In 1980, Dallas inventor Kenny Whitright devised the first viable color changer adaptable to basic fixtures, such as par cans, and not requiring individual operators for each fixture. The invention was the gel scroller. The scroller used a series of individual gels taped end to end to form a scroll, which was connected at the ends and wound around motorized rollers at opposite ends of a metal housing affixed to the gel frame mount of a par can. Each scroll was marked with a tab that passed through slot in the housing containing an optical sensor that relayed the current position of the scroll to a controller. With the use of gel scrollers, a LD no longer had to add a set of pars for each color wash he needed in a design. A single set of wash lamps could produce multiple user-defined colors. There were some drawbacks. The early scrollers tended to have frequent mechanical problems, including the scrolls disconnecting from the rollers, gels separating from each other, and scrolls deformed by heat and tension forming ripples at the edges that would jump out of the sensor track causing the scrollers to lose track of position. They were also slow enough that color bumps were not really feasible, particularly for non-adjacent colors in the scroll. Later refinements resolved most of these issues. The fundamental design was seized upon by a number of companies.

In 1981, the French firm Chameleon introduced a new intelligent luminaire, the Telescan, which relied on new computerised robotic moving mirrors and used three gel sets (one each of cyan, amber, and magenta) graduated from solid gel through gradually larger perforations and terminating in clear acetate. This mechanism was capable of smooth color crossfades using the subtractive method. The drawbacks were that gels have a fairly short life span, tending to lose saturation, and that the perforations tended to pucker under heat and tension, resulting in an undesirable diffusion effect. This system was also used in par can mounted scrollers.

Also in 1980, and also in Dallas, a small team of engineers from Showco began a project that ultimately resulted in the 1981 debut of the VariLite. The VariLite VL1 was a fully automated arc-sourced spot luminaire whose focus (pan-tilt), color, gobo, and intensity could all be controlled remotely. The color system consisted of three servo-driven wheels each containing five different color filters and one open slot. This resulted in the theoretical capability of creating 216 colors (6x6x6) by the subtractive color mixing method. The filter material was dichroic glass, a production lighting first. Dichroic filters have some interesting properties. They do not absorb light, as do other filters, but rather reflect all wavelengths they do not allow to pass. As a result, they have vastly greater life spans than do gels. They are also capable of creating much more precise colors as well as colors of much narrower bands of the spectrum. As with Whitright’s early gel scrollers, there were some drawbacks to the original design. While color changes were quite fast, the servos that drove the color wheels were noisy. Calibration was tricky, and the wheels were prone to a condition called “chatter”. The wheels were gear driven, and the mesh between the teeth of the wheel and the motor’s drive gear had to be carefully set. Too tight: the wheel would bind; too loose: the drive motor would oscillate against the wheel as it attempted to find equilibrium. If not properly damped, this oscillation could escalate into the violent (and noisy) vibration referred to as chatter. It was often visible as a pronounced vibration at the edge of a beam, and was destructive to the mechanism. Another problem was that the dichroic filters were small round discs RTV’d into circular cutouts in the wheels. As colors changed, the metal wheel material between the filters produced momentary flickering blackouts. Lastly, the various filter combinations were not all equally aesthetically pleasing. Many of the colors, and particularly those created with a three-filter combination, were either inconsistent or simply unusable.

Click here for Part 2

Email this story to a friend