I don't see it as a problem either, but it's significant enough that even stopped down to f/8 or f/11 near subjects in the extreme corners go out of focus. Really only a problem if you're really picky and compose with near objects in the extreme corners.
Andi Dietrich wrote:
I dont understand why that is such a big problem to people, did anybody even ever face a situation where field curvature at f 2.8 really matters?
Why is not not a problem?
U take some cityscape with buildings at infinity. Even up to f5.6, buildings in the center are in focus while those at the corners are out of focus.
weekh wrote:
Why is not not a problem?
U take some cityscape with buildings at infinity. Even up to f5.6, buildings in the center are in focus while those at the corners are out of focus.
Very easy. This will only happen when you focus in the center, a mistake that almost every one does. The secret for sharp photos all over the frame is to focus with field curvature in mind. This will practically mean that instead of infinity your actual focusing distance can be at 10-15 meters or less, (even 5m for UWA) and you'll get much sharper extreme corners if you stop down adequately. In order to determine the field curvature of your lens, you have to shoot an infinity scene wide open at infinity, and bracket a few shots focusing closer and closer until your MFD. You will be surprised that some lenses will give you infinity subjects at below 1 meter in the corners Of course results should be examined at 100x on screen.
So the moral of the story: If you want sharp corners, focus for them, if you want sharpness all over the frame use hyperfocal focusing
edwardkaraa wrote:
Very easy. This will only happen when you focus in the center, a mistake that almost every one does. The secret for sharp photos all over the frame is to focus with field curvature in mind. This will practically mean that instead of infinity your actual focusing distance can be at 10-15 meters or less, (even 5m for UWA) and you'll get much sharper extreme corners if you stop down adequately. In order to determine the field curvature of your lens, you have to shoot an infinity scene wide open at infinity, and bracket a few shots focusing closer and closer until your MFD. You will be surprised that some lenses will give you infinity subjects at below 1 meter in the corners Of course results should be examined at 100x on screen.
So the moral of the story: If you want sharp corners, focus for them, if you want sharpness all over the frame use hyperfocal focusing
sorry but I really don't get your solution. Unless u 're talking about a shot whereby the corner includes nearby objects and object in cneter is at infinity. However, if all objects are at infinity, your solution will not work.
I'm focusing with live-view, with objects, both the corner and center, are all at infinity. There's only 1 point for sharp focus. Focus in center, corner is out of focus. Corner focused, center will be out of focus.
Hyperfocus does not work well even with the lens stopped down to f5.6.
You got it the other way around. When the center is in focus at infinity, the corners are focused beyond infinity. Imagine an arc which closest point to your camera is at the center and going further away from the camera the more you move to the image periphery. The distance scale on the lens is effectively measured at the center. In order to get the corners in focus, you have to focus closer.
To give you an example about the field curvature effects, if you take a picture of a person at 2 meters in the center of the image, the curvature will make objects which are further away in the background be in focus the more you move out of the center.
To illustrate my point, I just took this photo with the ZA 16-35 at 16mm f/2.8. If you look carefully, you might discern the W shape of the field curvature instead of the simple arc shape of less corrected lenses. Now if you're shooting a landscape imagine that the infinity mark corresponds to the sharpest point in the center. In order to get the outer points in focus you have to focus closer than infinity. Note that curvature gets at its worst in the extreme corners, something which I couldn't show in the sample photo.
I don't understand - for years Zeiss 21mm has been the best thing around,
"OMG, look at the sharpness, OMG"
Now suddenly, out of the blue, the lines are not straight, the focus plane
is tilted... WTF?
snowboarder wrote:
I don't understand - for years Zeiss 21mm has been the best thing around,
"OMG, look at the sharpness, OMG"
Now suddenly, out of the blue, the lines are not straight, the focus plane
is tilted... WTF?
Not talking about the 21 in particular, because I know it's an outstanding lens, but as you say for years, people used lenses on 6-12mp cameras and wondered about the sharpness. Now with 21-24mp cameras, we will start to hear a different story.
edwardkaraa wrote:
Not talking about the 21 in particular, because I know it's an outstanding lens, but as you say for years, people used lenses on 6-12mp cameras and wondered about the sharpness. Now with 21-24mp cameras, we will start to hear a different story.
Surely, not a "different" story, but perhaps a more detailed (and complete) one. If the CZ 21 was sharper on the 6-12mp camera, it will still be sharper on the 21-24 mp cameras. The relative ordering will not sudden switch. If anything, you might be now seeing the limits of the lens that were not visible on the lower resolution sensors. As for the lines not being straight, people have always commented on that. It's not a lens designed with architectural used in mind and Zeiss has indicated as much.
snowboarder wrote:
.....the lines are not straight, the focus plane is tilted...
I came from the old 4x5 school. Even a Schneider 55mm Super Angulon lenses exhibited a curved field of focus. Schneider states that the Super Angulon moniker is their designation for 'slightly retrofocus' design. I also had a Schneider 47mmXL Super Angulon, and there was indeed a bit of barrel distortion in that lens.
This curved field of focus is simply another byproduct of retrofocus lens design, along with linear (barrel, mustache, whatever) distortion. Unfortunately, the German school of lens design suffers from those two issues more so than other lens smiths.
jjlphoto wrote:
I came from the old 4x5 school. Even a Schneider 55mm Super Angulon lenses exhibited a curved field of focus. Schneider states that the Super Angulon moniker is their designation for 'slightly retrofocus' design. I also had a Schneider 47mmXL Super Angulon, and there was indeed a bit of barrel distortion in that lens.
This curved field of focus is simply another byproduct of retrofocus lens design, along with linear (barrel, mustache, whatever) distortion. Unfortunately, the German school of lens design suffers from those two issues more so than other lens smiths.
I assume you mean to suggest that the "German School of lens design" might choose to emphasize other aspects of performance than field curvature and distortion in design trade-offs. Other schools of design might place greater emphasis on field curvature and distortion in their design decisions. Choose your lens based on the features that are important to you or your application.
Out of curiosity, what are the other schools of lens design you imply exist? Aren't all the possible others, essentially, copies of German design?
It is well known that other companies like Olympus and Nikon have done thier own lens designs. What German design did Olympus copy to make that 18/3.5 OM lens? The same goes for the Nikon 20/4. Take a look at the history of that lens on the NIkon website.
Now, I'm not talking about individual designs, but rather as was referred to, "design schools". Quite clearly, the model and "school" followed by Japanese lens designer/manufacturer was a German one. Is there now a "Japanese School"? That was essentially part of my question. If so, how is it distinct, and to what extent, from the German School they followed for so long, and when would we mark their design independence from their German origins?
Hehehe, Obviously Nikon started as a copy of Leica and Canon as a copy of Contax, if I remember well (not that I was alive that time). What I find really funny is that Japanese would copy a lens design and "improve" it, which probably means for them make it get higher points on the testing charts, and produce flat life-less photos in real life. (I know this is a gross generalization as lotusm50 will soon point out )
I wonder why some lens reviewers criticize Zeiss for insisting on introducing flawed "old-dated" designs that do not produce sharp corners on the testing charts, have some weird effects of LoCA, are not fully corrected for distortion, MF instead of AF... etc. As if Zeiss is not able to compute a design the same way Nikon or Canon does. What these "testers" either intentionally or not, ignore when bashing Zeiss lenses for the above, is that Zeiss is following a century long tradition of providing certain optical qualities to its glass. The other day I was reading a review of the ZF21 vs the 14-24 Nikon, and it was interesting to see that the Nikon generally produces equal or better results when examining the details at 100% but if you look at the photo as a whole, the Zeiss photo is so much more vibrant and appealing.
Japanese companies clearly started by copying succesful designs, but they have(had) famous lens designers that think (thought) completely out of the German design box. In fact, those German manufacturers have surely learned a design or two from the Japanese by now. And what about other companies like Angenieux? Those French lens designers worked independently of the German companies, in the same time frame. I wouldn't be surprised if the Germans learned a few tricks from them.
New designs only come from design schools. Many Japanese designs are proof that other schools of thought now exist beyong the old German design. You are looking for a distinct line, but it has been a long and gradual process. A good example of the disparity today is the Nikon 14-24G.
Lotusm50 wrote:
Now, I'm not talking about individual designs, but rather as was referred to, "design schools". Quite clearly, the model and "school" followed by Japanese lens designer/manufacturer was a German one. Is there now a "Japanese School"? That was essentially part of my question. If so, how is it distinct, and to what extent, from the German School they followed for so long, and when would we mark their design independence from their German origins?
Which review was this?
Diglloyd?
What did you like better about the zeiss? More contrast,color?
edwardkaraa wrote:
Hehehe, Obviously Nikon started as a copy of Leica and Canon as a copy of Contax, if I remember well (not that I was alive that time). What I find really funny is that Japanese would copy a lens design and "improve" it, which probably means for them make it get higher points on the testing charts, and produce flat life-less photos in real life. (I know this is a gross generalization as lotusm50 will soon point out )
I wonder why some lens reviewers criticize Zeiss for insisting on introducing flawed "old-dated" designs that do not produce sharp corners on the testing charts, have some weird effects of LoCA, are not fully corrected for distortion, MF instead of AF... etc. As if Zeiss is not able to compute a design the same way Nikon or Canon does. What these "testers" either intentionally or not, ignore when bashing Zeiss lenses for the above, is that Zeiss is following a century long tradition of providing certain optical qualities to its glass. The other day I was reading a review of the ZF21 vs the 14-24 Nikon, and it was interesting to see that the Nikon generally produces equal or better results when examining the details at 100% but if you look at the photo as a whole, the Zeiss photo is so much more vibrant and appealing....Show more →
Dont' know if this was already discussed, but a definitive point on which the new lenses generally beat their predecessors is the fact that they are 9 blades vs. 5 or 6, and they are usually curved, vs. straight. You can see this in the flare and in the starburst pattern of the sun. The old 21 c/y has six points, and the new zf or ze has almost too many to count (18?). But it's the avoidance of the old pentagon or hexagon flare that I like. I suppose there is some retro flavor to the pentagon flare, but personally, been there, done that.
This has been a very interesting discussion of Schools of Design in the science and
art of lens design.
This same thing (particular design approaches and trade-offs) shows up very
audibly in loudspeaker designs as well, as anyone with a serious interest in
audio systems can attest. The "Boston School," Canadian loudspeakers, British
(think B&W) loudspeakers, Japanese designs.
It's what makes different lens, speakers, yes even trumpets interesting to compare.
Toothwalker wrote:
>> Of course, you are absolutely right, but for a lens this wide, I don't think it would make a difference.
> That depends on the object distance.
Not really. When the lens is focussed closer, it will have less effective depth of field. Either way, using the centre to focus and then swinging the camera around to recompose will give you incorrect focus, even with a lens that has no field curvature at all. The wider the lens, the greater the angle of swing (recomposition), and therefore the greater the error. With wide angles, if you are interested in corner sharpness or off-centre subjects, focus recompose sucks. Smaller apertures hide the error, but it is still an error.
Of course results should be examined at 100x on screen.
)