Canon 200-400 f/4 reviewed by "Chasseur d'Images"

Started Jun 20, 2013 | Discussions thread
mattr Veteran Member • Posts: 3,523
Re: Canon 200-400 f/4 reviewed by "Chasseur d'Images"

Oldvic wrote:

mattr wrote:

Oldvic wrote:

July issue, pages 106 to 109, field and lab test. CI tests for an A3 print size.

On the 5DIII, without internal converter (200-400):

- vigneting: from 0.7-1 stop at f/4 to near zero at f/11;

- CA: near zero at all focal lengths and apertures;

- distortion:same as CA;

- sharpness: about 92-93% from f/4 to f/11 and ending around 88% at f/22. The error bars, which indicate consistency, are barely detectable, and there's virtually no difference between the center and the corners of the image.

5DIII, with internal converter (280-560):

- vigneting: same as above but without the f/4 to f/5.6 interval, obviously;

- CA: slight increase, the fringes stay below 0.03mm from f/4 to f/16 and end around 0.05mm at f/32;

- distortion: around 0.4% pincushion at all focal lengths;

- sharpness: same as above, with the new f/32 aperture introduced by the converter hovering around 80% from 280 to 460 and dropping to 65% at 560 (f/32 aperture, remember). Again only at 560, the error bars become a little bigger but not by much.

They tested for sharpness at 560mm using an external 1.4x converter and found no significant difference for most apertures, f/32 being the exception: the external converter reaches a sharpness level of 80% as opposed to the 65% of the internal one.

Also tested the internal-1.4x external converter combo (800mm focal length), with a sharpness of around 87-88% at f/16 and a minimum of about 87% at f/32.

Next they tested the lens with an external 2x converter (800mm) and the results are as above with the 1.4x-1.4x combo.

Finally, they tested the internal 1.4x-external 2x combo (1120mm): sharpness drops to around 40-45% at f/11, around 60-70% at f/16, and tops out at 70% for f/22 and f/32.

Thanks for the info but it is very hard to take this test seriously.

The focus on differences in the f/16-f/32 range is utterly crazy. It also would be interesting to know what the % numbers are.

The credibility of the test is CI's credibility. The differences were reported where they exist, and the percentages are my best visual interpretation of the graphs, which are rather small.

The problem is that resolution numbers at f/16-f/32 tell us very little about the optical characteristics of the lens. This is too far in diffraction territory, even with the relatively large 5D3 pixels. Differences will be caused mainly by errors or experimental flukes that affect the effective f-ratio, such as caused by focus distance. Of course, these results are also of pretty much zero practical relevance, IMO.

There is a reason why lens test sites that know what they are doing (e.g. Photozone) report MTF results usually only up to about f/11.

I'd recommend taking a look at the mag, as there's a lot more to the test than what I've reported here (test on a 700D, field use notes, etc.).

In order to test a very sharp lens (such as the 200-400/4) and to establish effects of TCs one should use the camera with the smallest pixels possible. So the test results from the 700D should be more informative. The 5D3 would be necessary only to analyze full frame edges.

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow