Say you have an LED light that emits green light. Are there sensors out there that could accurately determine the colour information of the green light? Such as the relative values of red, blue and green light being emitted from the LED.

  • $\begingroup$ Yes, its called a colorimeter. But there are other ways, such as using a camera or a spectrometer as well. $\endgroup$
    – joojaa
    Commented Oct 25, 2017 at 6:31
  • $\begingroup$ @joojaa A colorimeter I believe is used for something else entirely. I have considered using a camera but the light sensors of most digital cameras are quite limited. The light absorbed by the camera will also be affected by ambient lighting which is not ideal. An optical spectrometer sounds hopeful though. $\endgroup$
    – David Toh
    Commented Oct 25, 2017 at 6:39
  • $\begingroup$ No a colorimeter is to measure color. Color as physical phenomena is just a tad more complex than just RGB value. You don't need to shoot the object directly with a camera you can bounce it off a known color color card. $\endgroup$
    – joojaa
    Commented Oct 25, 2017 at 7:34
  • $\begingroup$ The data sheet for that LED will tell you as well... $\endgroup$
    – Solar Mike
    Commented Oct 25, 2017 at 8:02
  • $\begingroup$ The green LED is not emitting "red" light. At all. $\endgroup$ Commented Oct 25, 2017 at 19:09

3 Answers 3


What you are asking for is called a spectrometer. There are two main ways spectrometers work: with a prism or a diffraction grating. Either way, different wavelenght of light end up getting spread out spacially. You then detect the amplitude at various points along this spread-out spectrum.

Another crude way is with multiple filters, then detecting the light amplitude out of each filter. That's usually only good if you are looking for a particular wavelength, not trying to find the spectrum in general.


You want to do a light spectrum analysis on the led.

You can do this with a prism that will spread out the light into colors and then find where the peaks and valleys are in the spectrum.

That or get a Optical spectrometer which will give better data than just a colored barcode.

  • $\begingroup$ This is true for wider spectra but difficult to do with the relatively narrow band of an LED. $\endgroup$ Commented Oct 25, 2017 at 19:09
  • $\begingroup$ @CarlWitthoft if you have a better answer the answer box is right down there. $\endgroup$ Commented Oct 25, 2017 at 19:22

Indeed you need a spectrophotometer to analyze the spectrum of the light emited by the LED. However a spectrophotometer will not give you "color information" but only a spectral power distribution, something that looks like this:

Typical Green LED

Or like this:

Phosphor-converted green LED

Or even like this:

Narrow band yellowish green LED

Keep in mind, however, that this is not "color information" but only a representation of the amount of energy at different wavelenghts.

Notice also that there is nothing like "the relative values of red, blue and green light being emitted from the LED" in these graphs.

What you called "color information" can be determined by another technique called "spectro-colorimetry" or "colorimetry", which uses a representation of the human response to light at different wavelenghts to evaluate the impact of a physical light in terms of perceived color.

With colorimetry, you can start with the spectral power distribution and compute, if it exists, an equivalent "RGB" stimulus, using a defined color space such as, for example, the sRGB color space.

However, you will find out that many colors of LED lights (amber, violet, cyan, deep-red...) can not be reproduced by combining the classical RGB lights. In this case, these LED colors are "out of gamut" for your RGB system or colorspace.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.