As I read it, he is saying that best available resolution is only available
when the lines on the test target are aligned with the pixel array.
That reminds me. I was reading a magazine in the library the other day. It had an article about the freeze-frame techniques being used in the film and advertising industry (basically the "Bullet-Time" effect from The Matrix).
One guy was using an array of digital cameras, D60's if I remember correctly. He said that regardless of how well you align the cameras there will always be some variation in framing, because the mounting of the sensor in the cameras can result in upwards of 40 pixels of displacement from one camera to the next.
They use software tools to correct for framing variations between the cameras. The whole process is becoming pretty automated now due to the digital capture.
Cheers,
- Dave
http://www.digistar.com/~dmann/
