As a Nikon photographer, I’ve always been rather envious of what Canon can offer in its lens department, particularly the legendary 17mm Canon tilt-shift lens. I decided to bite the bullet last year and procured a pristine second-hand copy to pair with the Sony a7r, hoping that the combination of resolution and dynamic range similar to my Nikon D800E would live up to this demanding lens that no other Canon body could (except the Canon 5DSr, with the exception of dynamic range). A few weeks after I’ve dropped a sizeable investment of around S$3.5k for my architecture kit (Sony a7r, Canon 17mm TSE and Canon to Sony adapter), Nikon announced the 19mm PCE lens. With performance touted to be close to, or exceeding, its Canon counterpart, I thought I might regret my purchase decision until the announced the price. The Nikon 19mm PCE is priced at close to S$5k, significantly higher than the Canon, and eclipsing the entire kit that I’ve purchased just for a specialty lens. Any differences then, I figured, have to really justify that premium in order to convince me from getting one over the other.
As such, I’ve hoped for someone to conduct a simple, real-world review of both lenses with similar resolution bodies in order to determine the differences between the two, no matter how minute. However, as I scoured the Internet for such a review, finding even reviews on the Nikon lens proved to be an unsatisfactory affair. It seemed that this lens was priced out of reach for most people. Requesting a review unit from Nikon on the pretext of a comparison review with its Canon counterpart was met with silence, and so I resigned to hoping that someone might one day take the time to conduct a comparison between the two lenses.
As it turns out, a friend of mine happen to get his hands on the Nikon 19mm, and so I arranged for us to meet to test out both lenses. I’ll try to be as impartial as possible in this comparison review, but it will be conducted in real-life, outdoor conditions as I’m more of an environmental photographer as opposed to a studio photographer.
Both lenses are made of solid metal, with a hard plastic outer shell. Both have protruding bulbous front elements, and are manual-focus only. Knobs on both lenses are large and easy to twist. The Nikon’s knobs are a lot more damped as compared to the Canon’s, but that could be due to mine being quite a well-used lens, and that goes for its focusing ring as well, something I appreciate over the Canon’s somewhat loose focus ring which is prone to being accidentally spun when nudged. Other than that, build quality on both lenses is reassuring and impeccable for a modern-day lens. Both the Canon and Nikon tilt-shifts can shift to a maximum of 12mm in all directions and rotate in all directions. When paired with a mirrorless Sony body, the Canon lens have unhindered breathing space for freedom of movement and rotation, in contrast to the Nikon, where the camera can get in the way of the lens during rotation. As such, one will have to shift the lens back to its zero position before rotating it off-axis. However, I do not have a Canon body to test if the body will impede movement on the Canon tilt-shift. Still, the Nikon will allow for full movements to be used, although it can be a bit fiddly to achieve that.
Now for what I’m most interested in - the extent of shifting for both lenses which is crucial for shift panoramas. Shift panoramas are panoramas done on tilt-shift lenses where one shifts the lens in both directions to create a stitched image. I generally employ shifting and rotation techniques in order to create a field-of-view wider than the lens’ focal length by making use of the large image circle tilt-shift lenses provide. Since both lenses can shift a maximum of 12mm, but the Canon is wider than the Nikon, I would expect similar field-of-views for both when stitched, with the Canon being slightly wider. The results did hold true for that assumption:
The Canon 17mm TSE is slightly wider than the Nikon 19mm PCE when stitched, though not by much. What I am surprised is the rather strong barrel distortion I’m seeing from the Nikon, which is obvious from the composite. Compared to the Canon composite, the Nikon one shows some distortion in the Marina Bay Financial Centre buildings on the right.
Both lenses seem very sharp at the centre - hard to find any differences there except for the software stitching error in the Nikon image. It’s remarkable that the Canon, being designed in 2009, still has plenty to offer on a high-resolving body in 2017. These pictures are crops from the above composites. Both were shot at the same settings of f/16, 1/5s shutter speed. The Nikon one was taken slightly later, with more lights in the scene.
Both lenses also seem to be sharp towards their corners, which is a sign of remarkable optical engineering. The Canon seem to have slightly worse chromatic aberrations, but this is easily cleaned up in post. Notice how the Canon has more distinct 8-point starbursts from point light sources due to its 8-bladed diaphragm. The Nikon, being a 9-bladed design, renders 18-point starbursts, which are not as distinct. This is purely personal preference though.
Well, both lenses perform spectacularly well, being sharp from edge-to-edge and rendering contrasty scenes. The Canon, in particular, was impressive that despite its age still manages to beat the Nikon in terms of distortion control. Both have similar resolving power, and it seems that both can handle way more than the current highest resolving sensors can throw at them, which is a good sign that they will continue to stay relevant in years to come.
If I get the chance to conduct another comparison, I’ll try to compare their flaring properties, which should be interesting considering their bulbous designs. The Canon tends to fare poorly in this avenue, so I hope the Nikon can pick up where the Canon stops short. Still, considering the price differences and that the Canon can perform just as well as the Nikon for a significantly lower cost even accounting for the body, I’ll have to give the Canon the edge in the price-to-performance ratio. That, and the fact that it controls distortion extremely well.