DXOmark have tested several Zeiss lenses recently, and if they have anything in common, it's not 3D or high micro contrast, but rather abysmal performance. Someone on the Nikon forum at get dpi has shown DXO comparisons with equivalent Nikkor lenses, and they all sweep the floor with the Zeiss. The Distagon 21 and MP 100 in particular look really bad. It would be interesting to hear your opinions about this matter. I have my own theory but I will keep it to myself as it involves some pretty nasty language
My theory also involves nasty language My ZF21 and ZF.2 100 are brilliant, and anyone who tries to tell me otherwise is out of their mind. The 105/2.5 is also extremely good though.
Here is the link Edward mentioned, so you can see the comparative statistics for yourselves, and also read comments that make a lot of sense. I don't own any Zeiss lenses, had some of the Nikkors, but now use my old Leica lenses, so the whole comparison is not really important to me. I'm not interested in statistics and tables anyhow, I want to see pictures, and then I get an opinion, which is of course very personal.
A quick look at the resolution lp/mm makes me wonder if that is center, edge, corner or an average of some kind ... also, at what apertures?
My experience with my Nikon 20/2.8 AIS is that it has sharper central resolution than my Oly 21/3.5, but the Oly has more even sharpness across the frame. I find my Nikon's to be designed midway between my Oly's and my C/Y's in terms of sharpness distribution, vignetting, contrast, etc. The story repeats itself with my 28mm, 50mm and portrait glass.
Optics are always a series of compromises / tradeoffs. Engineers know this and brands typically design fairly consistently toward their philosophical drawing style. In that regard, the 'single' value of resolution doesn't tell the whole story ... 'grain of salt' might be warranted.
I recommend taking a look at those some lenses in question over on www.slrgear.com ... not sure I'd come to a 'floor mopping' conclusion after looking there. Particularly taking notice of the sharpness distribution mapping variance between the Nikon's & Zeiss. Central sharpness seems to typically be a horse race till you start moving outward at wider opertures. Pretty consistent with what my Nikons do.
The presentation form that makes most sense to me on DxO is the MTF plots.
If you want to look for microcontrast, you could check the modulation transfer for any spatial frequency you want.
What doesn't make sense, is the results. For starters: Practically all lenses tested on Canon 5DII, yield a modulation transfer of 91-92 % at 10 lp/mm in the center, green channel, at f/4 - f/5.6. How likely is that?
Oh, I don't think I want to continue. It could get nasty indeed.
First of all, I just want to say that I have NEVER been a fan of DxO's test results because they take very disparate aspects of the product they are testing, try to quantify its performance in each measure, and then they do some weird nonsensical number voodoo to come up with some final value that is supposed to summarize how "good" it is.
For example, just looking at their lens testing, you see they come up with a number for resolution, vignetting, etc., but it's only by looking at the details that you can get any understanding of what the lens is actually doing with respect to these measures. One number for resolution isn't telling you whether the center resolution is good/bad versus the corner.
In summary, DxO scores are pseudoscience, dumbed-down, prettified bar charts for non-technical photographers who are too naive or innumerate to know how to dig deep and question the analysis. The same can be said for Photozone.
BUT--I also find the reactions to these results just as telling. Several Zeiss users have come out and blasted the results and one really must wonder whether this sort of defensiveness would have happened if the lens brands were not disclosed. I think a lot of photographers get caught up in the fanaticism of their brand loyalty and have a very distorted idea of how well certain lenses *actually* perform. I hate to break it to some of you out there, but Zeiss lenses aren't magical, and the way some of you have reacted only demonstrates your personal bias and your inability to make objective comparisons.
Let me be absolutely clear--I think the DxO results are trash; fundamentally flawed. But I also think some of you so desperately want to believe that your lenses are so much better than anything else that you aren't willing to accept the possibility that you paid a lot of money for something that isn't necessarily all that worthwhile. It's a well known psychological phenomenon--buying something expensive, and then trying to rationalize that purchase by only seeing the good things about it, and refusing to see its shortcomings, or accepting the possibility that something cheaper could actually outperform what you bought.
Another assumption on everybody's part is that the words "looks good" which is a very subjective phrase has been truly quantified. What constitutes "looks good" to me is not necessarily yours. And at this point in our technology / aesthetics, I challenge anyone to put a number (or even a couple) on that.
Anyway, I started putting DxO into the Ken Rockwell folder when they specified that the noise from my D300 was worse than my D70....
The only good source for Zeiss MTF measurements are.. Zeiss MTF charts. They are the only ones that actually do proper MTF measurements. Could be because they are the ones actually making the equipment for doing MTF measurements.
Wickerprints, this is not the first time you are on an anti-Zeiss crusade. Blind hatred for a brand is just as misguided as blind loyalty.
wickerprints wrote:
It's a well known psychological phenomenon--buying something expensive, and then trying to rationalize that purchase by only seeing the good things about it, and refusing to see its shortcomings, or accepting the possibility that something cheaper could actually outperform what you bought.
Indeed, but what you are demonstrating is also a well known psychological phenomenon - it's called cognitive dissonance (i.e sour grapes).
Wickerprints, I agree that the biggest problem with DxO is how the results are read, but I'm not sure that I'd put their testing on par with Photozone. From a testing procedure standpoint, DxO is pretty legitimate, and it is a decent source if one knows how to dig deeper and interpret the numbers.
What's interesting to me is that DxO also has one of the most popular lens correcting softwares, so I wonder how this software and their Mark lens scores correlate??
denoir wrote:
The only good source for Zeiss MTF measurements are.. Zeiss MTF charts. They are the only ones that actually do proper MTF measurements. Could be because they are the ones actually making the equipment for doing MTF measurements.
Wickerprints, this is not the first time you are on an anti-Zeiss crusade. Blind hatred for a brand is just as misguided as blind loyalty.
Indeed, but what you are demonstrating is also a well known psychological phenomenon - it's called cognitive dissonance (i.e sour grapes).
Wow, your post is so -worthy.
You see, you've made a very simple error. I don't hate Zeiss lenses at all. In fact, I think they're quite nice. What you have mistaken for brand hate is actually a dislike for those who take their subjective feelings about a brand and arrogantly try to convince everyone else their choice is superior--i.e., brand loyalists such as yourself. The lenses themselves are fine. But like everything else, they are just lenses. They're not going to make your mediocre pictures look special.
I love my Zeiss lenses, but not because they are Zeiss lenses. They could be better, but they are already excellent (I have the 21 and 100), and I haven't seen any Nikon or Canon primes to match these two. Both are very sharp, and have very nice boke. What else could you want (leave "less CA" out of this )?
wickerprints and denoir, you should meet IRL and become friends
It would be interesting if both of you apply your knowledge and give us a little more insight into why the DxO lens measurement results are so weird, this is the topic of this thread
douglasf13 wrote:
Wickerprints, I agree that the biggest problem with DxO is how the results are read, but I'm not sure that I'd put their testing on par with Photozone. From a testing procedure standpoint, DxO is pretty legitimate, and it is a decent source if one knows how to dig deeper and interpret the numbers.
I agree that the testing procedures themselves are probably on the up-and-up, but the testing assumptions are rather suspect to me. Any review that purports to take such disparate metrics of a lens' performance and quantify it into a single number, to me, reflects either a willful desire to misrepresent one's data, or a very ignorant understanding of that data. It's not enough that DxO offers to break it out and let you dig into the details, because the mere act of summarizing is cause for suspicion.
What's interesting to me is that DxO also has one of the most popular lens correcting softwares, so I wonder how this software and their Mark lens scores correlate??
I honestly don't know. I'm not sure that they really do correlate all that much. The ability to correct for a lens' aberration characteristics is somewhat (but not entirely) removed from the ability to quantify how a lens performs. I'm sure they have great software developers who understand image processing techniques, but I'm not quite so convinced they have statisticians or optical engineers working for them.
wickerprints wrote:
You see, you've made a very simple error. I don't hate Zeiss lenses at all. In fact, I think they're quite nice. What you have mistaken for brand hate is actually a dislike for those who take their subjective feelings about a brand and arrogantly try to convince everyone else their choice is superior--i.e., brand loyalists such as yourself. The lenses themselves are fine.
That's a nice theory, but it doesn't hold up as this is not the first Zeiss thread that you are trolling. As for being a brand loyalist, from your profile it seems like you have only Canon lenses while my current selection includes in alphabetical order Canon, Leica, Minolta, Pentax, Sigma, Tokina & Zeiss.
But like everything else, they are just lenses. They're not going to make your mediocre pictures look special.
If mediocre pictures are your problem, then I agree that a lens won't help much. But that's your problem - there are photographers that know the value of a lens and how to use one.
Anyway, one should not feed the trolls. My bad. Carry on.
I love my Zeiss lenses, but not because they are Zeiss lenses. They could be better, but they are already excellent (I have the 21 and 100), and I haven't seen any Nikon or Canon primes to match these two. Both are very sharp, and have very nice boke. What else could you want (leave "less CA" out of this )?
Indeed. I have always been deeply impressed by the performance of the ZE Distagon 21/2.8. The corner performance is quite remarkable. It's only real flaw is the complex mustache distortion, which I think is forgivable given modern digital image processing technology.
The 50/2 MP in particular also impresses me, again mainly for its performance at f/2; but to be fair, I only mention it because it seems like no one else is making a really decent 50mm lens these days!
Lateral color, in my opinion, is not a huge deal--if the lens is delivering high corner resolution, it's easily corrected in post. Axial color is the more problematic of the color aberrations, as it's hard to correct and moreover, relatively few lenses control it well (with notable exceptions being expensive telephoto primes and other apochromats).
The thing to remember about Canon and Nikon glass is that they are designed for autofocus systems and low-light performance. It's a different design goal and it's just not going to give the same results as a manual-focus, slow-aperture formula. People tend to want it all--solid metal build (but keep it feather-light!), long focus throw (but still with AF!), perfect corners (but at f/1.4 and minimal vignetting!)--all for the low price of $19.99. It's not realistic.
If you're happy with Zeiss, be happy with it. But a lens--any lens--only projects the image you choose to let it see.
alundeb wrote:
It would be interesting if both of you apply your knowledge and give us a little more insight into why the DxO lens measurement results are so weird, this is the topic of this thread
I've never considered their lens tests to be worthy of closer examination. I gave up on their sensor tests a while ago when I saw that they had little in common with reality. And their sensor tests are supposed to be better than their lens tests.
If you want a solid performance evaluation of a lens, you need measured MTF curves and no review site active today provides them.
Feb 17, 2011 at 05:58 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
denoir wrote:
The only good source for Zeiss MTF measurements are.. Zeiss MTF charts. They are the only ones that actually do proper MTF measurements. Could be because they are the ones actually making the equipment for doing MTF measurements.
Wickerprints, this is not the first time you are on an anti-Zeiss crusade. Blind hatred for a brand is just as misguided as blind loyalty.
Indeed, but what you are demonstrating is also a well known psychological phenomenon - it's called cognitive dissonance (i.e sour grapes).
Sorry, I can't help myself here because I study cognitive dissonance. What wickerprints initially described--glamorizing the positive aspects of something you have chosen and failing to see or fully acknowledge it's faults--is what is a classic example of cognitive dissonance and it is a well known psychological phenomenon and we probably all do it to at least some extent at least some of the time. Sour grapes--or denigarting what you cannot attain as in the classic Aesop fable--could be argued to result from cognitive dissonance, but is not a classic example of it. Neither is there clear evidence demonstrating under what circumstances the phenomenon occurs. Simply put as a psychological phenomenon it is much more poorly understood.
My ZF21 and ZF.2 100 are brilliant, and anyone who tries to tell me otherwise is out of their mind. The 105/2.5 is also extremely good though.
-worthy.