Gary, I hadn't heard of camera shake associated with the number of pixels. Could you elaborate some on that?
It's not really about the number of pixels, but it's about the
angular resolution of the image taken.
The higher the angular resolution, the more likely it is that miniscule movement is captured during exposure.
Let's take an example of two imaginary cameras.
Camera 1 has a resolution of 6 Mpx, or 3000 x 2000 pixels.
Camera 2 has a resolution of 24 Mpx, or 6000 x 4000 pixels, and the sensor is
the same physical size as in camera 1.
Imagine that you're taking an image with 1/60s release time, that the subject is at the same distance for both cameras, and that you're doing so at the same focal length and aperture with the same lens.
The distance of the subject is such that one of its clear, contrast-rich edges is about one half pixel wide as seen through camera 1.
Imagine that moving the camera 1/10 of a mm results in an angular movement equal to about one half pixel, as seen through camera 1.
Then imagine that pressing the trigger causes the camera to shake just enough to move it 1/10 of a mm.
On camera 1, that movement will then probably not be visible.
On camera 2, the same movement is equal to a pixel's movement, which would definitely be visible.
But this is
only relevant if you peek at movement in terms of individual pixels.
If you take the 6 Mpx image and the 24 Mpx image and print them both at, say, 10x15 cm, then you'll be unable to see the difference anyway.
The problem is that when you have a 24 Mpx sensor, you're much more likely to want to do larger prints, or in the case of web images, you're much more likely to crop it and take a much smaller part of the image.
To confuse matters more, if the sensors' physical sizes differ between the two cameras, then this seems to change the angular resolution relationship.
For instance, the Olympus C-8080 Wide Zoom has a sensor output resolution of 3264 x 2448 at 8.80 x 6.60 mm, while the Canon EOS 1Ds MkII has 4992 x 3328 at 36 x 24 mm. At the same focal length, the Olympus has three times more angular resolution. Fortunately, you don't get a 200mm lens for this camera.
The Canon EOS 350D and 20D have 12.3% higher angular pixel resolution than the 1Ds MkII.
I have now ignored the basic issue of the narrowing of the field of view for smaller sensors, which of course is different by approximately the same as the so-called crop factor. So while the 350D and 20D can be said to have a higher angular resolution, they also cover a far narrower image at the same focal length.
But of more practical interest is that the 1Ds MkII has a higher angular pixel resolution than the 1Ds MkI, and therefore can be more vulnerable to camera shake
at the same focal lengths and FOV.
Next, I have started to wonder if there will be a next geneartion release to replace the 1D Mark II in the next 6 months. The $500 rebate that's being offered might be a strategy for clearing existing stock to prepare the way for such.
If a replacent is announced in 6 months, that's approximately 18 months after the previous release, or nearly two years after the announcement of the 1D MkII. The 1D MkI was announced about two years before the 1D MkII, so it's not unreasonable to expect a 1D MkIII (or a 2D, or some other name in the same market segment) announced before Christmas this year.