In a perfect world, a lens would, when aimed precisely at a target with square lines, reproduce the target with perfect square alignment. Due to the challenges of manufacturing optics and focusing light, this is almost never achievable, without a lens specially designed for flat-field focus, as seen in repro and copy lenses. In the case of these two lenses, we’re seeing significantly different performance in this area.

Canon EF 85mm f/1.2L II USM @ f/1.2

The 85mm f1.2 shot at f1.2 has a fairly pronounced “pincushion” effect, shown above by a local distortion and curvature curve that drops steeply as we move away from the center of the lens. Pincushion distortion is the effect that makes the image appear pinched in from the sides, and top/bottom of the frame. Shot at f1.8, we get the same effect, but to a slightly lesser degree.

Canon EF 85mm f/1.2L II USM @ f/1.8

The 85mm f1.8 has just the opposite distortion: a pronounced barrel distortion effect. Here’s the graph, showing the rise, or increase in outward curvature towards the centers of the sides and top/bottom.

Canon EF 85mm f/1.8 USM @ f/1.8

Barrel distortion is a common problem with most wide-angle lenses, but not so common in lenses of normal, short-tele or full telephoto range. If it was severe in a short telephoto it could cause some problems with distortion in a portrait or face proportions. However, in general use barrel distortion in a short telephoto won’t cause a significant problem- nowhere near the importance of correcting barrel distortion in a wide angle lens used for architectural interiors, for example.

That said, it’s interesting that these two lenses perform so differently in this area. The f1.2 is clearly the choice here, however, unless the lens is being used for architecture or reproduction work, it’s probably not particularly significant.