Comparing Canon's two 50mm EF lenses

 

Having been surprised at how well Canon's cheapest EF lens, the 50mm f1.8 II, performed in recent astrophotos, I decided to get the more expensive 50mm f/1.4 USM. I thought it would give me an extra stop in aperture. Unfortunately it does not because the distortion in the corners of the frame is really quite bad at f/1.4. It is just as bad at f/1.8. My first successful result with it had to be at f/2.8, two stops down from the widest aperture. On the other hand, stars at the centre of the frame are sharper with the f/1.4 lens than the cheaper f/1.8 one.

Stars are a severe test of any lens because we are expecting points of light to be imaged as points. In reality star images always have a small radius due to atmospheric turbulence but we expect them to be truly circular discs. Neither of these lenses can do that in the corners of a full frame image.

On this page I show corners and centres of single exposures of star fields. These are at high ISO settings, unstacked, and so are noisy. Ignore that, what matters is the shapes of the stars.

 Corners of full frames

EF 50mm f/1.4 USM lens at f/1.4

EF 50mm f/1.4 USM lens at f/1.8

EF 50mm f/1.4 USM lens at f/2.8

EF 50mm f/1.8 II lens at f/1.8
(fixed tripod => slight trails)

 Centres of full frames

EF 50mm f/1.4 USM lens at f/1.4

EF 50mm f/1.4 USM lens at f/1.8

EF 50mm f/1.4 USM lens at f/2.8

EF 50mm f/1.8 II lens at f/1.8
(fixed tripod => slight trails)

The photos with the f/1.4 lens were taken on an HEQ5 mount but the f/1.8 version was on a fixed tripod. I intend to get a better series of comparisons when I can.