fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Leica & Alternative Gear | Join Upload & Sell

1       2       3              5       6       end
  

Archive 2011 · Zeiss tested on DXOmark

  
 
denoir
Offline
• • • •
Upload & Sell: Off
p.4 #1 · Zeiss tested on DXOmark


ulrikft2 wrote:
The sets have been ciruclating on this forum for _quite_ some time now, so the "blind" part of the test is kind of off now..



True. I've never published the last set but the first three have been posted before.

I think it's an interesting question to pursue, so I'll make a couple of new sets and post in a separate thread.


We have had quite a few 105 2.5 vs. 100 2.0 tests over at the nikon forum with a recognition grade on par with the statistical probability.


Yes, but that only says something about how good mainstream Nikonians are at recognizing different lenses. The 105/2.5 vs 100/2 test here showed that many people had no problems telling them apart.



Feb 18, 2011 at 06:10 AM
Lars Johnsson
Offline
• • • • • • •
Upload & Sell: Off
p.4 #2 · Zeiss tested on DXOmark


denoir wrote:
That's because you are relatively new to the alt forum. Yes, there are some ventures to other forums, but to a large degree it's relatively isolated. In my experience the Canonistas seldom cross the borders while Nikonians are more likely to do so. Generalizing broadly, there is a certain hostility towards alt lenses in the Canon forum and a certain degree of contempt for Canon lenses in the alt forum.


How can you say that I am new here I did my first post in the alt forum before you



Feb 18, 2011 at 07:40 AM
trenchmonkey
Offline
• • • • • • •
Upload & Sell: Off
p.4 #3 · Zeiss tested on DXOmark


Using DXO data as a benchmark for a quality lens purchase
...a fool and his money are soon parted.



Feb 18, 2011 at 07:54 AM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #4 · Zeiss tested on DXOmark


Mostly it's actually about interpreting the results - and this is the part where it falls apart for most "users" of information.

To be blunt - most people have absolutely no idea what they're looking at, what scales and relationships those measurements have to "reality" as in perceived quality, and what the quantifiable distance/difference those figures mean in relation to other lenses.

The bottom line: NEVER LOOK AT "OVERALL SCORES" - and be sure that you understand what was actually measured.



Feb 18, 2011 at 08:00 AM
Lasse Eriksson
Offline
• • • •
Upload & Sell: Off
p.4 #5 · Zeiss tested on DXOmark


ulrikft2 wrote:
The sets have been ciruclating on this forum for _quite_ some time now, so the "blind" part of the test is kind of off now..

We have had quite a few 105 2.5 vs. 100 2.0 tests over at the nikon forum with a recognition grade on par with the statistical probability.


+1



Feb 18, 2011 at 08:25 AM
Lasse Eriksson
Offline
• • • •
Upload & Sell: Off
p.4 #6 · Zeiss tested on DXOmark


trenchmonkey wrote:
Using DXO data as a benchmark for a quality lens purchase
...a fool and his money are soon parted.






Feb 18, 2011 at 08:26 AM
RustyBug
Online
• • • • • • •
Upload & Sell: On
p.4 #7 · Zeiss tested on DXOmark


denoir wrote:
RustyBug got the Zeiss right.


Which makes sense / corroborates.

Since I have never shot with Leica and very rarely with Canon, while owning five Zeiss ... the Zeiss was easy for me to spot. I waffled heavily on the other two ... not because they are 'equal' but rather because I'm not that knowledgeable / familiar with their 'telltale' characterstics.

I simply guessed Canon for the first one because it seemed to have more vignetting to me, and in my mind (thinking 17-40L) Canon's are prone to vignetting more than Leica. Of course, someone who is more versed in Canon or Leica probably would have seen a multitude of other things that would have 'clued' them better than I ... as evidenced by the three that got it right.



Feb 18, 2011 at 10:41 AM
RustyBug
Online
• • • • • • •
Upload & Sell: On
p.4 #8 · Zeiss tested on DXOmark


h00ligan wrote:
It just reiterates that DxO tests, however they are done, are not representative of real world results.

Honestly, look at some zeiss landscape shots on full frame and get back to us DxO


I'm inclined to wonder to what degree the testing is done on a FLAT subject vs. Zeiss being designed more strongly to respond toward shooting more dimensional subjects ... probably wouldn't have any effect on the results ... ya think

Not knockin' DXO necessarily ... but it's a bit like testing a Gran Prix race car on a dragstrip. The car is built for negotiating transitions, while the test is for straight line driving. So is a Muscle Car better than a Gran Prix race car, just because it gets down the 1/4 mile faster ?? In that regard, Zeiss doesn't always fare as well in 'tests' as it does in other venues. A car newbie might say that the Muscle Car was the better car because it has a faster 1/4 mile time as an 'overall indicator' ... but is it ??

BTW, a Mustang GT Convertible will never test well on the 1/4 mile track or a Gran Prix course, yet it remains a favorite for what it does do in the real world.

Edited on Feb 18, 2011 at 11:26 AM · View previous versions



Feb 18, 2011 at 11:00 AM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.4 #9 · Zeiss tested on DXOmark


theSuede wrote:
Mostly it's actually about interpreting the results - and this is the part where it falls apart for most "users" of information.

To be blunt - most people have absolutely no idea what they're looking at, what scales and relationships those measurements have to "reality" as in perceived quality, and what the quantifiable distance/difference those figures mean in relation to other lenses.

The bottom line: NEVER LOOK AT "OVERALL SCORES" - and be sure that you understand what was actually measured.


Yeah, I had a feeling that was the case.



Feb 18, 2011 at 11:24 AM
nazdravanul
Offline
• •
Upload & Sell: Off
p.4 #10 · Zeiss tested on DXOmark


If I had any doubts about DXO, the Zeiss test results removed them completely : they are IDIOTS (unintentional or not, I don't care). After extensive usage , testing of primes / zooms from 14mm to 300, in CaNikon territory, on 20MP+ sensors, my 21 - 100 Zeiss lenses kick serious butt, and only the 200 f2 primes from CaNikon, have something extra in terms of sheer IQ compared to the (good) Zeiss lenses (not all Zeiss glass is created equal ) .
As trenchmonkey said , a couple of posts before me ... ^^^



Feb 18, 2011 at 11:52 AM
h00ligan
Offline
• • • •
Upload & Sell: Off
p.4 #11 · Zeiss tested on DXOmark


denoir wrote:
AF-ON, h00ligan, crazeazn, robinlee got 3/3 correct on set 1 while mshi got 2/3.
So in this small test, the correct recognition rate was ~93%. So it is definitely possible for a trained eye to differentiate between Zeiss and Canon drawing styles.

Set 2 seems to have been more difficult

crazeazn, h00ligan & robinlee got it right on all three. RustyBug got the Zeiss right while AF-ON missed all three.


Not bad for a rookie... in fairness though I own an x1, and I have been following with keen interest the renderings in the leica thread.. so I sorta cheated

Still contrast, sharpness, and color are relatively different imho for the zeiss.. and I am not brand loyalist... but overall I love the look.

I have probably seen the first test before, but I didn't consciously remember the result.. it's different having a result posted below. Also, set 2 was imo much more difficult because of the limited dof / subject matter.

I'd take the pepsi challenge again.



Feb 18, 2011 at 01:18 PM
crazeazn
Offline
• • •
Upload & Sell: Off
p.4 #12 · Zeiss tested on DXOmark


denoir wrote:
That's because you are relatively new to the alt forum. Yes, there are some ventures to other forums, but to a large degree it's relatively isolated. In my experience the Canonistas seldom cross the borders while Nikonians are more likely to do so. Generalizing broadly, there is a certain hostility towards alt lenses in the Canon forum and a certain degree of contempt for Canon lenses in the alt forum.


I have no contempt for Canon lenses whatsoever. Matter of fact I'm covered by 3 L zooms in the 17-200 range. I think most of the alt users want a completely diff. look, contrast or weight savings (oly users) that can be done without overdoing post processing to achieve similar effects. I for one love the zeiss contrast and the fact that adapted glass makes me concentrate on taking photos like I did back in the day with manual focus cameras. I still have my uses for my AF glass too.



Feb 18, 2011 at 01:25 PM
abam
Offline
• • • •
Upload & Sell: Off
p.4 #13 · Zeiss tested on DXOmark


...long-winded post deleted.

i suppose i don't care about DxO enough to get into a debate about it, or its chosen metrics.



Feb 19, 2011 at 06:18 AM
Mirek Elsner
Offline
• • •
Upload & Sell: Off
p.4 #14 · Zeiss tested on DXOmark


madamasu wrote:
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.

Thomas

http://forum.getdpi.com/forum/showthread.php?t=23642


I only looked at the "score" and resolution. I don't shoot Nikon, but if the tests show results at center and short distance:

- I read on this forum that ZF 50/1.4 is not great on MFD and it can explain the first pair of results, no?
- Albeit at longer distances, in my testing Canon 50/1.4 was as sharp or sharper than ZE 50/2. I don't know anything about Nikon 50/1.8, but if the second set showed Canon 50/1.4 at 2.8 instead of the Nikon, the results would not be surprising for me
- Can't comment on the third set, the only 85mm I have is Canon's
- In the 100mm comparison the Zeiss is set to f/2 and Nikkor obviously to f/2.8. Again, comparing to some Canons at f/2.8 (135L, 100L, even the new 70-200), not Nikkors, the result would be no surprise




Feb 19, 2011 at 02:03 PM
scalesusa
Offline
• • • •
Upload & Sell: Off
p.4 #15 · Zeiss tested on DXOmark


There are many testers out there trying to do their best, and trying new test methods. I respect them all, but I also have my own opinion as to who does the best job(its my opinion, so I'll keep it to myself, but its not DXO).

Normally though, after you read the test method, camera used, and dig thru the data, the data pretty much agrees, even if the artifical rating they assign seems to favor different paramaters.

So, let the actual data do the talking, read it, understand it, and compare it and a pattern will emerge as to which is best for you.

I think we should be happy that we have many testers testing lenses and to let us know how they rate them.

As to lens manufacturer's MTF charts, consider that they are not stupid, the charts reflect the strongest performance of the lens, they are not about to show the weak points, I wouldn't either. They are completely accurate, as far as they go.



Feb 19, 2011 at 11:11 PM
Lotusm50
Offline
• • • • •
Upload & Sell: Off
p.4 #16 · Zeiss tested on DXOmark


I was trying to stay out of this. I've voiced my opinion on this enough times in the past. But I really feel I should address the logic of the statement below. While seemingly sounding reasonable, it really is in my opinion, misguided.

"Many testers out there trying to do their best" is useless if they all are doing it wrongly, or inappropriately, or inaccurately. Many people doing it badly, no matter how well intentioned merely further confuses and obfuscates truth and reality, and is worse that just one person doing it badly. Further, these "testers" don't preface their testing and analysis by saying they are "trying new test methods". They present this stuff as Gospel, as the highest scientific fact, and how they are revealing a hidden truth that others don't see and manufacturers actively conceal.

Some of these sites don't even mention what their test methods are, other secrete it deeply in their sites so you really have to look for it, others still describe it incompletely so you don't get the full picture. But beyond all this, most of their readers know even less than the testers. So how are they going to effectively assess methodology, even if they can find it? When someone who has knowledge, experience and insight calls these testers on the flaws in their approach, they vociferously fight back defending their "integrity" and self-anointed roles as "experts". Dare I say that Carl Zeiss knows a lot more about lens testing and measurement that the PZ boys? The PZ boys it would seem, might just give me an argument.

What I don't know if whether these testers 1) don't really know better (ignorance), 2) know what might be a better approach, but don't have the equipment, capital, time, etc. to do it properly, or 3) are completely cynical - know they are doing it poorly, but it can be done cheaply and they can sell it. I'm not sure which is worse.

If many people are doing it badly, and one is doing it right, what does that look like to the consumers who don't fully understand optical science? Majorities can be wrong. Many people coming to the similar conclusions using variations of similar flawed methods doesn't make it right.

I also reject the cynical statement that manufacturer MTF's "reflect the strongest performance of the lens". Manufacturers measure each of their lenses the same way. In particular, they are all measured (or calculated) at infinity even if the lens is optimized for a different focus distance. Special purpose lenses deviate from this, to address their special purpose -- Macro lenses are given additional measurements at close-focus distances, and tilt/shift lenses are often tested to a larger image circle (becuase they create and use a larger image circle -- even if this shows the weakest area of performance). Each manufacturer may use a somewhat different methodology to create their published MTF's, but each manufacturer measures all their lenses with pretty much the same methodology regardless how it might present a given lenses. If you actually examine these manufacturers' MTF's charts they often do show the lens' weak points. But, of course, you have to know what you are looking it. Certainly some corporate cynicism is warranted, but this is not really the most appropriate area to express it.

No, I am not happy, and we should not be happy, that so many rank amateurs and charlatans are out their presenting themselves as experts and flooding the marketplace with bad and/or misleading information. One spouting nonsense can be easily ignored. But many, propagated by the Internet, is creating a pandemic of misinformation.



scalesusa wrote:
There are many testers out there trying to do their best, and trying new test methods. I respect them all, but I also have my own opinion as to who does the best job(its my opinion, so I'll keep it to myself, but its not DXO).

Normally though, after you read the test method, camera used, and dig thru the data, the data pretty much agrees, even if the artifical rating they assign seems to favor different paramaters.

So, let the actual data do the talking, read it, understand it, and compare it and a pattern will emerge as to which
...Show more



Feb 20, 2011 at 12:37 AM
crazeazn
Offline
• • •
Upload & Sell: Off
p.4 #17 · Zeiss tested on DXOmark


at the end of the day, if you like the lenses and it takes excellent pix who gives a shit if some website refutes it with some hokey data.


Feb 20, 2011 at 03:05 AM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.4 #18 · Zeiss tested on DXOmark


Joe DSLR Public would *not even know* Zeiss made lenses for their cameras, if Canikon or the main photo publications (for whom they are big accounts) had anything to do with it. With the exception of the excellent c/a digital photography (which tests ZE lenses against Canon with predictable results (!), so all interested in the 'small experiment' thread should check them out) I see very little even basic news of CZ or anyone other than the big two.

Look, might it not be that they - consciously or unconsciously - devised weightings on metrics that tend to favour non-Zeiss lenses and ignored whole areas of image quality entirely? A lot of people hate vignetting, for example. Many people reading here think the CZ lenses to be nothing special, so DXO is probably that way inclined also.

If it has not linked in this thread here are some thoughts from David Kilpatrick on how the various companies went about lens design, written in 2007:
http://www.myspace.com/richardavedon/blog/412656651



Feb 20, 2011 at 04:52 AM
sebboh
Offline
• • • • • •
Upload & Sell: Off
p.4 #19 · Zeiss tested on DXOmark


philip_pj wrote:
Joe DSLR Public would *not even know* Zeiss made lenses for their cameras, if Canikon or the main photo publications (for whom they are big accounts) had anything to do with it. With the exception of the excellent c/a digital photography (which tests ZE lenses against Canon with predictable results (!), so all interested in the 'small experiment' thread should check them out) I see very little even basic news of CZ or anyone other than the big two.

Look, might it not be that they - consciously or unconsciously - devised weightings on metrics that tend to favour non-Zeiss
...Show more

all testers are bias in one way or another and there is no way to get around human nature, but i doubt it's worth the money for canon/nikon to try and influence testers to purposely make bias tests. all the tests i've seen in which zeiss lenses perform poorly (which isn't that many due to a lack of interest on my part) have obvious reasons for the poor performance if you look at the testing methods (which are usually hard to find but consistently applied across brands). the most obvious and common seems to be testing wide and ultrawide angle lenses that are optimized for infinity at an extremely close test distance. in any event there is a very good reason why zeiss and others don't get near as much attention as the big two - there aren't near as many people interested. especially since zeiss lenses are manual focus (except for the sony/zeiss lenses), which means 99% photographers will automatically ignore anything they do.

your linked article was very interesting and sounded about right to me until it started talking about the number of aperture blades. other than making sharper looking bokeh how does 5 or 6 aperture blades lead to an enhanced sharper look? also, during the time period he highlights as the days of minoltas best color matching most of their lenses had 6 aperture blades (not more as he suggests) while nikkors had 7.



Feb 20, 2011 at 12:21 PM
wayne seltzer
Offline
• • • •
Upload & Sell: Off
p.4 #20 · Zeiss tested on DXOmark


Lotusm50 wrote:
I was trying to stay out of this. I've voiced my opinion on this enough times in the past. But I really feel I should address the logic of the statement below. While seemingly sounding reasonable, it really is in my opinion, misguided.

"Many testers out there trying to do their best" is useless if they all are doing it wrongly, or inappropriately, or inaccurately. Many people doing it badly, no matter how well intentioned merely further confuses and obfuscates truth and reality, and is worse that just one person doing it badly. Further, these "testers" don't preface their testing
...Show more

+1

Tried to stay out of this one too, as it is pretty obvious that their testing methods and data are pretty bad and their arbitrary weighting of lens characteristics including just using the wide open number for the resolution component is retaaaaarded(use your best Boston accent here).
I had to LMAO at the corner results in the f8 field chart for the ZE 21.
I would link to it but I think FM doesn't work with png files.
As an engineer, we are taught to analyze the data we get from testing to see if the testing is being done in a way which represents real performance. I don't know how the boys at DXO could get corner results like they did at f8 for the ZE 21 and think their testing is showing real use performance.
Wish we could put our feedback on their website to tell them and people who look at these numbers how lame they really are.






Feb 23, 2011 at 02:22 PM
1       2       3              5       6       end




FM Forums | Leica & Alternative Gear | Join Upload & Sell

1       2       3              5       6       end
    
 

Welcome back
Log in to your account