Wednesday, April 7, 2010

Disk Color Mixture and Beyond

Don Hall, Former President of Applied Color Systems, Inc. (ACS) and ACS-Datacolor

[Rotating a disk for fun, knowledge, dizziness, and profit. This month we’ll hear from Don Hall, who has engaged in all of the above and even patented some of it. Don starts by reviewing a recent paper by Rolf Kuehni, and then reaches farther… - MHB]

Rolf Kuehni [1] recently published a comprehensive and well researched paper on the historic use of a modified child’s toy, a spinning top, to investigate the mysteries of color.

Kuehni starts with Ptolemy’s second-century, first-recorded, observation that a fusion of color occurs when a spinning multi-colored potter’s wheel reaches a certain speed. Eight hundred years later Alhazen, a Persian natural scientist, made a similar observation. After another 700 years, experimenters tried to understand why disk mixtures don’t ‘properly’ correlate with pigment mixtures. After this was understood, disk color mixture found practical use. In 1763, Antonio Scopoli, an Austrian physician and natural scientist, used disk mixture in classifying insect colors. Two years later, Chevalier D’Arcy, a physicist, measured the persistence of a rotating visual image using a glowing piece of coal. He needed at least eight revolutions per second for the image to fuse. (See [2] for more on the time resolution of vision.)

Over the next hundred years, Kuehni continues, disk-mixture studies revealed “accidental colors”, afterimages and complementary colors. During this period attention gradually shifted to empirically based color order systems.In 1810, the artist Philipp Otto Runge proposed such a system but had difficulty correlating Newton’s colors with pigment mixtures and decided to experiment with disk mixtures. In the process he lightness-matched the chromatic segments against the black & white segments.

Several years later, in 1855, James C. Maxwell reported using the famous disk-mixture device of his own design to advance color theory. By 1860 he switched from a spinning
disk to a visual colorimeter to match spectral colors with spectral primaries. A few years later O. N. Rood used disk mixture to reconstruct or “correct” Maxwell’s diagram.

In 1900 A.H. Munsell patented his Color Sphere which, when spun, produced neutral grays of decreasing Value from the top to bottom of the sphere. Later Munsell used disk mixture to create the color panels that populated his 3D Color Tree. In subsequent years the Munsell Company offered his color chips in a circular form with a center hole and slitted radius for mounting on a spinning disk device.

By the 1920’s and 30’s, visual photometers, tristimulus colorimeters and spectrometers obviated the need for spinning-disk mixture. However, as Kuehni reports, there was one final gasp for that old technology. In 1977, Applied Color Systems Inc. (Princeton, NJ) began to develop an instrument for measuring color materials that were not readily measured on a reflectance spectrophotometer because of texture, pattern, size or geometry. After nearly two years of unsuccessfully exploring color CRT and color projection systems, Ralph Stanziola, ACS’s Executive Vice President and Technical Director, decided to take a page out of the past, to the amazement of his associates, by developing a “Maxwell Disc” connected to a computer to match colors. Although the Visual Color Simulator (VCS-10) took three years to develop, it was technically successful as a visual color measuring input device for computer color matching and also as an accurate color simulator that could rapidly transmit visually simulated colors to remote locations. As Kuehni correctly points out, it may have worked well but it was too expensive to be widely accepted.

During the development of the VCS-10, some visual anomalies were encountered, particularly during start-up while the color disk pack accelerated to the flicker-fusion rate. Some observers complained of nausea and vertigo, and others saw different colors in their peripheral vision. To avoid these perturbations, the power to the controlled illumination lamps was not activated until the rotation was fast enough for flicker fusion.

The effect we were avoiding by turning the lamps off is related to a much weaker disk-color effect that is not properly disk color mixture (and which Kuehni doesn’t mention). Produced in 1894 by toy maker C.E. Benham, the “Artificial Spectrum Top” [3] was a disk that was one half black and the other half a white background overprinted with four areas having a series of three concentric black arcs arranged in a step-wise fashion. When this disk was spun below the flicker-fusion rate, concentric circles of weak colors appeared. When rotated the opposite way the illusionary colors reversed order.

Benham’s top is still a subject of scientific investigation [4, 5] and disk color mixture may be a matter for history. Nevertheless, I still have a Swiss Made “Optischer Farbmischer” [] on my desk as a reminder of the pleasure of working with Ralph Stanziola on the VCS-10 and in a sense reliving some experiences of the scientific pioneers that Kuehni documents so well.

1. Kuehni, RG. A brief history of disk color mixture, Col. Res. Appl. 35, 110-121 (2010).
2. Morgan Eye Center, Univ. of Utah, Temporal Resolution
3. Benham C.E. The artificial spectrum top, Nature 51, 113-114 (1894).
4. LeRohellec J, Vienot, F. Interaction of luminance and spectral adaptation upon Benham subjective colors, Col. Res. Appl. 26, S174-S179 (2001).
5. Kenyon,G, Hill, D, et al. A theory of the Benham top based on center-surround interactions in the parvocellular pathway, Neural Networks 17, 773-786 (2004).