Half inch lenses compared

[Luc De Wandel]

This might interest those with a Sony PMW-320: I made some comparison test shots for sharpness, with the three 1/2-inch Canon lenses that I have for this camera: YH19x6.7KRS (SD), YH16x7 KRS (SD), and KH10ex3.6 irse (HD).

Shots were made with aperture 4; focusing on a 42 inch monitor; zoom ratio different of course, because of the different focal lenghts; identical camera settings; screen grab from XDCAM browser; desaturated in Photoshop to concentrate on sharpness alone.

You'll notice that there is very little difference. In fact the minor difference in sharpness between the 3 lenses is overruled completely by even the smallest mistake in focusing. I focused on a 42 inch screen to get the ultimate precision, but when focusing with the 320’s viewfinder, it is virtually impossible to reach the precision that is needed to bring out the difference between these lenses. Of course, with scientific equipment that can actually measure sharpness, the differences will be more apparent (I presume)...

[Alister Chapman]

An interesting test, but a much more interesting test is to do something similar but with much lower contrast between the black and white lines, perhaps between the same patterns but instead of black and white, slightly different shades of grey. This then starts to show the real world lens performance, in particular contrast performance as it is contrast that makes an image look sharp more than resolution alone. The ability to resolve small contrast changes, for example between blades of grass on a lawn or the texture of a surface is much harder to achieve than black and white line resolving.

You should investigate doing your own MTF tests using Imatest. There's a free 30 day trial version and it's very easy to use, but above all else repeatable and the results are conisitent. Imatest - Imatest Studio

[Luc De Wandel]

I wanted to use Imatest in the past, but discovered that it works only on a Windows platform. I hoped it would be released for Mac too, but to no avail... And I don't have 'virtual Windows' on my Macs.

I'll certainly do another test as you suggest, Alister, as I agree 100% that 'real life' sharpness is more than just resolution. Anyway, I found it significant that the most critical factor in this test was the purely physical limitation of focusing precisely enough to see the differences between the lenses. A movement of 1 mm of the focusing ring - wile checking sharpness on a 42 inch monitor - would have led to completely different results. This leads me to the conclusion that (HD-)lenses for high-defintion use should have a different gear ratio for the focusing ring...

[Alister Chapman]

If you move up to a camera with a super35mm sensor the focus issue becomes even worse as your DoF and CoC decreases. That's why decent PL mount lenses have focus scales that can cover 270 degrees of the lens barrel and the lens barrels are generally quite large, so the overall focus throw is increased. This is often the biggest failing of using DSLR lenses on these cameras.

Generally speaking one of the things your paying for with more expensive lenses is small things like better focus scales, less breathing, reduced CA and better contrast as opposed to just better resolution.

[Luc De Wandel]

Some of the DSLR Canon lenses are quite good for this aspect, as a matter of fact. I used to have the Canon 85mm f 1:4 and that just kept focusing for ever, I think it did 3 turns or something. But is was a fantastic lens, that I should never have sold... But for my use, at the time, (concert photography) it was utterly useless. Now it would have been a gem for my 5D mkII video shooting.

Canon 85mm f/1.2 L II

[Luc De Wandel]

DV-member Jim has sent me a screen grab of a comparable chart, shot with the PMW-320 kit lens in comparable circumstances. Looks very sharp, too...

[Jim Forrest]

Afraid the chart I used is not very good so the contrast levels don't look right.