Posts Tagged ‘barrel distortion’

Canary Wharf

Friday, October 4th, 2019

I was going to the opening of an exhibition at the Museum of London Docklands that evening and since it was a nice afternoon decided to leave early and take a walk around Canary Wharf on my way there. And again when I left I took a few pictures on my way to the station.

It’s quite a while since I’ve seriously gone there to take photographs, though occasionally I’ve walked through parts of it on my way elsewhere, taking the Jubilee Line from Waterloo and walking to nearby places or changing to the DLR.

I hadn’t seriously intended to take photographs, but I did have a camera with me, though only the Olympus OMD E5II with a tiny 17mm f2.8 pancake lens – a 34mm equivalent. I’d recently got this secondhand (it seemed so cheap I couldn’t pass it by) and was interested to see how well it performed. It makes the camera almost pocketable.

I read a review which promised high levels of chromatic aberration and barrel distortion, and thought that these would probably not be important for most of my photography. But trying them out at Canary Wharf they might be a problem.

Of course they were not, as you can see from some of these pictures (and the others on My London Diary in the post Canary Wharf. Both are lens faults which software can handle well, and it is hard to find much evidence of either in the results. If I open the RAW files in Photoshop, clicking the ‘Lens Correction’ icon in the Camera Raw dialogue tells me “This raw file contains a built-in lens profile for correcting distortion and chromatic aberration. The profile has already been applied automatically to this image”.

But then I looked at the RAW files, and they seemed fine too, with no hint of barrel distortion and the same small amount of colour fringing, not normally noticeable. I think my viewer software will actually be looking not a the raw data but at a jpeg incorporated into the RAW file, which I think means that any correction needed has been done in the camera. A few clicks on the web and I find that Micro-4/3 images are auto-corrected according to the stored lens profile (lens firmware) and a review on Optical Limits which shows the effect. As they say, barrel distortion “is only rarely noticeable in field conditions.”

The reviews also say the lens isn’t that sharp, at least when viewing large prints, but even examining these pictures at 1:1 on screen the sharpness in both centre and corners seems acceptable. There may be some uses for which it isn’t sharp enough, but I think it’s good enough for most of us. If anything it is the actual file size of the camera (4608×3456 pixels with the EM5II) that will be a problem for really large prints, not the lens.

Canary Wharf