RGB colors vs Wavelengths

[jfloots]

I was wondering if anyone perhaps knows the equivalent RGB colors on a pc for the different wavelengths? For example:

Red / 650nm / #FF0000
Blue / 473nm / #0000FF
Green / 532nm / #00FF00
Blu-Ray / 405nm / #8800FF

etc., but obviously there is a difference between 650nm red and 670nm red etc, so all of them can't be on #FF0000, but if you knew which nm was exactly at #FF0000 etc. then you might be able to determine where the rest is at???

[mixedgas]

etc., but obviously there is a difference between 650nm red and 670nm red etc, so all of them can't be on #FF0000, but if you knew which nm was exactly at #FF0000 etc. then you might be able to determine where the rest is at??? [/QUOTE]

It will not be a linear 1:1 correspondence as the wavelengths do not shift, but the intensities do, and unless you have a new LCD monitor with LED backlights, the monitors give off bands of color, not monochromatic lines like a laser. There is a great deal of complex research done on matching monitor and printer colors, and the math is complex. It is different depending on the eye's viewing conditions and source power levels as well. for example, photopic vs scotopic viewing.

For a test of passband wavelengths of a typical led based LCD, look here:

http://www.xbitlabs.com/articles/mon...meters_15.html

For P22 standard phosphors, the data is here:

http://us.geocities.com/columbiaisa/crt_phosphors.htm


Some flouro backlight spectrum shots are here, scroll down

http://www.chemistryland.com/CHM107L...ctroscope.html

From what I have posted here, the folks on PL that specialize in color (Tocket et al) will help you with your question. I imagine it will be quite a discussion. I don't want to scare you, but you have opened quite a can of worms.




Steve

[jfloots]

Wow that is complex!!! Ok let's close this can of worms then immediately!! :-D

What I was actually just trying to get at is which wavelength of red is the prettiest? Some reds will be more towards the orange side where others will be more towards the neon pink side and I imagine those ones would look nice indeed?

[Doc]

red can be "orangy", but it can't be pink without blue. Pink is a mix of colours, orange can be prime.

[allthatwhichis]

I don't think there should be a "neon pink" shade of red; 640 is becoming a favorite as it doesn't have the orangish look of 635 but has the nice beam size like 660 but is much brighter to our eyes. I have done a side by side with 660 and 640 and they almost look exactly alike.

[tocket]

Simply put, the color of lasers can't be represented accurately in the sRGB color space, which is a pretty good model for what computer displays can show.

A light source can be characterized in color by its coordinates in a color space. There are many such color spaces, but the most commonly used is CIE 1931. The coordinates for reds almost all the way up to 600 nm lie outside the gamut in such a fashion that their closest match is in fact #FF0000. That's how poor computer display devices are, or how good lasers are, depending on what way you look at it.

If you want to convert wavelengths to RGB colors you can actually use my Chroma application (search for it). Enter the wavelengths, press update and open new plot window. Then use the data cursor to click on the background (not on a point or line). You will then see the CIE coordinates and the RGB values of the select point. You can move the cursor around with the arrow keys until you hit the pink square that is the desired wavelength and read the RGB value.

The value will be in a decimal format with 1 as the maximum value, but that's easily converted to an HEX value.