When I heard recently that Canon was upgrading some of their tilt shift-lenses I got very excited. As you will soon learn, I am a bit of a nut for tilt-shift lenses and I wouldn’t want work without them. My 24mm and 17mm get almost constant use for my architectural work, and the 90mm is my lens of choice for food and product. I love the control they give me, and while they lack the convenience of autofocus these lenses cause me to slow down and take my time and hopefully take better pictures . Adding to the list of positives is that fact that some of the recent Canon lens upgrades have been nothing short of miraculous in terms of optical improvement (I’m looking at you 24-70mm 2.8l II).
The chart I used, while arguably not the fanciest was helpful for my simple sharpness test.
The three updates are a 50mm (to replace the 45mm), a 90mm (which updates a classic to a macro L version) and a 130mm. Looking at their charts the optical performance of all three lenses should be nothing short of amazing. I also do a good bit of pixel peeping while researching lenses at “the digital picture” which confirmed my suspicions that the 90mm was super sharp. Similar to my eyes to much more expensive boutique lenses like the Zeiss Otus. As I looked more into the 90mm macro I started to daydream about how I might use all of this resolving power and before you know it I had decided to unload some lesser-used gear and get myself one of these beautiful new tools.
Top right corner zoomed in to over 100%, darned hard to see the difference. f11 at 1/125
My first impression of the new 90mm ts-e was very favorable, the lens is much bigger and heavier than the classic with a more solid feel. When compared to the original 90mm ts-e it seems almost twice the size and weight. I don’t much mind the weight because I generally use this type of lens on a tripod. The fit and finish of the new lens is more in keeping with it’s L series counterparts. The focusing is stiff but smooth and reassuring, the feel is quite different than the effortless focusing on the older version. On looks alone this new lens is a knockout.
Here is the view after shifting the lens, I chose to inspect the top right hand corner to see the performance shifted to the outer limits of the lens.
I was already a fan of the original 90mm ts-e, so the opportunity to upgrade to a L lens with better durability, sharper optics and weather sealing was exciting to say the least. Though due to the design quirks of a tilt-shift lens, the weather proofing is not as effective as on a typical L lens. My initial feeling about the lens once I had time to do a little shooting, was that it was the best lens I had ever worked with. But I have learned over the years that it is easy to jump to conclusions that prove to be incorrect. So I decided to test my perception by a simple head-to-head comparison.
Here are the results at a little more than 100%.
I decided to start with a lens calibration chart so I could judge sharpness from the center to the edges. I shot with the camera tethered to a iPad so I could zoom in and get perfect focus. Tethering allowed for triggering without touching the camera and introducing any additional motion or blur. I used strobes for lighting that also have the benefit of further reducing motion blur because of how quickly they make the exposure. Finally, I used mirror lock-up to insure that the motion of the mirror wouldn’t effect the results. I know that several of these steps are redundant when working with strobes but I wanted to make sure that I got as clean a shot as possible.
Here is some nice crisp text for the next part of my test.
Once all of the pieces were in place, I made some images of the chart. The aperture was set at f11 which should provide the best sharpness results for both lenses. I shot one image with each lens with the shift centered and another shifted fully in order to test edge to edge sharpness. I pulled up the results in camera raw and was surprised by the results. Rather than the huge jump in quality I had anticipated, I could barely see a difference. The new lens had the edge, but just barely. It was so small a edge that I could barely see a distinction when zoomed in at 100%. Just to be sure, I took the files from both lenses and layered and aligned them in photoshop so I could go quickly back and forth between the files and the result was the same. There was an almost imperceivable difference, with the very slight edge going to the new 90mm ts-e macro.
Here is the jar portion of the image. the older legacy lens on the left and the new macro on the right.
Surprised by my results, I decided to expand my test. I found some sharp contrasty lettering on a jar in my studio, I focused and shot with both lenses, stacked and aligned the images and sure enough I got the same outcome. Both lenses were amazingly crisp, but with such similar results, is the upgrade worth it? At this point I had my doubts.
The closest focusing distance of the new macro, the detail is gorgeous, especially when you zoom in to see all of the detail.
Macro capability is a feature of the new 90mm to-e l, so I decided to test both lenses at their closest possible focusing distance. Here, the new lens comes out on top. While the lens isn’t technically a 1:1 and therefore not a true macro, it doesn’t disappoint when it comes to close ups. In using the older 90mm ts-e I would commonly hit the close-focusing limit and resort to using an extension tube to approach macro magnification. The newer lens seems to have a nice range, without resorting to extension tubes.
Here is the same subject photographed at minimum focusing distance for the older 90mm ts-e.
While there is no appreciable difference in sharpness at f11, at f2.8 the two lenses are quite a bit different, the new 90mm is a lot sharper. As the aperture goes up, the difference between the lenses shrinks to become almost negligible. The important consideration is determining how this type of lens is most likely to be used. For product, I am always shooting at an aperture of f7.1 or higher, at which point there is very little distinction between the lenses. Both lenses do exceedingly good jobs with the precise technical demands of product photography but the newer lens adds the ability to shoot (near) macro. For photographing food, depending on the style it might be more common to shoot at lower apertures where the extra sharpness of the new lens might be beneficial, however for the more editorial style of shooting food the original lens is plenty sharp and makes files big enough for almost any use. It really seems like the only place in food photography where upgrading to the new lens might be necessary is to capture highly technical commercial work for packaging or advertising.
I absolutely love both lenses, but the macro is a welcome addition.
While the new 90mm ts-e macro is a engineering marvel, testing both lenses side-by-side really convinced me how great a deal the legacy 90mm lens is. Some older lenses haven’t kept up with the resolution needs of the higher megapixel cameras of the last few years, but that isn’t true of the original 90mm to-e. The excitement around Canon’s new line-up has drawn attention away from a real gem. Used versions are selling on EBay at sub $600., even though the image quality easily rivals its far more expensive successor. All this is not to say that the new 90mm ts-e is not worth the money but that upgrading is far from the obvious choice that it seemed.