Talk:Perceptual paradox
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||||||||||||||
|
"Improperly Called"[edit]
Not to put too fine a point on it, since that appears to be where this discussion is erring, it's NOT improper to call a photon "green" or some other color. Because, just as has been pointed out, color is an issue of perception, simply stating that something has a particular color implicitly assumes we are talking about an issue of perception. That's the point which should be made, not that it's inappropriate to call an object or a light source or a photon red or blue or green, which it is not.
65.202.227.47 (talk) 13:47, 1 October 2008 (UTC)mjd 2008.10.01 09:47EDT
No Contrast Invariance Paradox[edit]
There are no units in which either Ia or Ib can be negative, and can only approach zreo at the limit of visual perception. If they actually are zero then you cannot perceive the object.
There are units in which Log[I] can become zero (in fact, if we're careful to pick them cleverly then any measurement can result in a zero or negative Log) but they are units which are not relavent to the problem. The eye can never perceive less than one photon and under bright light conditions the conversion efficiency per photon necessarily decreases. Choosing the appropriate units resolves this apparent conundrum.
That having been said, I'm not sure you have the appropriate formula for contrast. Perhaps you can reference that?
65.202.227.90 (talk) 16:32, 1 October 2008 (UTC)mjd
Chromatic Aberration[edit]
The eye does not have a simple lens. It's actually quite complex, and it doesn't act alone. The cornea is also a refracting object and in conjunction that actually do a lot to correct aberrations. Is the statement that color separation can be many photoreceptors wide based upon a thin lens calculation or is it based on real measurements? This page needs a great deal of cleaning up in this regard in general.
65.202.227.21 (talk) 16:49, 1 October 2008 (UTC)mjd
The eye also has gradient index structures which add yet another layer of optical correction. Please supply supoprting evidence for various "problems" regarding abberation, etc.--65.202.227.151 (talk) 20:50, 3 November 2008 (UTC)mjd
Source problem[edit]
Yet, the point-source is perceived clearly, and can be discriminated from other such points as little distance away as 1/10th the visual angle of a photoreceptor. According to Roorda, in a "perfect" eye, a foveal cone subtends 30 arcseconds, and the diameter of the central "Airy" disk varies between 10 cones (1mm pupil used for reading) to about 1.5 cones (8mm pupil used during fight/flight response).
Yet here: http://voi.opt.uh.edu/VOI/WavefrontCongress/2003/presentations/roorda/Roorda_Optics.ppt#375,63,Slide 63
While Roorda agrees with the assertion that a foveal cone subtends 30 arcseconds (120/deg) he then states quite clearly that the Nyquist theorem limits visual perception to 60 arcseconds, not the 3 arcseconds that the "1/10th the visual angle of a photoreceptor" statement implies. In fact, everything in Roorda's presentation agrees completely with standard optics and there is no suggestion of any kind regarding a perceptual paradox. In fact he makes it quite clear that visual perception is limited in the best case to ~20/10. He then makes this even more clear by mapping a 20/20 (limit of "average" perception) (here: http://voi.opt.uh.edu/VOI/WavefrontCongress/2003/presentations/roorda/Roorda_Optics.ppt#370,57,Slide 57) and 20/5 (here: http://voi.opt.uh.edu/VOI/WavefrontCongress/2003/presentations/roorda/Roorda_Optics.ppt#371,58,Slide 58) letter "E" onto the fovea showing that a 20/20 "E" maps out a reasonable impression of the letter "E" on the fovea but that a 20/5 letter "E" is completely imperceptable as an "E" shape because it falls below the Nyquist sampling criterion (the bar frequency in the "E" is greater than the half the cone frequency) and therefore the bars cannot be separated JUST AS TWO POINT SOURCES CLOSER THAN TWO RODS APART COULD NOT BE SEPARATED in directed contradiction to the assertion above that point sources 1/10 cone separation can be resolved. Roorda is NOT a valid source for this unsubstantiated claim. —Preceding unsigned comment added by 65.202.227.160 (talk) 20:07, 5 January 2009 (UTC)
Land Experiment[edit]
The Land experiment was done with 500 nm and 557 nm (not 500 nm and 520 nm as reported) light sources (http://www.answers.com/topic/edwin-herbert-land). One is cyan (blue-green) and the other is greenish-yellow and each is almost exactly centered on one of two distinct cones - one at 498 nm and the other at 564 nm (http://mindprod.com/jgloss/rgb.html). This gives the visual cortex enough information to determine the composition of the light source and to back the light source information out of the perceived field of view - i.e., the visual cortex interprets colors correctly because there is enough color information available in the field to do so even though there are only 500 nm and 557 nm photons received by the eye.
This is fundamentally different from and not remotely related to the RGB theory of color composition wherein colors are produced by combinations of red, green, and blue light. When all you have are red, green, and blue sources (such as in an RGB display) then color can be effectively and correctly produced in that way. When you have additional information (such as the colors inherently present via reflective/absorptive properties of objects) then the color of the light source is less important but NOT unimportant. This is yet another non-paradox that is partly explained by well known physical properties of the eye and partly by observed action of the visual cortex. —Preceding unsigned comment added by 65.202.227.100 (talk) 21:11, 5 January 2009 (UTC)