Corner Bokeh Quality of Canon 35mm f/2.0 IS vs. Canon 35mm f/1.4 ii

Some eight months after buying the Canon 35mm f/2.0 IS lens, I’ve sold it to upgrade to the new Canon 35mm f/1.4L ii.  In many ways, it’s with some regret.  The f/2.0 IS served well on the whole.  I’d bought it primarily to shoot submissions for my Getty Images stock library; a nicely portable lens that could shoot a variety of subjects, in low light when needed, with the ability to blur away the more distracting elements of an unwanted background.  Over the eight months I used it, shots like these made their way into my Getty gallery.

Canon 35mm f/2.0 IS - Smoke Without Fire
Canon 35mm f/2.0 IS – Smoke Without Fire
Canon 35mm f/2.0 IS - Hazard
Canon 35mm f/2.0 IS – Hazard
Canon 35mm f/2.0 IS - Rain Without Sky
Canon 35mm f/2.0 IS – Rain Without Sky
Canon 35mm f/2.0 IS - Hard Light & Water
Canon 35mm f/2.0 IS – Hard Light & Water

For all the successful shots, however, the Canon 35mm f/2.0 IS has a significant weakness; the quality of the corner bokeh.  Especially exposed when shooting close up with a wide aperture for a shallow depth of field, the corners can take on a weird geometric pattern.  Giving images the sense that they’d been poorly photoshopped, and incredibly hard to remove, the effect was something I could not live with.  Here are some examples;

Canon 35mm f/2.0 IS - Strobist cow parsley
Canon 35mm f/2.0 IS – Strobist cow parsley. Geometric pattern widespread in top corners.
Canon 35mm f/2.0 IS - Girl picking blueberries
Canon 35mm f/2.0 IS – Girl picking blueberries. Geometric pattern evident in top corners especially.
Canon 35mm f/2.0 IS - Girl picking blueberries
Canon 35mm f/2.0 IS – Girl picking blueberries. Geometric pattern widespread.
Canon 35mm f/2.0 IS - Girl at Kimmeridge
Canon 35mm f/2.0 IS – Girl at Kimmeridge. Geometric pattern evident on cliff’s behind girl’s head, and through waves/rocks on left.
Canon 35mm f/2.0 IS - Girl running in Paris fountains
Canon 35mm f/2.0 IS – Girl running in Paris fountains. Geometric pattern evident in buildings, especially far left.

Whilst more or less apparent from image to image, in some cases the strength of the pattern was such that the image was essentially unusable for commercial purposes.  When the new Canon 35mm f/1.4L ii was launched then, my first question was whether this issue would be fixed.  Most articles compare the old and new f/1.4, or compare the new Canon f/1.4L with the Sigma 35mm f/1.4 ART lens.  For those f/2.0 IS users experiencing similar issues as those I faced, I hope the following images from the new 1.4 ii help inform their thoughts on a potential upgrade.

The short story is that I have yet to see this issue with the new L lens.  Here are some examples, shot as tests rather than portfolio pieces.  Many were shot on the evening the lens was first delivered, at night and therefore ISO 1600.  As such I would not use these as examples for sharpness necessarily.

Canon 35mm f/1.4 L ii - Test shot
Canon 35mm f/1.4 L ii – Test shot
Canon 35mm f/1.4 L ii - Test shot
Canon 35mm f/1.4 L ii – Test shot
Canon 35mm f/1.4 L ii - Test shot
Canon 35mm f/1.4 L ii – Test shot
Canon 35mm f/1.4 L ii - Test shot
Canon 35mm f/1.4 L ii – Test shot
Canon 35mm f/1.4 L ii - Test shot
Canon 35mm f/1.4 L ii – Test shot

More to follow as I continue to shoot with the new lens.

Test Until Destruct – How do repeated jpg saves degrade image quality

Though I shoot exclusively in RAW and edit all within the same Lightroom/Photoshop platform, at some point I end up saving my files to jpg.  Whilst I’ve always understood that that first save to jpg inflicts some pixel damage to the image, I hadn’t fully appreciated that the loss of quality stacks with each subsequent save to jpg.  A little while back, having been told about the effect, I thought I’d run a test – repeatedly saving, reloading and resaving a series of my images to jpg to see what happened.

The white figures indicate the number of repeated jpg saves.

Original image Meltwater, Hertforshire.
Original image Many Glacier, Montana.
Original image Greenhouse Effect.

At this scale the artifacts weren’t obvious for the first few saves. Seen at 100% or greater areas of the sky had become unacceptably degraded by as few as five saves, exhibiting the kind of level of damage seen at 10 saves in examples of this scale.  The damage is a result of the jpg’s compressed nature. Keeping files in RAW, PSD or other non-compressed file formats will preserve quality and prevent this kind of degradation.