Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!elroy.jpl.nasa.gov!jpl-devvax!mark
From: mark@jpl-devvax.jpl.nasa.gov (Mark R. Rubin 4-7794)
Newsgroups: comp.music
Subject: Re: Perfect Pitch
Message-ID: <1991Apr3.171021.9733@jpl-devvax.jpl.nasa.gov>
Date: 3 Apr 91 17:10:21 GMT
Reply-To: mark@devvax.JPL.NASA.GOV (Mark R. Rubin 4-7794)
Organization: Jet Propulsion Laboratory, Pasadena, CA
Lines: 68


Another $.02 Increment About Absolute vs. Relative Color Sense

o Back in the early '80s I attended a public talk at Caltech given by
  Dr. Edwin Land (sp?), inventor of the Polaroid camera. All of what
  follows is filtered through 10 years of a memory clouded by too
  many late night debugging sessions, but ...

o Most of the lecture was about color theory. A lot of it was work in
  progress, and he didn't have a tidy, all-encompassing conclusion to
  summarize with.

o The most striking demonstration he gave, however, concerned
  relative vs. absolute color sense. He had some slides that he had
  dubbed "Mondrians", after the painter. More like the early abstract
  Mondrain paintings (before the thin stripes with little color
  squares in them), they were overlapping rectangles of solid colors.
  Anybody who's written the following code to test out a graphic
  system:

	LOOP:
	x = random() ;
	y = random() ;
	width = random() ;
	height = random() ;
	color = random() ;
	rectangle(x, y, width, height, color) ;
	goto  LOOP ;

  knows what they looked like, but Land emphasized that the colors
  had _not_ been chosen randomly.

o To demonstrate, he pointed to a particular color rectangle in one
  of the slides, and asked the audience what color it was. Everyone
  said the obvious: "Green" (or some color -- the point is there was
  no uncertainty). Land then took a piece of cardboard with a small
  round hole, held it in the beam from the slide projector, and
  manipulated it until all that shone through was a small portion of
  the "green" rectangle.

o A gasp arose from the audience. The round spot of light on the
  screen was bright orange. (Again, the particular colors,
  green->orange, aren't clear in my memory, but they were that
  radically different. Not red->orange or the like.) It was as
  astonishing as any of the classical optical illusions (necker
  cubes, equal length line segments overlaid on perspective
  converging railroad tracks, etc.) Land removed and re-inserted the
  cardboard, and each time the color changed.

o He then aimed a sensor at the spot which displayed the red, green,
  and blue components of the light. (Three photocells with different
  colored filters on them.) Once again, he put the cardboard in and
  out, and the color changed back and forth. The red, green, blue
  readings didn't move.

o I don't know what all this proves, but I've been convinced ever
  since that color perception is based on visual environment. Note
  that this is different from the "take off your orange sunglasses
  and the world looks blue" phenomenon -- there was no adjustment
  period.

o BTW, all of these comments come from an amateur musician who's
  still trying to develop reliable _relative_ pitch, much less
  absolute. Any tips welcome.



Mark R. Rubin		mark@jpl-devvax.jpl.nasa.gov