Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] experiments with dyes (was: happy equinox)



Michael Porter suggested (see below) that magenta and cyan might be misconstrued. AFAICT, they are almost always misconstrued.

A long time ago (and I don't know the reference right now) a Physics Teacher article suggested which theatrical gels under the trade names of Roscolene and Roscolux did the best job of color mixing using both the additive system (red, green, blue) and using the subtractive system (yellow, cyan, magenta). I purchased large sheets of the suggested gels from a theatrical supply dealer many years ago and I still use them. I cut some of them into 2" X 2" squares that I can distribute to a class.

I have tested the filters and they are not bad. My judgment of good/bad is done by using a spectrophotometer, and looking for the following...

The red filter transmission should fall off rapidly below approximately 600 nm.
The green filter transmission should fall off rapidly above approximately 560 nm, and below approximately 500.
The blue filter transmission should fall off rapidly above approximately 470 nm.

If you have filters like this you can use three adjustable-brightness light sources, one for each filter, and you can overlap them and adjust them to get white where they all overlap, and you can get yellow where the red and green overlap, cyan where the green and blue overlap, and magenta where the blue and red overlap. This is another way to judge them without using a spectrophotometer. My set works beautifully in this manner.

***However*** when practically any novice looks at these for the first time, they accept the green as green, but they say the blue is violet, and they say the red is way too dark to be an "ordinary" red.

For a good set of subtractive primaries, as determined with a spectrophotometer, the filter should do a good job of blocking the complementary additive primary, and have good transmission throughout the other two additive primary regions. For example, the cyan filter should totally block red (block from 600 on up) and transmit equally well through the blue and green regions (from 400 to about 560.) Likewise, yellow should block blue (everything below about 500) and transmit everything above that. Magenta should wipe out the middle (block the green region from about 500 to 560.)

If you have these, you can use them subtractively (hold two together) to show yellow and cyan make green, yellow and magenta make red, and cyan and magenta make blue.

***However*** novices looking at these for the first always claim the cyan filter is blue. They also claim the magenta filter is red, although they admit is perhaps more of a fuchsia-red. This is the misconception Michael Porter mentioned.

If you have both the subtractive set and the additive set you can also show that holding up the blue and the yellow together do not make green, but make black. ***However*** novices will claim getting black from yellow and blue (when added subtractively) is cheating because they refuse to admit the blue filter is blue. They only want to accept the cyan filter as blue (albeit "sky blue" according to some) and claim the blue filter is something else (usually violet). The best way I have been able to convince them that cyan is not blue, is to project a bright spectrum on a screen using a slide projector with a slit and a diffraction grating, and having them look at the spectrum through the cyan filter, and seeing that the cyan filter passes every bit as much green as it does blue. I ask them why they are calling it blue when it is obvious they could just as easily call it green. At this point some see the light (sorry), but others continue to think I am somehow tricking them... that the cyan filter is indeed blue, but my demonstration that it passes as much green as blue is some sort of magic trick.

Another good "trick" to do with a yellow and cyan filter is to combine them additively. Put each in separate slide projectors. Power at least one slide projector with a variac so you can adjust the brightness (you usually need to dim the yellow one). Before class, adjust the brightness so the addition of the two produces a good white color (with no yellow cast or cyan cast). Then aim the slide projectors so they don't overlap. When the class comes in, they can see the yellow and cyan projected on the screen. When class starts ask them what color will result of you move one of the projectors so that the two colored regions will overlap. Virtually everyone will say the overlap will be green. When you push the two colors together to produce white they can barely believe it. That's how I generally start my first class session about color.

For the nitpickers, it is clear that the full story of color perception and the ability to create any hue, saturation, etc. is more complicated than simple additive or subtractive mixing with three "primaries." However, the three-primary model of additive and subtractive mixing is a good place to start. I think it is difficult to move very far toward understanding color until the three-primary model is understood. Then you can examine the limitations and move beyond it.

Summary... obtaining a good set of additive filters and subtractive filters is a very useful way to start... but you still need to be prepared for a large number of disbelievers, and if you have time you should try to move a little bit beyond it.

Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Bluffton University
1 University Drive
Bluffton, OH 45817
419.358.3270
edmiston@bluffton.edu


--------------------------------------------------
From: "Michael Porter" <listmoe@gmail.com>
Sent: Monday, March 24, 2008 12:12 PM
To: "Forum for Physics Educators" <phys-l@carnot.physics.buffalo.edu>
Subject: Re: [Phys-l] experiments with dyes (was: happy equinox)

<snip>

I always thought that a casual glance at the true subtractive
primaries might lead to the misconception -- cyan looks blue, magenta
is kind of red...