Erik's web

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

Sonnar 150 CB on DSLR using Arax tilt adapter

E-mail Print PDF
Article Index
Sonnar 150 CB on DSLR using Arax tilt adapter
Orange crops
Flower crops
Bokeh crops
Resolution crops
Scheimpflug
The Orange Mystery
Issues with flare and ghosting
All Pages

NOTE!

The previosly reported problem with lens flare was probably caused by internal reflections in the adapter used. The problem will hopefully solved by painting all the interna anodized surfaces matte black.

NOTE!


Background


One of the reasons for these tests is to find out if Zeiss lenses have some mythical properties compared to contemporary japanese designs. The other reason is to find out if a Hasselblad MF lens on an Arax tilt adapter is a worhwhile alternative to more expensive TS options. 


The lens chosen for this test was one of the better of the classic Zeiss lenses for the Hasselblad, the Zeiss Sonnar 150/4 CB. I got this lens used, at a very good price. Now, is this lens OK? I think so. 


I compared the MTF curves for the Sonnar 150/ measured by Zeiss to the MTF curves for the Sony Alpha 70-400/4-5.6G APO mesaured by the Swedish periodical Foto. Foto tests at the Hasselblad factory and their measurements on Zeiss lenses are very close to Zeiss data. Foto unly publishes data at 20 lp/mm, but according to their data the 70-400 at 150 mm should be a snudge better than the Zeiss lens.


Click here for Zeiss own description of the lens.


I tested the lens on my Sony Alpha 77, an APS-C crop sensor camera with 3.9 micron pixel pitch. That camera was choosen for two reasons:


  • I wanted to use the most demanding sensor I have.
  • I wanted to use live view for pinpoint focus accuracy.


To sum up shortly here are some MTF figures:


Lens10 lp/mm20 lp/mm 40 lp/mm80 lp/mm
Sonnar 150/49081
5623
Sony 70-400/4-5.690795220
Minolta 80-200/2.8APO90815318
Zeiss Data Sheet for Sonnar ;-) 918055???


From this it is pretty clear that the Sonnar 150 tested is a pretty decent sample, being slighthly better than the other lenses at f/8. My MTF data is a ver close match for those published by Zeiss, coincidence or not?!. The Sony/Minolta lenses may hold a small advantage at larger apertures, however. (I'll come back to that issue)




Note: The images used for Imatest MTF evaluation are not sharpened. The curves that can be compared are the black ones. 


The figure belows combines the MTF curves into one diagram. Note that scaling here is different, it shows LW/PH (Line Widths/Picture Height) and lalso goes to twice the Nyquist limit. The curves indicate that the Zeiss Sonnar is performing best of the three and the Sony 70-400/4-5.6G is the weakest performer. 


The crops on the following pages are sharpened with "landscape preset" in LR4.3.


Orange crops


These crops are intended to show a possible "3D" effect.


The image below is reduced to fit page width, please click on the image for an "actual pixels view"!







Flower crops

These crops are intended to show rendition of fine detail.

The image below is reduced to fit page width, please click on the image for an "actual pixels view"!







Bokeh crops

These crops are intended to show "bokeh"
The image below is reduced to fit page width, please click on the image for an "actual pixels view"!



What I found:

Both the Sonnar 150/4 and the Minolta 80-200/2.8 have significant longitudinal chromatic aberration at large opening but this is essentially gone at f/8. The 70-400/4-5.6G has little fringing. At f/8 the Sonnar still has some color fringing, the 80-200/2.8 shows some double contours while the 70-400/4-5.6 looks just fine.



Resolution crops

These crops are intended to show resolution and microcontrast
The image below is reduced to fit page width, please click on the image for an "actual pixels view"!



What I found:

The resolution crops are the position I put focus using "live view" at maximum magnification. 

The images are pretty close. Sonnar has most detail, followed by 80-200/2.8 and 70-400/4-5.6. It seems that the 80-200/2.8 has most "snap", but this can come as much from processing as from exposure. The differences are small.




Scheimpflug


Tilting the lens is also to tilt the focal plane. There is a principle called Scheimpflug principle that focal plane, aperture plane and image plane cross along a single line. In practice that mean we can tilt the plane in focus by tilting down the lens. Look at the image below and you are encouraged to click on each image for actula pixel views.

Oh these images turn out dark. I'm going to try to make them brighter!

f/4f/11
Untilted
Tilted


The first crop shows to focus point which is the "WMF" mark on the knife.

f/4f/11
Nontilted
Tilted


This crop shows the first cup 

f/4f/11
Non tilted
Tilted


The second one is even more out of focus

f/4f/11
Nontilted
Tilted


And the third one, it cannot even be seen it is a cup! 

f/4f/11
Nontilted
Tilted




The Orange Mystery


I published some of the crops on LuLa and got some objection regarding the orange in the orange crops. So I made a very simple experiment.

I took the actual orange on the picture and made a fews spectral samples using my Color Munki Photo. The corresponding Lab values under D50 lighting would be 62, 35, 72.

So I made a small crop from the image and opened in Photoshop and did the following

  • Converted to LAB
  • Added a new selection on a layer and filled it Lab(62, 35, 72)
  • Measured a large sample of the original layer and found it to be around Lab(84,29,90)
  • Created a new layer with a new selection and filled that area with Lab(84,35,72), that is the measued values but using L-value from the image sample to compensate for different lightness.

The resulting image is shown below:


Next I blurred the layer containing the original image, so texture was eliminated:

I think this demonstrates that the rendering of the orange is correct. Mystery solved?


Possible causes for the mystery:

  • The original images were posted as 16-bit TIFF in Prophoto RGB, by mistake. My browser was handling them correctly. Hopefully it is fixed now?!
  • Could be the my oranges have an odd color. They may have had a bleak color when I bought dem. Also, they have been around for a while and may have gone a bit dry in the pealing.

Conclusions (preliminary)

It seems that the Hasselblad Sonnar 150/4 is at least as good than the other tested lenses. Regarding MTF it is a tiny bit better. It has issues with color fringing, but it goes away when stopped down to f/8. A very good lens at a low cost of 300€! 

  • The images from the three lenses tested are very close, I don't think the differences would be visible in A2-size prints.
  • I posted two questions on Luminous Landscape forums regarding which of the images were preferred. Most posters preferred the left image, that is coming from the Hasselblad lens. Very few poster preferred the Sony 70-400/4-5.6 and no one the old 80-200/2.8 APO. This may indicate that classic Zeiss designs may have qualities that make them attractive.
  • The lens seem to work well with the ARAX adapter. A usable tilt configuration at very low cost.
  • I have noticed a lot of ghosting when shooting outside with sun shining from the side. Got hints in a LuLa posting on this being an issue with this lens, too. This lens really need a lens shade. 



Issues with flare and ghosting

The sample I have has more problems with flare and ghosting than any other lens I have seen. I don't have a lens hood for this lens but I have one on order, hopefully it helps. It is possible or even probable that the sensivity to flare is related to my sample, it may be one of the early T* coated lenses. 
Sonnar 150/4Sony SAL 70-400/4-5.6

Bild2




Last Updated on Tuesday, 23 April 2013 10:24  

Statistics

OS : Linux u
PHP : 5.2.9
MySQL : 5.0.67
Time : 01:57
Caching : Enabled
GZIP : Disabled
Members : 1510
Content : 72
Web Links : 1
Content View Hits : 381616

Login

Banner1

J!Analytics