How about a 20 mm lens test?

I've been meaning to perform this test for quite some time, but I've finally gotten around to it. As some may know, I've been searching for the holy grail of wide-angle lenses for some time. I had a 17-40L which turned out to be bested by my 16-35L. I had a Sigma 20/1.8 that didn't quite cut the mustard. I had a Canon 20/2.8 which came close but didn't quite make it either. Against those 3, my 16-35 came out on top.

So now, I took out the Zeiss-Jena 20 mm f/2.8 Flektogon and gave it a whirl against the Canon 16-35L at 20 mm (well, 19 in a couple of shots, but close enough). Here's how it fared.

Testing was done in such a manner that I tried to keep things as equal as possible. The Zeiss consistently exposed a bit higher so I tried my best to compensate and equalize the exposure either with exposure compensation in-camera or in DPP. All shots were raw, and both lenses were taken through a series of f/2.8 to f/8 in 1-stop increments. In hindsight, I probably should have shot f/11 and f/16 as well. Maybe next time.

Test camera was the 5D. CZ was manually focused using the angle finder at 2.5X magnification with a couple of runs to make sure I had the best focus. Focus was on the power line pole. The Canon was auto-focused as this is how I intend to use it.

First, the test scene, from a tripod off the back porch:

That was an excellent review. It just goes to show that the "holy grail" is built on reputation more than fact.

I always knew the 16-35 was an awesome lens. Just wish I owned it instead of the 17-40, which is an excellent lens in and of it self, but I do miss that extra stop.

Just to note, this 20 mm Flektogon is not to be confused with the 21 mm Contax/Zeiss. Contax is a much better manufacturer. That Contax-Zeiss 21 mm lives up to its reputation, at least from all the great tests I've seen.

I agree about the 16-35 - it is underestimated, and provides a very good lens, especially for an ultra-wide, fast zoom. Mine performed a tiny bit better in the corners at 16 mm & f/2.8 on the 1D II compared against my 17-40 at f/4. That is why I've stuck with it.

This test backs up my opinion of the Flektogon design--exceptionally sharp in the center and soft in the corners. Your pictures indicated to me that the Flektogon was sharper than the Canon in the center of the image.

When I tested medium-format lenses in the Pentacon Six mount, I found my Flektogon 50mm f/4 lens to be among the sharpest in the test in the center, and among the softest at the edge.

If your subject is in the center, the effect is not necessarily a problem. As a landscape photographer, however, I prefer a lens with even performance edge to edge, even if it's not quite as good. It's quite likely that having the very sharp center at the expense of sharpness at the edges was an intentional design compromise. Remember this lens was designed in the early 50's when nobody expected such wide-angle lenses on SLR's, and gladly took what they could get.

A group of us who collect all this old ex-Soviet stuff are conducting a big lens test next week where we will attempt to quantify the limiting resolution at MTF values of 50 and 10 percent. 50mm and 65mm Flektogons (both of which are wide angle in the 6x6 format), including a number of samples of each, will be tested. I'm curious to see if the sharp-center/soft-edge performance is a congenital trait of the design, as I think it is. I'll report back in this thread if I find anything interesting.

Rick "who thinks the CZJ Sonnar is world class, but the Flektogon a bit dated" Denney

All true, of course, but I don't think the Flektogon suffers from poor manufacturing. Poor collimation or alignment would adversely affect the center sharpness, and this lens does not show a problem there. I'm convinced the effect you have seen was a design issue.

The East Germans were never motivated to update their designs. They made changes to the barrels and added better coatings over the years, but I seriously doubt they ever fundamentally recalculated the lens formulas. West-German Zeiss, on the other hand, kept their R&D process going throughout, making steady improvements. When the Zeiss factory workers fled the Russian zone (with the help of the U.S.), it was primarily the scientists and designers who did the fleeing. The production capability stayed in Jena for the most part. The early-50's Distagons were no better than the Flektogons, I suspect, but they had the tools to improve the designs over the years. Back in the 50's, the Flektogons for the (then) Praktisix had as good a reputation as the Distagons for the Hasselblad.

Of course, the modern Distagons are vastly pricier than the old Flektogons.

Rick "who thinks the flare performance is pretty typical of Flektogons" Denney

an expensive extra stop though
(twice the price of the 17-40)

Indeed it is, and if not for the improved corner performance, I'd have kept the 17-40. It really isn't much different in performance over most of the frame.

Looking at the review sites Ive seen mixed reports on whether its better than the 17-40, seems in some cases it's worse

And then there's some that say they are basically the same except for the stop. A lot depends on the copy(ies) they used.

And yah, you are paying for a stop. That goes for any lens though, pretty much.

That is the point - apart from the extra stop don't banlk on getting a better performer, you might - yuo might not. it's a gamble

Ok, I'll qualify that - my copy of the 16-35L is better in terms of edge and corner performance at f/2.8 than my 17-40 was at f/4 at the widest focal length when I compared them on the 1D II. One would expect the differences to be greater on full-frame, but at the time I compared, I didn't have full frame available. The vast center of both lenses is very good.

Also consider light falloff - the 16-35 does perform better in this respect, and that is not a trait that varies significantly from copy to copy. This is not an issue on 1.6X cameras.

That could make it a better performer depending how you use it. I'm sure with what I do, I would be able to push it to limits where the f/4 might not stand up(holding all other things constant and no flash).

Obviously the 2.8 becomes more important on the longer telephoto lens, but there's always that difference between the stops in the 2.8 versus 4 or 2.8 versus variable zooms.

see here:

https://photography-on-the.net …php?p=1020720#p​ost1020720

If he isn't happy with the 16-35L, he's probably not going to be happy with the 17-40 or anything else 20 mm or wider. There's of course some light falloff with the 16-35 at 16 mm, f/2.8, which diminishes as stopped down (as expected due to 1-cos^4 physics - see http://www.vanwalree.c​om/optics/vignetting.h​tml ). Same on the 17-40, but to a larger extent especially if used at identical stops.

There's been some discussion of these two lenses on FM amongst people with full-frame cameras.

CeeCee is right about the 35/1.4L though - it's a phenominal lens. Though it does vignette on full-frame at very wide apertures as well, but its insignificant at f/2.8 according to Castleman's tests. Even has a little coma wide open. But no lens is perfect, and the 35 is as close as any.