Tuesday, April 29, 2008

The mistery of stable images

We know optically the eye is like a simple meniscus camera that projects an image onto the retina. We also know that on the cortex there is a holomorphic map of the visual field. However we know very little of what happens in between. For example, in the lateral geniculate nucleus (LGN) there are seven layers, and if we stick a toothpick in a point like in a club sandwich, the layers are geometrically aligned.

But the human visual system (HVS) is not like a camcorder. Here is a simple test: take a camcorder and film what while you move down the street. Try doing this while walking, running, riding a camel, riding in a sports car. While you are doing this, each time you see the same scene.

Now remove the context, i.e., sit on a chair and watch what you filmed. The sports car piece will probably just be blurred, the other pieces will probably give you motion sickness and you will not make out much visual information. If you ever used a virtual reality system you know what I mean.

Add to this that you keep moving your head, and between fixation points your eyes keep saccading. We must conclude that the map on the cortex is not an image of the retina but an image of the real world. How does the HVS perform this feat?

Tim Gollisch and Markus MeisterTim Gollisch of the Max Planck Institute of Neurobiology and Markus Meister of the Department of Molecular and Cellular Biology and Center for Brain Science, Harvard University have solved this riddle and published their data on page 1108 of Science 22 February 2008.

The trick is performed by the retinal ganglion cells, which come in pairs of fast OFF cells and biphasic OFF cells, which receive inputs from both the ON and OFF pathways, and the slow OFF and ON cells. It turns out that they have different spike latencies, which can be e a powerful mechanism to rapidly transmit a new visual scene. Moreover, certain neurons in visual cortex are exquisitely sensitive to the coincidence of spikes on their afferents (30), which is one possible readout mechanism for a latency code.

The autors conclude that it is conceivable that early aspects of sensory processing operate on the basis of the classification of spike latency patterns.

Friday, April 4, 2008

Promoting happiness

Last December I wrote a short post about a Science paper providing neurophysiological evidence for the importance of social comparison on reward processing in the human brain. The last print version of Science has a paper teaching us how we can be even happier.

Elizabeth W. Dunn and Lara B. Aknin of the University of British Columbia, Vancouver, BC, and Michael I. Norton of the Harvard Business School in Boston write in their paper Spending Money on Others Promotes Happiness that although much research has examined the effect of income on happiness, they suggest that how people spend their money may be at least as important as how much money they earn.

Specifically, they hypothesize that spending money on other people may have a more positive impact on happiness than spending money on oneself. Providing converging evidence for this hypothesis, they found that spending more of one's income on others predicted greater happiness both cross-sectionally (in a nationally representative survey study) and longitudinally (in a field study of windfall spending). Finally, participants who were randomly assigned to spend money on others experienced greater happiness than those assigned to spend money on themselves.

They conclude that given that people appear to overlook the benefits of prosocial spending, policy interventions that promote prosocial spending — encouraging people to invest income in others rather than in themselves — may be worthwhile in the service of translating increased national wealth into increased national happiness.

The United States Declaration of Independence already postulates a right to pursue happiness, and the IRS encourages charitable donations by allowing a generous tax deduction for them, so our Government has already made the correct policy interventions. Now, when I look at my bank transactions for March, I am convinced I must be the happiest person in the world!

bringing gifts

Thursday, April 3, 2008

Color Chart: Reinventing Color from 1950 to Today

Carinna Parraman wrote: "Check out the Internet version of the Color Chart exhibition at MOMA in NY, it is beautifully executed and certainly worth a visit."

Follow this link to MOMA to see what you can do with creativity, color, and Flash.

Carrie Mae Weems, blue detail from Moody Blue Girl, 1988

Blue citation from Carrie Mae Weems, Moody Blue Girl, 1988

Wednesday, April 2, 2008

Administrative note and color lawsuits

First an administrative note. Most feedback we get from you, our esteemed readers, is in the form of personal email. Only rarely are we able to generate sufficient controversy to spark a debate in the blog comment section, such as with Non-local realism, An On-Line Color Thesaurus, or yesterday's Revolutionary White Reflectance Standard for Metrology. Therefore, we are happy for every good comment we get. However, as you are aware our blog server is rather crafty, and it is difficult for us to find comments when you replace the post title with your own title. This summer HP will be upgrading to commercial blogging software and this blog will run smoother, hopefully even multilingually. In the meantime here is my answer to a comment on color lawsuits I was unable to locate.

IANAL (I am not a lawyer), so I cannot tell you how many colors the Constitution on the Laws thinks you are seeing or entitled to seeing. From a color science point of view, color does not exist in nature, it is an illusion that is elicited in our visual system.

Colorimetry is the art to predict an illusion from a physical measurement, hence what we do in color reproduction is to try to build models that allow us to make statistical predictions of this illusion. Our supreme authority is the Commission Internationale de l'Éclairage (CIE), which in Definition 845-02-18 defines (perceived) color as follows:

Attribute of a visual perception consisting of any combination of chromatic and achromatic content. This attribute can be described by chromatic color names such as yellow, orange, brown, red, pink, green, blue, purple, etc., or by achromatic color names such as white, gray, black, etc., and qualified by bright, dim, light, dark etc., or by combinations of such names

Perceived color depends on the spectral distribution of the color stimulus, on the size, shape, structure and surround of the stimulus area, on the state of adaptation of the observer’s visual system, and on the observer’s experience of the prevailing and similar situations of observation

Perceived color may appear in several modes of appearance. The names for various modes of appearance are intended to distinguish among qualitative and geometric differences of color perceptions

As all the et cœteras in the definition reveal, there is nothing that allows you to count how many colors you can see. A metric you could use is to enumerate all the color names you can tell, i.e., the size of your color lexicon. However, the color lexicon is acquired, so its cardinality depends on your experience. The cardinality also depends on time, as we name more colors the more evolved the civilization gets.

For example, 2000 years ago, the Romans could not distinguish between blue and green. More recently, 1000 years ago, the Japanese, which had three colors — white/pure (shiroi), black/dark (kuroi), and colorful/red (akai) — added a fourth color to their vocabulary because tree leaves, sky, and the sea are colorful but not red, hence aoi became the name for those things, without distinction between green and blue. Even today midori (green) is not an attribute but a substantive; the color of an unripe apple is blue (aoi), not green (midori) because it does not make sense from a grammatical point of view.

aoi ringo

We can use psychophysics to start with one color, then change is slightly until we perceive a just noticeable difference (JND), increment the counter by one, and start over with the next iteration step. This way we could determine that we can see something between 6,000 and 10,000 different colors. But when you look what populations name distinctly in a color thesaurus, you typically find a 700 to 900 word dictionary.

If we go back to the lawsuit discussed in that blog comment, we could take the position of the electrical engineer and look at the addressable number of colors. When a display can address 16,777,216 colors, this is many more than you can actually see, and even 262,144 is much larger than 10,000. From a color science point of view, the point is moot.

Turning the argument around, a printer can only put down or not put down cyan, magenta, yellow, or black marks. Yet you would not claim the printer is capable of printing only four colors. The trick is that the human visual system (HVS) has a limited resolution and therefore you can halftone colors with dithers.

In display monitors and TV the spatial resolution is much lower than in printers, so instead of spatial dithering, temporal dithering is used, but you still see the same color, it was just cooked differently. Remember, color is just an illusion elicited in the HVS.

One complaint in that post is that you can see artifacts when you can address less than 16,777,216 colors. That is moot too. Give me any two colors on an 8 bit display capable of displaying only 256 colors and I can produce a completely smooth gradient between your two colors, with each step below one JND.

This is the HVS as we know it in color science. IANAL, and I do not know if the law refers to a human visual system. For that matter, I do not even know if lawyers might live in a world where the photons are colored.

To close the loop, if you have comments, do not send me email, post them to the blog without changing the title.