to search for an explanation.
sharpness actually doesn't get better stopped down. I did them at 4 and 5,6 too, but there is almost no point showing them, since there is only less oof.
The blur quality closeup is actually very, very nice. Also, with that impressionistic look you don't really need Makro Planar levels of sharpness. Real world vs. test charts right there.
"Actually it might be hard to find out as both lenses appear to outresolve the present sensors, except maybe in the +-extreme corners. If they donīt there is something wrong with the specific lens at hand."
Indeed, yet one sees 'differences' in images shot with different lenses, contributed to by other factors. theSuede recently opined that his 7D did not outresolve his lenses - that camera is the FF equivalent of 46Mp. CZ simply say 'we need more pixels'.
We prefer one over another not only because of resolution but the other factors we know about.
Another aspect of Herb's findings is that diffraction on a high Mp FF DSLR is much much better than on a lower Mp FF DSLR, due to the better MTF of the sensor. So final image quality should be better at small apertures *compared* to what we are accustomed to seeing in low Mp cameras.
The extent to which one can use careful sharpening to compensate for diffraction is a very good question. Total near-far sharpness is the realm of TS lenses, of course - for now anyway. A lot of landscape images actually beneift from a gentle fade out of sharpness, as the 21mm does so effectively. A way of training the eye what not to focus on, much like differential sharpening. Agree about 'received wisdom', nothing beats seeing.
edwardkaraa wrote:
Why is it so difficult to believe that the 25 may be sharper than the 21. It's not like many people own both and have done extensive comparisons. I have no problem believing that, until proof of the contrary
I would say the burden of evidence is on the other side. We have found out that Herb cannot help me, as I am not interested in how subjects at infinity look when the lens is focused at HFD for his acceptable CoC at infinity, and he is not interested in making tests for other cases.
edwardkaraa wrote:
I think the "common knowledge" is the heritage from Contax days when the 21 was the king and the 25 was below average. The current situation is different but no one seems to bother investigating it. The 25 is still handicaped by the bad reputation of the Contax version, while the new 21 is undoubtedly similar in performance to the Contax with a very incremental improvement, if any at all.
I think rather than the Contax reputation it is verdicts like atPhotozone where they state "The resolution capabilities are simply sub-standard for such an expensive prime lens". But if you compare their resolution numbers with those of the 21, they are actually very similar (3327 vs 3378 at f2,8 and 3087 vs 3091 at f11). What is different, is the rather bad numbers for borders and extreme borders, which again reach the 21 level at f8 and f11.
People read the verdict, don't look at the numbers and go back to their forum reporting that according to Photozone the 25 is a bad lens not worth its prize (actually half of what the 21 costs).
@bluetsunami: the bad border sharpness works out nicely in this example with its even structured leaves and the smooth wooden table in the background, but it can work out rather bad with gras-like or twiggy structures, which can lead to some disappointment with flower close ups.
One thing photozone fails to mention is that the border and corner resolution is exceptional at infinity even wide open, but of course they wouldn't know that with their mini test chart shot at MFD.
fortunately not at minimum focus distance
but as I said before, the strength of the 25mm is not at close distances up to about 3 m.
I tried shots like flower bush at 1/2m with surroundings, and those shots were disappointing. But I also used a polarizer, which I will not do again, and the object was lavender, which again has that grassy, branchy structure the 25 doesn't render well at close distances, especially with contrasty light.
can you stand more flower close up attempts?
just a quick setup against the light through the living room window, which was a bit much actually. Limited artistic value, but you can see the focus, oof rendering. Once you are not right at minimal distance the focus gets better.
I just did an examination of a series I shot 2 weeks ago. It was not meant for the forum, so I just did it quick and dirty. Handheld and viewfinder focused. So not very accurate, frames are not exactly matching, focus is not dead on correct.
It was also shot against the light, not ideal.
See what you think.
will upload the comparison in a minute just wanted to test with this shot whether I can tick "private" at Zeiss images and still have it show up. I can.
This thread needs a cat! Cat and master did not agree on where the field of sharp focus should be. Cat won... 25ZF + 5D2, snap shot in live view x10, moving target
hej Z, thats not a moving cat!
toss it a couple of treats and get it moving! and with your third hand you fokus while you toss
all viewfinder focused at 1600 ISO, ZF 25 @ f4 and with energy saver lamp which is impossible to whitebalance. slightly oversharpened too, I wouldn't have shown them, but since cats are on
sorry no more flowers at hand, but I will look what I can do
thats from back to front. If you acount for the focus at 5,6 being slightly more to the front, you can see, that using 5,6 for an image that should be sharp from frame to frame might not be such a good idea with the 25. It would be much worse still at close range with different fosusing.
I can also conclude that 2,8 and 4 are really bad at the front corners. 5,6 I need to test again. 8 and 11 are both fine. All apertures are fine at infinity in the far corners. If DOF is enough
will depend on the framing, that is how much to the front do I want things sharp.
This is 100 %. I am not sure it shows at full size. Should be 1078 pixels wide. Its a screencapture out of Capture one. Not the best method, but you can at least see some things. Sorry for the red indicating blown highlights, couldn't switch that off.
sorry, I kept adding stuff in the editing... at least now the files are also showing up full size. and I hope the layout is now alright.
Thanks for the test photos. But what they really show is that it's impossible to have the DOF cover the entire frame at f/2.8. If you take a photo where both center and extreme corners are at infinity, or close to, you would be surprised of how good the corners are even at f/2.8.
