A very sharp Nikon prime vs. the Beercan

Started 9 months ago | Discussions thread
ForumParentFirstPreviousNextNext unread
Flat view
Bruce Oudekerk
Senior MemberPosts: 2,677Gear list
Like?
A very sharp Nikon prime vs. the Beercan
9 months ago

Lately I've been using my old Beercan for reprographics because it has the lowest geometric distortion at 100mm of all the lenses I own at that focal length. To ensure everything is in focus I have been shooting at f8.0.

Concurrently with this, and obviously unrelated, I was out shooting in the bitter upstate NY winter cold using my trusted old friend, the Minolta 28-135 on my Sony a850. Very atypical for this go-to lens, I had a lot of fuzzy images. I was so concerned, the next day I spent time testing (indoors:) and comparing it to my Sony 70-300G, the Minolta 70-210 Beercan and even the old Minolta 100-200 f4.5, which I never use. I tested them at 100mm to keep it simple and speed up the process. The quick results are that at 100 mm the 100-200mm was maybe the sharpest (of course its actually more like 105mm, I think), the Beercan was next and the 70-300G after that, but I'm really splitting hairs here. Last, but not least, is the 28-135. All four put in what I considered an excellent showing for zooms. But there's always more to the story.

I had created a lens testing setup using a version of the old USAF 1951 resolution test chart where I rotated that image 0, 30 and 45 degrees and the camera is placed at an arbitrary distance for comparison between various lenses. It is closer than ideal but still stresses any available camera/lens combination. In this case the tripod was set up for the 100mm testing.

A few nights later, friends came over and one of them had his Nikon D3X. I wanted to compare his wonderful Nikon 85mm f1.8 to the almost 30 YO Minolta 70-210 f4.0 'Beercan' in this very artificial setting as some sort of reference. Mounting these lenses on the 24MP Nikon D3X and its sensor cousin, the Sony a850, seems like a fair comparison.

This image shows my test setup, such as it is, and the highlighted area in the central portion of the chart that I will be observing

Further below is a small composite portion of test images from these two 24 MP cameras. The top of the image is at 100%. The bottom is an enlarged detail of the detail. The zoom EXIF reports 90mm but that was as close as I could get to the field of view of the 85mm prime.

The Nikon was exposed at ISO 200...tungsten balanced and the Sony at ISO 160 at 3300K. In ACR I turned off default sharpening and white balanced the 'paper white' on both images. On the Sony, I then adjusted exposure to match the 'paper white' on the Nikon and then adjusted the black level to also mimic the Nikon...effectively normalizing the two images.

In Photoshop, I aggressively sharpened the Nikon and then tried to duplicate that look with Sony by using even more aggressive sharpening settings. (USM 200, 0.4, 3 vs. 230, 0.4, 4) I don't show this here but I am amazed that all of the chart sections retain relative relationships between both lenses, including the rotated chart portions and near corner.

You will need to look at this at 100% to make anything of the image!

The results are interesting to me in that the old a-mount Minolta zoom holds up well to a current and highly regarded Nikon prime that many would consider a near benchmark lens. The differences are interesting. While sharpness and resolution in this case give only a slight nod to the Nikon, this lens is clearly much newer and possesses more sophisticated optical coatings and formula (not to mention its a prime) and thus has MUCH, MUCH better micro-contrast. I can see an anomaly on the screen at 100% but can not really perceive WHAT it actually is. Resampled using nearest neighbor to 1000%, it becomes very obvious what is happening. In any real world scenario this is (probably)meaningless but is necessary to show what I can barely perceive. Perhaps it is obvious to some.

As an aside, I tried straightening the Nikon image just to see how it would negatively affect resolution perception (it was off by a whopping 1.35 degrees...OK so I'm anal retentive). Surprise, surprise, if anything it made it better. Surprising, in that PS could be this subtle and not surprising, in that it straightened a chart that has a horizontal/vertical orientation.

...and for those wringing their hands, the newer Sony 70-300G had the best micro-contrast of my lenses tested but was inferior to the Nikon prime. DXO seems to show a more or less similar relationship to what I found between the Nikon 85mm f1.8 and the 70-300G at f8.0. The surprise is the Beercan which, to my eyes falls in the middle.

Take away what you will from my testing. I found it interesting so I'm sharing this for whatever will be made of it.

Bruce

Nikon D3X Sony Alpha DSLR-A850
If you believe there are incorrect tags, please send us this post using our feedback form.
ForumParentFirstPreviousNextNext unread
Flat view
Post (hide subjects)Posted by
ForumParentFirstPreviousNextNext unread
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow