Erik's web

  • Increase font size
  • Default font size
  • Decrease font size


E-mail Print PDF
Article Index
Flower samples
Side by side
Orange crops
Zeiss quality control
Why do I have 5 70 mm lenses
Why these tests were made and other observations
So, what is the outcome for me?
All Pages

Please note: I just begun writing this article. I will add more samples later and the samples may be corrected. I will also make all relevant raw files available for download.

Note 2: Hopefully I can post some new samples with A Zeiss Sonnar 150/4 for Hasselblad included in a week or two. 

I have read a lot about Zeiss lenses having magic qualities, like 3D or different color rendition. It is not clear to me if these qualities are myth or reality. Nor is it clear to me what is meant by 3D effect. Most articles on the net are about short DoF and show small JPEGs as evidence. Some other articles discuss microcontrast. 

I have realized that I have two lenses with a degree of Zeissness so I decided to compare those to three other lenses that were neither designed, made or certified by Zeiss.

All the tested non Zeiss lenses are excellent performers at the tested focal length of 70 mm. So we are comparing the Zeiss lenses to a strong competition.

Also I would add that there are many things to like about Zeiss. Good information, nice people. I also got the information that Zeiss is a foundation intended to take good care of employees and promote research in optics. I do think Zeiss is a great company ding great things and making excellent lenses.

How the test was made

There were a couple of aspects I wanted to investigate:

  • One was about color rendition. A statement I read was that T* coated lenses would have significantly better transmission in the blues compared to other lenses and would give a significantly different color rendition.
  • The other issue I wanted to find out if there was a "magical" look to in focus images.
I used a setup including a small USAF test target, a one Dollar bill, a color checker, an orange and small flower (pelargon). All these were carefully aligned to be at same focal plane using a laser distance meter. The camera used was the Sony Alpha 77 SLT. Focus was done using live view at maximum magnification with focus being on the USAF test target.

Exposure was ETTR so blinking highlights were avoided on the white patch of the color checker. The images were adjusted in LR4 to match on white and black patches. Color balance was set on light gray patch (second from left). 

The camera used was an APS-C camera. It has a pixel of 3.89 microns, making it one of the most demanding sensors. Using APS-C obviously ignored edges and corners on full frame, but I wanted to find out about "peak performance.

Testing for weakness or for strength?

As I see it, I can test for weakness or for strength. 

Testing for weakness is intended to find out limitations. Most lenses have a weak performance in the corners. Whenever I'm thinking about buying a lens I check out the corners. If the corners are good, the rest of the lens is probably OK. If the corners are bad, I would reconsider the lens. Corners often improve on stopping down.

The other way is to test for strength. Finding out what the lens can deliver when it is at it's best. Almost all lenses perform best at medium apertures, the best lenses peak at f/5.6 or so. Most lenses perform best at the center. The area where the lens reaches peak performance is often called the sweet spot. Optimally the sweet spot would cover the whole sensor area. 

The image below is a screen dump from the DxO-mark website and shows the "sweet spot" of the 24-70/2.8ZA and that of the Sony 85/1.4 ZA lens, both at f/5.6 on a Sony Alpha 700 (APS-C):

The image indicates that both lenses resolve about 50 lp/mm. The 85/14 has significantly better corners while 24-70 may have a somewhat better sweet spot.

The next screen dump shows the same image but with the text subject overlaid. We can see that the test subject is inside the sweet spot of the 24-70/2.8 at f/5.6.

Finally the image below shows the resolution plots of the Zeiss 85/1.4 and the Zeiss 24-70/2.8, the resolution plots indicate that the ZA 24-70/2.8 is slightly better at the center compared with the 85/14, when both are stopped down to f/5.6.

It has been argued that primes should be used for the comparison. What I show here is that the tests are using a part of the image where the zooms actually keep up well with the 85/1.4 prime.

Also, all the lenses used here are zooms. The telephoto zooms are at their short ends. 

Zeiss characteristics

I have read some articles by Erwin Puts, and Lloyd Chambers. Also I have studied a lot of MTF data coming either from Zeiss or Hasselblad. I'm not intending to say that I'm an expert of Zeiss, but I have made some reading.

The impression I have is Zeiss most often aims for:
  • Reasonable designs with good performance
  • Avoid designs which are very demanding on tolerances
  • Designs lenses with a wide sweet spot but often not very good corner performance
MTF measurement

The camera used was a Sony Alpha 77, a 24 MPixel APS-C camera, very demanding.

MTF measured on a vertical slanted edge off axis (mid field). Two values given, LW/PH where MTF drops to 50% and LW/PH where MTF is 18%. No sharpening was applied on these samples.

lensLW/PH 50%LWPH 18%
ZA 16-80/3.5-4.515042657
Minolta 80-200/2.8 APO15422722
SAL 70-300/3.5-4.5G SSM14602737
SAL 70-400/4.0-5.6G SSM15772922
ZA 24-7014502706

From the MTF data I would draw the conclusion that the two Zeiss lenses are the weakest contenders at 70 mm. Now both of these lenses are zooms at maximum focal lengths while the others are zooms at shortest focal length.

Last Updated on Friday, 28 December 2012 17:48  


OS : Linux u
PHP : 5.2.9
MySQL : 5.0.67
Time : 13:52
Caching : Enabled
GZIP : Disabled
Members : 1510
Content : 72
Web Links : 1
Content View Hits : 359340