From: Charlie Stross (charlie@antipope.org)
Date: Fri Jan 23 1998 - 03:42:42 MST
On Thu, Jan 22, 1998 at 10:12:36AM -0800, John K Clark wrote:
>
>>I think there are some figures for visual data sensitivity in Foley,
>>Van Damm et al, but it's a while since I had to read that (read:
>>since I last had anything to do with real-time photorealistic
>>rendering for simulator graphics engines). About 4kx4k pixels
>>springs to mind ...
>
> That would mean we could distinguish between 16 million objects without
> moving our eye at all! The letter "l" in "resolution" just does not seem to
> me to be made of 16 million parts, something is not right.
Dead right it isn't; becuase the letter "l" is fairly simple. Instead,
try looking at a very high-resolution true-colour image of the
Mandelbrot set. You can fit it inside the 2% or so of your visual
field that the fovea covers, and you'll see a lot more detail on it
than that nice courier-medium letter "l" on your terminal ;-)
> Even in the
> unlikely event that the fovea was capable of such a prodigious feat, I doubt
> it could transmit the description, and even if it could the brain is
> certainly not able to handle such a glut of information.
The fovea handles about 50% of the sensory bandwidth of the entire
retina, if I remember correctly. The rest isn't just "motion sensors",
but it's not far off. (You can get an idea of this by wandering around
in the near-total dark; the cones in your fovea are no use in low light
levels, so you're reduced to using your peripheral vision (mostly rods).
-- Charlie
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:48:30 MST