Our human interface with reality

Our subjective experience of the world that surrounds us is incredibly rich and detailed.

If we estimate the equivalent information rate of this experience in the ubiquitous units of bits per second, we find a staggeringly high rate, 1000 Gigabits per Second or more.

The bit rate of our subjective visual experience of reality

What bit rate is our subjective visual experience of reality equivalent to?

There is no way that we can measure an equivalent information rate directly. However, we can approach the question by considering the quality of our visual experience, or more specifically to ask:
What information rate would be required to create a dynamic image of such quality that we would be unable to distinguish it from reality?
Despite all the hype, even the best of today’s High Definition TVs only provide us with a moderately sharp picture. No-one looking at an image of Mount Everest would be fooled into believing that they were actually seeing Mount Everest itself even with the latest 3D technology. However, the ultimate high-definition television would provide a dynamic image so accurate, that we would be unable to visually detect the difference between the image and a real view of a scene, even with the rapid action in a football match.

The High Fidelity TV display calculation:

We can base the calculation on the following properties of the human eye: The ability to resolve fine spatial detail, the extent of our field of view, the speed of response to changes in the scene, the accuracy with which we can resolve colour, and depth.

Spatial information:

With our head and eyes stationary, our entire field of view extends 180 degrees horizontally and 130 degrees vertically. Multiplied together this gives a figure of 23,400 square degrees, more conveniently described as 4.5 steradians of “solid angle”.
The angular resolution of the human eye is 0.1 milliradian (10-4 radians), or 10-8 steradians solid angle.
So to provide a high resolution image with sufficient resolution would require a 450 Megapixel display (the ratio between these two figures).

Now add colour information:

We can resolve approximately one million colours [1]. So to define each pixel sufficiently to achieve perfect colour and brightness would require 20 bits (as 220= 1 million approximately).
So one static High Fidelity image with perfect colour and resolution would require a 450 Megapixel display and 9 Gigabits of information.

Changes in Brightness:

Our eye can sense fluctuations in brightness as fast as 50 times per second, so to unambiguously display changes in brightness this fast would require this image to be updated at twice this rate, so at 100 times per second.
To update our 9 Gigabits of information one hundred times a second would require an information rate approaching 1000 Gigabits per second (1 Terabit per second).

So our visual experience of reality is equivalent to an information rate of One Thousand Gigabits per second!

(Eric Nussbaum of Bell Labs performed a similar calculation in 1987 and published a similar figure [2]).
Some estimates of colour resolution are ten times greater [3], and if we add eyeball rotation, it adds another 50%, so taking the total very close to the huge figure of 16 Terabits per second proposed by Jim Crowe in the context of a High Fidelity “Telepresence” communication system, in which two people at different places on the planet could experience sharing the same space[4].

References:
1/. “On the number of absolutely identifiable spectral hues”, Halsey, R. and Chapanis, A. (1951), Journal of the Optical Society of America, 41, 1057-1058.
2/. “Integrated broadband networks & services of the future”, Eric Nussbaum, Paper MB2, Proc. OFC/IOOC 1987, Reno Nevada, Jan 19-23, 1987
3/. “Color in Business, Science and Industry”, by Judd, Deane B.; Wyszecki, Günter (1975), Wiley Series in Pure and Applied Optics (third edition). New York: Wiley-Interscience. p. 388. ISBN 0-471-45212-2.
4/. Quoted by Joshua Shapiro in Wired magazine, Issue 6.11, Nov 1998. http://www.wired.com/wired/archive/6.11/crowe.html

Comments: