7D fubar?

And after getting excited about the 7D the other shoe has now formally dropped. It's pretty clear that much of that extra pixellage is wasted with a strong aliasing filter. This is well-illustrated by looking at the DIII/DIV/7D/5D2 comparison down the page at:

http://www.the-digital-picture.com/Reviews/Canon-EOS-1D-Mark-IV-Digital-SLR-Camera-Review.aspx

Which shows that the 7D is far noisier, and quite a bit softer than the 5D Mark II (unless you ramp up the sharpening, presumably). The 1DIV looks like it might be a little noisier than the 5D2, and perhaps not quite as much raw sharpness, but the results are close.

The results are in line with Darwin Wigget's review of the 7D so there's no reason to think they are erroneous.

Since the tests are done at the same focal length, the sharpness of the characters is a good match for the intrinsic sharpness feather detail would be in the final image. The 7D falls short by a decent margin, especially over the 5D2. The only upside to this is that the high cost of the 1D IV looks a little better if I don't buy the 7D as well - there's no point me getting a camera that has substantially worse image performance than my current one.

I'm also interested in what/if Canon will announce for the 1Ds IV. Although $8K-ish is a lot of money for an SLR, compared to $5K-ish for an SLR that isn't as sharp as the one I even own (1D4) it would look rather more compelling if it did resolve more than my 5D2.

Northern Bobwhite

Northern Bobwhite, adult male, TX in April

I was on a ranch in Raymondsville TX looking for Ferruginous Pygmy-Owl when a small covey of Bobwhites dropped into the feeder. In NJ (where I've seen precisely one) the birds are skittish but they were obviously not feeling threatened and used to the human activity in this area. They weren't totally tame, and sped up as they walked past me. I ended up lying on the grass shooting the quail hand-held with the 1D Mark III as they wandered past me or fed on grain. This is always a little more exciting in a region that has fire ants, but I got away with it this time.

Short-billed Dowitcher

Juvenile, NYC in September

This is a classic juvenile Short-billed Dowitcher shot from a low angle (lying in the mud) at Jamaica Bay WR toward sunset. This was shot with the digital Rebel XTi, which doesn't always produce exceptional images but does well on well-lit subjects at low ISO.

Wilson's Phalarope

Tests on 5D Mark II settings

Prompted by discussions with a friend concerning his new 7D I did the following tests on my 5D Mark II :

1. Default Picture Code settings do not affect RAW file appearance in Lightroom. Not a great surprise this, since RAW files are supposed to evade most or all processing policies that effect in-camera JPG files. I shot an IT8 target with RAW+JPEG. Of interest was that there was some difference in color reproduction in the reds between "Faithful" and "Neutral" in the JPGs.

2. High ISO Noise Reduction settings do affect RAW appearance. This test was shot against a defocussed white wall (i.e. "featureless" image) at 2 stops underexposure at ISO 1600. There was a marked difference in appearance of the JPG files, with much less evident grain in Standard and Strong settings than at Low or Disabled. Standard was fairly close to Strong. This pattern was also evident in file sizes (more noise = larger files). For the RAW files there was a subtle difference for the same shots - more or less invisible for the lower ISO (400?) that I tested the previous evening but just visible at ISO 1600. I did not test sharpness reduction, but that's the usual trade-off. Canon, however, in this document, indicate that Adobe Camera Raw ignores this setting.

3. Noise patterns at all ISOs. Everything from 100 to 3200 in 1/3 stop increments with the same -2 stop underexposure as the previous test. Notably, stops 1/3 below the "standard" ones showed less noise, but also a little more exposure. In fact for ISO 160 the featureless noise shots perhaps showed nicer noise appearance than ISO 100. The 5D Mark II has this tendency to produce grid noise artifacts at lower ISOs. The ISO 1/3 stop above the standard ones fared least well. This pattern has been commented on previously at the Dpreview.com forums, and despite the high bs level much of the content in those forums, it appears to have a basis in fact.

Central Park, October 6th

Black-throated Blue Warbler (first fall male)


Scarlet Tanager (first fall male - notice the molt limits in the coverts)

Sora

DeKorte Park, Lyndhurst, Sept 26th 2009