Sometimes I put up posts basically because I find myself answering the same email over-and-over. One of the emails I get a lot is: "I want the best copy of the xxxx zoom lens."
People get irritated because my response is usually something like this: Which focal length do you want the best copy at? The long end, short end, or in the middle? By best copy do you want the highest center MTF, the best corners and edges, or is the middle of the image the most important to you? Or do you want the one with the flattest field? And again, at which focal length do you want whatever is the most important thing to you to be the best at?
Generally, people miss my sarcasm and just respond that they want the best overall lens. I could probably satisfy that request for a prime lens (it would be splitting hair, but it could be done). But there's no way to do it with a zoom; there are just too many variables, and the variables interrelate and counteract each other in complex ways.
So today I thought I'd just put up some pictures and demonstrate how one single variable, field tilt, changes on a given lens at different zoom distances. We chose Canon 70-200 f/2.8 IS II lenses because they are excellent. So you fanboys that want to take this out of context and say Canon 70-200 f/2.8 IS II lenses have tilt variation, trust me, it's less than the zoom you are a fan off. Every zoom lens made is like this.
Field Curvature Examples
All of these lenses had been tested just prior to running these field curvatures both with test-chart photographs and on our MTF optical bench. All of them were optically at least average, if not slightly better than average by my standards, which are quite a bit higher than what the factory standards are (we reject about 5% of lenses that factory service tells us are fine).
The first lens I'll show you was the weakest of the 12 we tested (although I'll mention again, it was still an "average" copy). It looked fine on all of our usual optical tests, but on the optical bench the MTF chart had some detectible astigmatism at 200mm. The field curvatures at 70mm, 120mm, and 200mm zoom lengths show why. There's a bit of field tilt at the wider focal lengths (about normal for most zooms) but at 200mm the tilt becomes more dramatic and the tangential field is much more tilted than the sagittal field.
About a year ago, Andrew from SLR Magic came and spent a couple of days in our lab, testing some of the SLR Magic lenses and getting input on improving the optical adjustments on their lenses. At the time Andrew made it clear their company wanted to up their quality and compete with the better lenses on the market. He contacted me a couple of weeks ago to let me know he was sending a copy of their new 50mm APO T2.1 Cine lens for optical bench testing.
During our conversation it became very obvious that Andrew was very comfortable this new lens was going to test very well. You know me, there's nothing I like better than using our optical testing equipment to crush the comfort right out of a manufacturer, so I was pretty eager to take a look at this lens.
While this isn't a review, I will say the SLR Magic lens was mechanically well made with smooth, steady resistance focusing and aperture rings and had a nice compact size. As an apochromatic lens, one of its major advantages is that longitudinal chromatic aberrations are neutralized, so out-of-focus highlights in both the foreground and background remain color-neutral. But this lens is also designed to be high resolution, capable of shooting excellent 6K images, so some MTF testing seemed in order.
Optical Bench Results
We will compare the SLR Magic 50mm T2.1 with a couple of similar lenses: the Zeiss 50mm T2.1 and the Leica 50mm f/2 APO Summicron-M. While I don't know what the final pricing on the SLR Magic 50mm will be, I assume it be similar to the Zeiss CP.2 (about $4,000) and certainly less than the Leica (over $8,000). Let me note that we tested a single copy of the lens, so I can't make any comments on sample variation.
As you can see in the MTF comparison below, the SLR Magic lens definitely is holding its own with the Zeiss 50mm. In the center of the image they are nearly identical at lower frequencies, while the SLR Magic lens is actually a bit better at higher frequencies. The Zeiss holds it's higher frequencies a bit better in the middle 1/3 of the image. The SLR Magic is superior in the absolute edges. (Remember, the 20mm distance from center we show would be at the very edge of 35mm full-frame edge image. If you are shooting a smaller sensor the outer areas of these MTF charts don't matter to you.)
The 50mm Leica Summicron-M f/2.0 is arguably the best 50mm lens we've tested, particularly at higher frequencies. When we compared the SLR Magic to the Leica, the results shocked me a little - not that the Leica was better, but that the SLR Magic lens was actually fairly close to it, except at the higher 40- and 50- line pair/mm frequencies. That's an outstanding performance.
As long as we had it on the machine, we did field curvature maps of the SLR Magic 50mm, too. They were very nice, with tangential and sagittal fields very similar, as you would expect from the MTF curves. The field curvature demonstrates that this lens is designed to give a nice flat field from edge-to-edge with just a tiny bit of 'mustache' curvature. This particular copy had just a very slight amount of tilt in the field (0.01 degree), which is excellent -- we don't think anyone could notice less than 1 degree in an image.
A Bit of Summary
This is not a lens review but since we are able to do these tests and most reviewers can't I thought them worth putting up. When we met with SLR Magic way back our primary input was on methods to minimize copy-to-copy variation. Testing one copy doesn't give us any insight into the progress they've made there (although I completely agree with the steps they've taken and am optimistic). However, the perfectly flat field with no tilt in this sample gives some evidence that things are looking good. With many lenses almost every single copy has some tilt.
Until final pricing is established there's no way for me to say if this lens is a bargain or not. But assuming the price is reasonable, it certainly provides excellent image quality from an MTF standpoint. Some early reviews indicate the goals of the APO lens in removing color aberrations have also been met. It looks to me like SLR Magic has upped their game quite significantly here.
The new Canon 11-24mm f/4 is their widest ever beast of a zoom. Roger has tested it and said it's as sharp as we had hoped, but I wanted to know how it looked in a real world comparison with the other beautifully bulbous ultra wide, the Nikon 14-24mm f/2.8.
Canon 11-24mm on a 5DmkIII & Nikon 14-24mm on a D750
I'll be honest. I'm pretty excited about the Canon 11-24mm f/4 L lens. I love shooting ultra-wide and the chance to shoot this wide with a rectilinear lens on a full-frame camera has me pretty excited. But I'm also very aware of how near-impossibly difficult designing a lens this wide would be, so my expectations were tempered a bit. There's a reason I'll often stitch together a couple of 24mm shots for a landscape rather than take one 16mm shot. Okay, there are several reasons, but image quality is high among them.
So I couldn't wait to get the new lens on the optical bench to see if it was even close to acceptable at the wide end. But there is a bit of a problem there. We've never had the opportunity to test anything at 11mm focal length before. So what do we compare it to? I decided we'd compare the 11-24 to itself. We'd get an idea of how well it did at the long end compared to other 24mm options, and then compare those results to the wide end. The wide end can't possibly be as good as the long end, of course, but we can see how close it is.
This is not a detailed lens review, of course, just a nice quick assessment of resolution with the new lens.
This Sucker is Heavy, BTW
Looking at a comparison of the wide f/4 zooms most of us are familiar with, the new Canon 11-24mm is a little shorter and wider than the Canon and Nikon 16-35 f/4 lenses.
Left to right: Nikon 16-35 f/4, Canon 11-24 f/4, Canon 16-35 f/4 Roger Cicala, Lensrentals.com, 2015
The new Canon lens actually most resembles the Nikon 14-24 f/2.8 lens, not just in size and shape but also in the very protruding front element.
Nikon 14-24 f/2.8 (left) and Canon 11-24 f/4. Roger Cicala, Lensrentals.com, 2015
What the pictures don't show is the weight. The new Canon weighs in at 2.6 pounds, which is twice the weight of the 16-35 f/4 IS lens, and half a pound heavier than the Nikon 14-24 f/2.8.
Our Just the Lenses posts are optical tests where we compare various lenses on the optical bench. Unlike DxO or Imatest test results, no cameras are involved, eliminating one of the major variables. It's particularly useful when we're looking at third-party lenses that can be used on various cameras. It's hard to extrapolate the results of a test made using a third-party lens on a Canon 5D Mk III when you are trying to determine how it might compare to one shot on a Nikon D800, for example. Testing on the optical bench gives a direct comparison between lenses without any other variables.
You know Aaron and I love doing teardowns of almost anything. The two types we look forward to most, though, are Sony cameras and Canon lenses, because those tend to be on the cutting edge of engineering elegance. For a couple of weeks now, we've been wanting to sink our screwdrivers into the new Canon 100-400 IS Mk II lens, and yesterday we finally got a few free hours to do it.
For those two or three of you who don't like a little lens strip-tease I'll give you the quick summary: the build quality on this thing is amazing. I usually laugh when people describe a lens as "built like a tank" because what I know is the lens they are describing has a thick, heavy outer metal shell filled with tiny delicate pieces that break and wear out with great frequency. But this lens is built like a tank inside and out. Continue reading →
When I posted our Imatest results from the single copy of the Nikon 300 f/4E PF ED VR that we had received, I pointed out that the copy was slightly decentered and therefore the results were questionable. We got a second copy of the lens in today, and this one was perfectly centered.
I had planned on waiting until we had a half-dozen samples tested before writing anything else, but apparently I stirred up such a hornet's nest with the first results that I thought I'd go ahead and post the results of this well-centered copy. Especially since it looks like it will be several weeks before we actually have six copies in stock to run a series.
My guess when testing the previous copy was that a well-centered copy might or might not be better at peak resolution, and probably would be better in the corners. Turns out, for those who want the short version, my guess was pretty much correct. Continue reading →
As most of you know, we generally test multiple copies of a lens when we evaluate it, simply because we are so aware of copy-to-copy variation. But I got caught between a rock and a hard place this week. We received exactly one copy of the new Nikon 300mm f/4E PF ED VR lens and the purchasing department told me it would be at least a week and probably longer before any more come in. But I've been dying to know how it measured up against the existing Nikon 300mm f/4 ED AF-S lens.
So I did what I had to do and ran the one copy through Imatest, comparing it to a batch of several copies of the older model. We used the new high-resolution, backlit film charts that we've been using for telephoto lens test lately, and shot all the tests on a Nikon D810 test camera so we would generate the highest possible resolution.
I always like to mention my expectations going into testing. The existing 300mm f/4 AF-S is a very, very good lens. The new PF ED VR brings a lot of new technology to the table, and we know from experience with Canon that Phase Fresnel (what Canon calls Diffraction Optics) technology has historically not give the resolution that a good optical lens does. On the other hand, Canon's second generation DO lenses seem to be every bit as good as their standard optical telephotos. So I hoped the new lens would be optically as good as the old one, but wouldn't be shocked if it were a bit better or a bit worse.
Even if it wasn't quite as good optically, the PF ED VR has a lot of attractive features: an excellent VR system, a tiny size, better coatings, and an all electric aperture system with no mechanical linkages. (Most of you won't consider the latter a great advantage, but most of you don't work on lenses for a living.)
For emphasis: this is somewhat of a failed test. The results we got for the 300mm f/4E PF ED VR lens are good, but other copies may well be better. I'll speculate on how much better, but until we get more copies, that's all it is, speculation.
A Quick Look at the Lenses
The side view shows at a glance just how much smaller the new 300mm lens is. It can't show you the weight difference, but the new lens, at 1.66 pounds, weighs about half as much as the original version.
Nikon 300mm f/4 AF-S (left) and 300mm f/4E PF (right) with hoods. Image Roger Cicala, Lensrentals.com, 2015
The view from the mount shows that, like the other "E" lenses, there is no mechanical aperture lever. The aperture in this lens is completely electronic.
Image Roger Cicala, Lensrentals.com, 2015
Below is a simple table showing the Imatest results as we usually present them: Center point, Weighted average of 50 points across the front of the lens, and average of the 8 corner results (4 horizontal, 4 vertical) given in Line Pairs / Image Height from unsharpened raw images.
300mm f/4 AF-S
300mm f/4E PF VR
The results seem somewhat straightforward at first glance. The new lens is a bit sharper in the center, not quite as sharp in the corners, but the two lenses are very close in resolution. Certainly with only one copy of the new lens tested I wouldn't want to say more than 'equivalent in resolution' to the original version. It's advantages in size and VR speak for themselves. I'd want one if I shot at this focal length a lot.
But . . .
One of many Imatest shots taken with the 300mm f/4E PF VR lens
There's more to it (or at least there should be more to it) than just letting the computer spit out numbers. When we look at the actual graphs of the Imatest results, the 300mm f/4E PF VR graphs all show the sharpest point not near the center of the lens, but off to the right side of the image. It looks like the lens is slightly decentered or tilted. This isn't bad, by any measurement, but it's noticeable.
Our Imatest set-up self-checks to be certain we are correctly aligned, but just to be sure we tore it down, set it back up, and got similar results. We then put the lens on OLAF, our 5-micron-pinhole-collimator testing machine, which confirmed the lens was just slightly decentered.
OLAF image of the 300mm f/4E PF VR at center point.
If you look carefully it's apparent that the rings for each dot are not perfectly symmetrical; they're slightly shifted to the left. Again, this isn't a horrible result by any means. A really bad lens would be smeared 300% or more times on one side compared to the other, this is not bad at all.
Still, it's apparent this lens is slightly decentered.
What Does That Mean for the Results?
I'm going to speculate here, because we don't have experience with decentered Fresnel lenses. When we get another half-dozen copies we may find they are all just like this - that a slight degree of decentering doesn't affect the lens at all. (There are several lenses like that, where every copy is slightly decentered and the resolution is just fine.)
We may find that a well-centered copy is far superior to this one, particularly in the corners, but perhaps even in the center.
What we can say is that the new lens is at least the equal, from a resolution standpoint, to the old lens. That's about all for now.
When I get a half-dozen copies in and tested, I'll be able to say a lot more. What I suspect I will find is the center resolution after testing multiple copies will be about what it is here -- just slightly better than the original 300mm f/4 AF-S. I think that because our maximum resolution point, while off-center, was not up near the edge of the image. My experience is that the center resolution won't change much on a well-centered copy, it will just return toward the center of the image.
I expect we will see somewhat better numbers in the corners, and therefore in the overall average of the lens, probably making it slightly better than the original lens in all areas. Not greatly better, possibly not enough that you'd even notice it in a photograph. But at least as good, in a much lighter lens with excellent VR. That will definitely make this a worthwhile purchase for a lot of people.
The real bottom line? Testing one lens creates about as many questions as it does answers, at least for an OCD guy like me.
A while back we did a teardown of the Sony A7r, and were very impressed with its clean modular design. Now that Sony has released the A7 II we thought a similar teardown would be worthwhile. There are certainly going to be some differences. First among them, of course, is that the new camera has 5-axis, in-body image stabilization, which is definitely going to create some differences. There also is a more robust magnesium alloy shell and lens mount, and perhaps (or perhaps not) improved weather resistance.
In an earlier post we were most impressed with how much better the new Canon 400mm f/4 IS DO II resolved compared to the original version. I mentioned that we hadn't been able to get many copies and didn't have time to do any other comparisons right then, but that we would do more as soon as we could.
We still haven't received any of the other couple of dozen 400mm DO IS II lenses we've ordered, so the results are for the same pair as the original post. For all other lenses in this post we tested four copies and averaged the results. For the 300mm f/2.8 with 1.4X TC III test we used four lenses and four different converters. All tests are done using the same backlit film chart as the previous test, using a Canon 5D Mk III test camera.
400mm DO II vs. 300mm f/2.8 IS II
The first comparison we made was between the 400mm f/4 DO IS II and the 300mm f/2.8 IS II at their native focal lengths - 400mm and 300mm. I do want to point out that this puts testing distances at roughly 19 and 15 feet respectively. This is not ideal working distance for super telephoto lenses, so take these results with that tiny grain of salt.
We tested the 300mm lens at both f/2.8 and at f/4 to level the playing field a bit.
400mm f/4 DO II
300mm f/28 IS II f/2.8
300mm f/2.8 IS II f/4
These results are about what I expected, since we already knew that the 400 DO II is really excellent. Shot at its native f/4 it has a bit better resolution than the 300 f/2.8 does shot at f/2.8. Stop the 300mm lens down to f/4, though, and it's a bit sharper than the DO.
Could you notice these differences in a photograph? Probably if technique was equal and you pixel peep a bit. But these are all spectacular results. If you can't tell the difference between a 300mm f/2.8 shot at f/2.8 and at f/4 (most of us can comparing side-by-side shots), then you sure can't tell the difference between either one and the DO.
The question most people (and by most people, I mean me and a couple of others) wanted answered was how the 400 DO compares to the 300 f/2.8 with a 1.4 X teleconverter added. Other people wanted to know how it compares with the tried-and-true, bargain-priced Canon 400mm f/5.6 L, so we did those comparisons too.
Let's again point out that there are some differences in these tests. The 300mm f/2.8 with teleconverter is actually shooting at 420mm, so it gives a bit more reach. The 400mm f/5.6 is being compared one stop down compared to the other two, which are being shot at f/4, which gives it a bit of an optical advantage.
400mm f/4 DO II
300mm f/2.8 with 1.4X
400mm f/5.6 L
The results, again, are fairly triumphant for the DO. The DO version I, which I shot with frequently, definitely gave up some image quality compared to a 300 f/2.8 with teleconverter. Most of us who shot the DO were willing to do so because it weighed less, and the weight was distributed near to the mount making it easier to handhold. The version II 300 f/2.8 is much lighter than its predecessor, so the weight savings isn't quite as significant. However, it's clear that from a resolution standpoint at least, the 400 DO is slightly better than the 300 f/2.8 IS II with teleconverter.
Again, let's remember that with DO lenses, field tests in varied lighting conditions may well reveal other issues that a simple resolution test can't. But the resolution test certainly suggests that shooting the DO could give results at least as good, and perhaps a tiny bit better, than the 300 f/2.8 IS II with teleconverter.
The results with the 400mm f/5.6 weren't surprising to me at all. It has always been regarded as a very sharp lens. Given that it's shooting at a 1-stop narrower aperture, no one should be shocked that it can hang with the other two lenses in the center. The price is certainly far more attractive, although the narrower aperture and lack of image stabilization make it a very different lens than the other two.
Would I sell my 300 f/2.8 IS II and move to the 400 DO? I doubt it. But if I was considering the two for purchase and realized I was going to shoot at 400mm, I'd be leaning towards the DO. It's still easier to shoot handheld, the IS systems are now equal, and it doesn't appear to give up anything from a resolution standpoint.
Can I get equivalent shots with a 400 f/5.6? If the light and technique are good, I can certainly get very close, if not equal shots. But the other two lenses will get shots in a lot of conditions that the 400 f/5.6 could not shoot well in.