Last weekend morning I went to Stanford to take comparison shots with my Contax N 50/1.4 and ZE 50/2 MP to show the different rate of focus falloff between the two lenses especially stopped down. This topic came up again recently in the Zeiss thread but I first noticed it last year while shooting comparison landscape shots and mentioned in a discusion Samuli were having about the two lenses on pg 26 of Zeiss Z* thread. Samuli then posted some forest comparison shots which showed this behavior where the 50 P focus falls off at a faster rate from the focus point than the 50MP.This helps better isolate the subject from the background and shows spatial relationships between objects in a scene from front to back. It also renders volume and shape better in objects.
I have shot these comparison shots on a tripod with MLU and focused using LiveView at 10x with a Hoodman Loupe. No PP except downsizing and sharpening for web.

#1 50P at f1.4 (full scene)
#2 50P at f2 (full scene)
#3 50MP at f2 (full scene)

#4 50P at f2.8(full scene)
#5 50MP at f2.8 (full scene)
#6 50P at f4 (full scene)
#7 50MP at f4 (full scene)

#8 50P at f5.6 (full scene)
#9 50MP at f5.6 (full scene)
#10 50P at f8 (full scene)
#11 50MP at f8 (full scene)

I'm surprised by how much the far-background blur discs change in size between the P and MP lenses; I suppose this could be mostly due to the ~6"-1ft focus point difference between the two sets of images (P appears to be focused closer than the MP), which is a fairly substantial shift at this focus distance.

#12 50P at f11 (full scene)
#13 50MP at f11 (full scene)
#14 50P at f16 (full scene)

#15 50P at f1.4 (far crop)
#16 50P at f2 (far crop)
#17 50MP at f2 (far crop)

#18 50P at f2.8 (far crop))
#19 50MP at f2.8 (far crop)
#20 50P at f4 (far crop))
#21 50MP at f4 (far crop)

#22 50P at f5.6 (far crop))
#23 50MP at 5.6 (far crop)
#24 50P at f8 (far crop))
#25 50MP at f8 (far crop)

#26 50P at f11 (far crop))
#27 50MP at f11 (far crop)
#28 50P at f16 (far crop))

quite intersting, thanks.
totally subjective... but does the mp have more/less 3d than the P?

With all else being equal, wouldn't you expect to see more structure in the OOF areas from the lens with a significantly smaller front element?

wayne seltzer wrote:
This helps better isolate the subject from the background and shows spatial relationships between objects in a scene from front to back. It also renders volume and shape better in objects.


You're saying that about the MP, right?

Nice work, Wayne.

If I had to guess based on your images, I would sat that the 50MP had perhaps a half a stop (maybe a little less) worth of additional DOF over the 50P.

And as much as people have issues with the bokeh produced by the 50mm Planar, it looks quite reasonable in your samples.

freaklikeme wrote:
With all else being equal, wouldn't you expect to see more structure in the OOF areas from the lens with a significantly smaller front element?

Yes and no. The smaller front element makes the MP more likely to vignette at f/2, which expands DOF towards the corners of the frame (the blur discs are cut off). But there is no theoretical difference in the middle of the frame, if you don't account for different design in terms of correction for spherical aberration.

I think much of the "phenomenon" is because they are not focused exactly the same. For some reason everyone that tries to compare these lenses have the same difficulty to focus them accurately to the same distance, so something is clearly going on with the designs.

Thanks Wayne. Very interesting comparison. I just got my MP, and I know why!

Makten wrote:
I think much of the "phenomenon" is because they are not focused exactly the same. For some reason everyone that tries to compare these lenses have the same difficulty to focus them accurately to the same distance, so something is clearly going on with the designs.

The N f/1.4 P apparently has no floating elements. The MP does. And itīs a fair guess that the P has undercorrected spherical aberration to enhance the bokeh, as has e.g. the new 35/1.4. As a consequence it has some degree of focus shift, which obviously changes the field of best focus when the aperture is changed.

Focusing is critical when doing these kind of tests. Iīm almost desperate with my testing of the 250mm Superachromat vs Vario-Sonnar vs... I see no other way than focusing tethered on my 27" iMac. Not very convenient. But 10x live view is not up to snuff when doing critical comparisons. Not if you really want to know.

Thanks for the responses, I didn't get to post the mid-crops yet which I will do now.
One thing about the focus of the 50P is that this is the Contax N version which does not allow stopped down focusing, aperture only closes down when shot is taken.
So if there is any focus shift, then I can't do anything about it since I have a 1ds3 and not a 5d2, I can't use the movie-mode trick. I will provide mid and then near crops.I will take a close look at the near crops to see if focusing has shifted or is different between each of the two lenses.
Middle crop is the 2nd cup from the front.

#29 50P at f1.4(mid crop)
#30 50P at f2 (mid crop)
#31 50MP at f2 (mid crop)

#32 50P at f2.8 (mid crop)
#33 50MP at f2.8 (mid crop)
#34 50P at f4 (mid crop)
#35 50MP at f4 (mid crop)

#36 50P at f5.6 (mid crop)
#37 50MP at f5.6 (mid crop)
#38 50P at f8 (mid crop)
#39 50MP at f8 (mid crop)

Thanks Wayne! Interesting. To me the MP seem to have more 3D/pop.

A

#40 50P at f11 (mid crop)
#41 50MP at f11 (mid crop)
#42 50P at f16 (mid crop)

I accidentally had #33 and 34 in view in stead of #32 and 33. They look very similar in most aspects (P at 4, MP at 2.8).

Have a look or two at Nasseīs paper Depth of Field and Bokeh, perhaps p35 and forward. My guess: The P has under corrected SPA, the MP (almost?) corrected, or at least far less SPA. That would affect the lightness distribution on the OOF point spreads in significantly different ways. Although we seldom look at the quality of closer than sharp field bokeh, this might add some insight into the matter on differences between the P and the MP. My prediction is that the P should have uglier close up bokeh than the MP.