Thanks to Mike for continuing to share your tests. I have compared an Oly 21/3.5 with the CZ 21/2.8 (but not at really close focus like this test). The Zeiss 21 is better overall, but the little Oly 21/3.5 is really close when stopped down. The Oly requires a little more post-processing (CA correction and sharpening), but when looking at prints most folks would be hard pressed to tell the difference between the two lenses.
I agree with Mark and Tim. It seems that folks who own Zeiss get locked into its superiority because of a kind of lemming mentality... Yes it is the better RESOLVING lens in the extreme corner, but I think the more relevant fact is the slight differences that shows up here at actual pixel view will not show up in a print at 200+ PPI resolution. And no matter what, the differences while perhaps arguable as to their significance, are only visible in the extreme corner.
However, when you factor in the added distortion of the CZ at those same extreme edges the "superiority" of the Zeiss in OVERALL imaging quality gets significantly diminished -- bear in mind distortion, unlike resolution, WILL remain visible in a print! IMHO it really boils down to which factor you place more significance on. You of course are entitled to your own opinions...
Any way you cut it, the Oly 21's -- BOTH of them -- are attractive options to the CZ when you compare size, weight and OVERALL imaging performance even EXCLUSIVE of the obvious cost advantage to the Oly.
Edited by Jack Flesher on Aug 07, 2005 at 06:27 AM GMT
Tim Ernst wrote:
After 30 years in the business I have not sold a single 100% crop on screen though. In a print you simply cannot tell the difference - folks around here sometimes seem to forget all about the real world - we sell prints and images in books and calendars and magazines and not pixels.
this is very true. having just recently printed a bunch of 8x10s for my portfolio, my zeiss 21 shots don't appear to be any more sharp than my 24 TS-E. this is the largest i've gone with these shots so far, but i'm not expecting the results to differ much in larger formats, especially since the 240lpi of most offset presses (namely a heidelburg) will kill that sub-pixel micro-resolution anyway.
either way, it's interesting that a $350 zuiko 21 can be so close to a $1600 zeiss 21. i'd like to see some tests against CA and flare to help make a better decision on which is the overall best lens since sharpness is only one factor.
It seems that folks who own Zeiss get locked into its superiority because of a kind of lemming mentality...
You know, it would be nice if folks here could try to disagree without being disagreeable.
What I see in this thread is a test clearly showing the Zeiss to be the better lens, and then a bunch of Zuiko fans all claiming it doesn't matter. I don't think that's a reasonable basis for calling those who DO see the Zeiss's superiority "lemmings".
Yes it is the better RESOLVING lens in the extreme corner, but I think the more relevant fact is the slight differences that shows up here at actual pixel view will not show up in a print at 200+ PPI resolution.
And again, the same thing could be said of the Canon lenses. If your standard is "good enough" that's one thing. People shell out for exotic lenses because they want the best.
However, when you factor in the added distortion of the CZ at those same extreme edges, the "superiority" of the Zeiss in OVERALL imaging quality gets significantly diminished.
Sorry, can't agree with that at all. Distortion is not that hard to correct, but blur is impossible to correct.
charlesk wrote:
Sorry, can't agree with that at all. Distortion is not that hard to correct, but blur is impossible to correct.
Agreed. Vignetting and barrel distortion are easily corrected in Photoshop. Chromatic abberation is trickier. Flare and blur are bad news. The absence of flare and blur is where the Zeiss excels.
And again, the same thing could be said of the Canon lenses. If your standard is "good enough" that's one thing. People shell out for exotic lenses because they want the best.
We're not suggesting you can extrapolate this across the entire image surface -- we're talking (arguably) minor resolution differences in the extreme corner performance and the tradeoffs there, not tradeoffs across the entire image. Here the Canon does have trouble.
Another argument is how much you can enlarge before this overall lower resolution becomes significant to your images. FTR, I have printed 1Ds images taken with the 16-35 to 24"x36" and they look great at that print size, though you can see the "problems" in the corners at actual pixel view. So your claim may in fact be correct that it doesn't matter
Sorry, can't agree with that at all. Distortion is not that hard to correct, but blur is impossible to correct.
Uhhh okay... So tell us exactly what happens to the edges of your image when you do the distortion correction? Oh yeah, they get distorted so they are no longer square and the entire image needs to be cropped back rectangular before displaying! In the CZ you lose enough of the original image to make it what, about an effective 23mm lens?
Falloff OTOH is easy to correct without losing any image.
Edited by Jack Flesher on Aug 07, 2005 at 06:47 AM GMT
it's all so darn subjective. we shoot with what works for us. for me, i shot with and owned two copies of zeiss 21/2.8 and 18/4. there was simply no benefit for me, landscape shooting, over my canon 16-35L. and i print large. i never pixel peep on screen, that's not real world imo - i look at the printed image. in fact, i shared some 20x30 prints, same scene, exposure, etc, from yosemite with several pros here in ny, and all three picked the canon 16-35L over the zeiss 21.
there are a lot of tradeoffs, for some, there are some benefits. and i don't begrudge anyone who wants to use this glass! but i've seen very few real world examples of the zeiss in action - everyone talks about it but they don't share the results. guy's shown the airplane cockpit (impressive shot!) but other than that i've only seen "tests"
i do agree that there is a unique economy going for the zeiss glass - fed by the fuel from folks here and on other fora ... many new shooters see this hype, and want to try it for themselves.
for me, i'll take the af and metering of my 16-35L over the zeiss 21 - it's not worth the tradeoff imo. but that's just my opinion.
mike, thanks for sharing your results, it should make for an interesting discussion. btw - the results of the zuiko look impressive, you probably have just caused a 50% jump in the going price for these now
Jack Flesher wrote:
Uhhh okay... So tell us exactly what happens to the edges of your image when you do the distortion correction? Oh yeah, they get distorted so they are no longer square and the entire image needs to be cropped back rectangular before displaying! In the CZ you lose enough of the original image to make it what, about an effective 23mm lens?
ok, so either take this into account when composing the original shot or just crop the shot and be happy. the resulting focal length, be it 21mm or 23mm, will only matter to the photog and not to the end viewer of the photo. i have yet to hear someone ask me what was in the other 1/4" to the left or right of my photos.
charlesk wrote:
I think the zeiss is sharpest in the corner but not by much.
It doesn't look very close to me at all. Look at the text in the poster on the Zeiss shot and the Zuiko 3.5 shot.
Look at the color, contrast and definition of the black line on the left edge of that poster in the Zeiss shot and Zuiko shot. The Zeiss line is black -- the Zuiko line is basically the same color as the guy's pants!
And this is at f/8? I can only imagine what the differences would be at f/3.5 or f/4.
Disclaimer: I own Zeiss glass and not Zuiko glass. But the differences seem rather obvious to me. The Zuiko lenses may be a better value than the Zeiss lens but they are not in the same league in terms of performance. --c...Show more →
Charles - you are right that the Zeiss 21 is sharper than the Oly 21/3.5 at a close-focus range. But check out the Zeiss 21 vs the Oly 21/2 - do you still think it's not close?
Also - in a longer-focus range shot, the Oly 21/3.5 performs better, although it's still a tad softer in the corners than the Zeiss.
ok, so either take this into account when composing the original shot or just crop the shot and be happy. the resulting focal length, be it 21mm or 23mm, will only matter to the photog and not to the end viewer of the photo. i have yet to hear someone ask me what was in the other 1/4" to the left or right of my photos.
Except my point was if you do that, you are "losing" the very extreme corners which is where you're basing your CZ defense from.
Just to be clear, nobody is claiming the CZ21 is other than an excellent lens -- in fact I can say it is the sharpest 21mm lens out to the extreme corners I have seen. However I will also state that IMO the resolution increment above the Oly in the corners is not going to be significant in most prints, at least for me.
Also, the CZ's waveform or moustache distortion may not be problematic in most images but certainly can be in architectural renderings. Here one needs to take the crop factor of the distortion correction into account and know that they may end up with a slightly tighter FOV than 21. And if you crop it to say 22.5mm and do the same with the Oly, then the advantage of the CZ's corner resolution has been nullified, blurring the distinctions between the two lenses further.
Lastly, for folks looking to obtain a good 21 a few other factors should be considered, namely availability... The lack of availability of the CZ has driven its price to the point of non-affordability for many shooters even on the rare occasion when they do crop up for sale.
All I'm suggesting is that the Oly offers a very attractive wideangle alternative for many photographers.
So your claim may in fact be correct that it doesn't matter
But obviously to the people who are interested in exotics, it DOES matter.
Giving up "auto everything" matters a great deal to a lot of people. If you are going to give that up, it has to be to get substantially better results.
I won't argue that the Oly may be better for the money. But Mike's own test results put to rest any claim that the Oly is close to the Zeiss irrespective of price. IMO, of course.
And once again.. it's f/8. I'd love to see this comparison between the Zeiss and the Oly 2 redone at 2.8.
Uhhh okay... So tell us exactly what happens to the edges of your image when you do the distortion correction? Oh yeah, they get distorted so they are no longer square and the entire image needs to be cropped back rectangular before displaying! In the CZ you lose enough of the original image to make it what, about an effective 23mm lens?
I don't really see that much distortion in the Zeiss shot, to be honest. And at least all of it is very usable, which cannot be said for the Oly if you care about corner sharpness.
Falloff OTOH is easy to correct without losing any image.
I'm not talking about falloff, I'm talking about the fact that the far corner in the Oly shots are not sharp. You would have to crop those and then you are back to 23mm anyway. --c
However I will also state that IMO the resolution increment above the Oly in the corners is not going to be significant in most prints, at least for me.
To each his own.
I would still love to see these compared wide open.
Lastly, for folks looking to obtain a good 21 a few other factors should be considered, namely availability... The lack of availability of the CZ has driven its price to the point of non-affordability for many shooters even on the rare occasion when they do crop up for sale.
I think this is the only valid argument for the Oly over the Zeiss.
As for the "rare occasion" that they come up for sale, that's in the past. There are plenty of them available for sale, and the price has come down substantially, as you know. --c
Charles - you are right that the Zeiss 21 is sharper than the Oly 21/3.5 at a close-focus range. But check out the Zeiss 21 vs the Oly 21/2 - do you still think it's not close?
It's closer in terms of sharpness, and the Oly 21/2 may well be a better choice for the money. At least at f/8.
But how about chromatic abberation and contrast? Look at the black line on the left edge of the poster -- it's lavender in the Oly shots. --c
What we really see here is the truly abysmal nature of Canon's wide L zooms. I should be inured to the sensation by now, but every time I see comparative results I'm amazed afresh.
How many of us have now travelled the same path from Canon's glass through to Zeiss and Leica and now Zuiko are becoming the optics du jour? Everyone should of course do their own testing, and long may results like this be publicised - even if only to persuade Canon to pull their finger out.
A big heap of consensus is mounting:
• There is nothing close to the Zeiss 21, including the Olympus - as this test demonstates - c'mon, the Zuiko 21 is clearly getting way less information to the sensor.
• The next best thing might be the Olympus, or it might be the Leica 19 or it might be the Sigma 20, or it might even be the Leica 21-35. As soon as my Olympus adaptor arrives, I'll post some comparative samples of the next best . . . . Certainly my Sigma 20 is very close to the performance of the Leica 21-35 . . .
• Canon's best efforts fall woefully short in this arena. What sense is there is developing higher and higher density sensors unless they are also developing a range of much better lenses for them?
As for the "rare occasion" that they come up for sale, that's in the past. There are plenty of them available for sale, and the price has come down substantially, as you know. --c
Charles: Dohhh... I just realized you are the same Charles selling your CZ21 over at RG... Sorry for not making that connection earlier. I don't post here very often any more and this was in a gear review thread so was just sharing my opinions on lens performance...
FWIW, yours he cheapest one I've seen lately at $2999 and looks to be a very good buy for somebody wanting a CZ21.
As for wide open performance, I did actually test that a while back, but did not save the images to post. In a nutshell, the CZ was a bit better overall at f2.8 but both were fairly poor compared to stopped down. From f4 up both are very usable and the differences were again in the extreme corners, about the same and what Mike shows in his f8 comparisons here. Of course the Oly kicked the CZ's butt at f2
All joking aside, the Oly 21/2 at f2 generates a unique image reminiscent of the Leica Noctilux M (50mm f1.0) -- sharp in the center with a notable softness and vignetting from about half-way out -- for those interested in its creative or available light abilities.
Aug 07, 2005 at 11:24 AM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
Jack Flesher wrote:
All I'm suggesting is that the Oly offers a very attractive wideangle alternative for many photographers.
Precisely!
I continue to be amazed at the image quality offered by my Oly 21/3.5. Even though I'm not currently shooting FF (so comparisons of corner sharpness are not as big of an issue for me at the moment) I never the less continue to be impressed by this little gem of a lens.
After going through three different copies, I too have an extra sharp copy of the Canon 16-35mm. In both my real world shots as well as my own set of controlled test shots, the Oly 21/3.5 edges out the 16-35mm. I’m sure the Zeiss 21mm would perform similarly.
While not a huge difference in sharpness at some apertures, the other advantages offered by this lens make it a real winner. Price alone is one advantage, but the size difference is clearly another.
I just got back from a two-day road trip to the wide vistas of the Texas hill country (i.e., the San Antonio area). I left my Canon16-35mm at home and took the Oly. I was glad I did. The Oly 21/3.5 performed wonderfully.
Price alone is one advantage, but the size difference is clearly another.
Good point Taz.
I regularly back-pack my gear into remote locations, and due to its diminuative size and minscule weight the Oly 21 always is in the pack.
Here is a shot of the lenses I carry when shooting landscape. The Oly 21 (with EOS adapter attached) is outlined in red. The Canon 16-35 is to its right for size comparison purposes, and that's an EF 25 tube behind it; you can see thay are nearly the same size. To the left of the Oly 21 is the equally diminuative Oly 18 which I sometimes carry too.
Charles: Dohhh... I just realized you are the same Charles selling your CZ21 over at RG... Sorry for not making that connection earlier. I don't post here very often any more and this was in a gear review thread so was just sharing my opinions on lens performance...
I think that's great. I have no problem with that. I am just pointing out that this lens is no longer rare or hard to find. Is it expensive? Certainly. Is it worth it? That depends on the individual.
As for wide open performance, I did actually test that a while back, but did not save the images to post. In a nutshell, the CZ was a bit better overall at f2.8 but both were fairly poor compared to stopped down.
Well, IMO the CZ is already more than a bit better at f/8 in the corners so I bet it would be very noticeably better at f/2.8. Obviously every wide angle is better at f/8 than wide open.
Lars, do you have anything to contribute to this thread other than innuendo? --c
--c
--c