Wavelength of blue filtered light

Question

Does blue filtered white light have a certain wavelength, or is this highly dependent on the filter and source of light?

Answer 1

When white light is incident on a blue filter, the transmitted light consists of a continuous band of wavelengths, not just a single wavelength. The spectral characteristics of the transmitted light depend on both the spectrum of the incident white light and on the transmittance spectrum of the blue filter.

To illustrate this, consider the following system: a MagLite flashlight, with tungsten lightbulb, and various blue filters, as shown in the photo below:

The white light output of the flashlight is shown in the photo below:

and its output, through the round blue plastic filter, is shown in this photo:

The blue plastic filter was provided with the flashlight when the flashlight was purchased. Evidently, it does what it was designed to do: make blue light. But it is not possible to just look at the blue light, reflecting from white paper, and learn all that much. Instead, it is essential to use a spectrometer, spectrograph or spectroscope to spectrally disperse the light.

The next photo shows how the white light from the flashlight is input to the fiber optic and collimator of a homemade echelle spectrograph.

The echelle spectrograph disperses the light arriving through its fiber optic input cable and produces an echellogram, i.e., a two dimensional spectrum. The result is shown below:

Clearly the white light consists of a continuous band of wavelengths from violet through red, i.e., the typical rainbow.

The next photo shows blue light, through the blue plastic filter, incident on the spectrograph’s input collimator:

and its resulting echellogram:

Comparison of the two echellograms immediately leads to two conclusions. First, both the white light and blue light are broadband, consisting of continuous bands of wavelengths. In particular, the blue light is far from being single wavelength. Second, the blue plastic filter did surprisingly little filtering, i.e., it attenuated some of the longer wavelength region of the white light’s spectrum, but perhaps less than might have been supposed a priori.

Of course, there are better blue filters. The next two echellograms show what happens when blue glass filters #1 and #4, shown in the first photo, are used to filter the flashlight’s white light:

These clearly attenuate longer wavelengths better than the blue plastic filter could achieve, but neither gets anywhere near producing a monochromatic output. Even inexpensive laser pointers are not all that good as approximately monochromatic light sources:

Answer 2

What most of us think of as blue light corresponds to a wavelengths of around 450nm:

How wide the blue band is depends on where you draw the line with violet on the low wavelength side and cyan on the high wavelength side.

I found the diagram at

Answer 3

In John Rennie's answer you can see what monochromatic blue light is. When monochromatic light of (say) 450 nm falls on our retina, the intensity is weighted in three different ways by three different detectors in our eye, called cones.

(image from wikipedia).

It is seen that for this wavelength the signal from the S-cone is around 20 times as large as that from the L-cone, and maybe 12 times as large as that from the M cone. Any other combination of wavelengths that gives the same proportion will be perceived as exactly the same colour, maybe a very intense light around 400 nm combined with a distribution of wavelengths around 500 to say something.

The light coming out of your filter will have a spectral distribution whose support is a subset of that emitted from the light source, and its color will generally depend on the spectral distribution of incoming light, not only on how it looks to us. The only way to guarantee blue light to come out when white-looking light comes in is to block out everything away from a narrow band around 450 nm. This will cause some white light to be blocked out entirely however, and in general the light intensity will be greatly reduced, so I would think this will rarely be desirable.