How to multiply spectrum with an "emissive light spectrum"?

[ademmler]

Hi,

this are my first steps and tries in Colour - forgive me if I am doing fully wrong ;-)

How can I multiply a given spectrum of a color with a "spectrum of a light source"?

... sd_blue = colour.SpectralDistribution(data_blue, name='Blue') sd_light = colour.SDS_ILLUMINANTS['B'] / 100 sd = sd_blue * sd_light print(repr(sd))
I get this error:
RuntimeWarning: invalid value encountered in less_equal
V = np.where(L <= 0.0031308, L * 12.92, 1.055 * spow(L, 1 / 2.4) - 0.055)

As far as I see the spectral data does not match.
Do I have to use shape and if yes how to get them ready for multiplication?

[KelSolaar]

Hi @ademmler,

The warning you get is probably related to a different section of the code, it is coming from the colour.models.eotf_inverse_sRGB definition.

Multiplying two spectral distributions should be straightforward: You simply multiply one with the other! There is one important thing to consider though: Whether their spectral shapes are the same.

• If the spectral shapes are differents, you will get NaNs where the bins do not match.
• In that case, you can leverage the fact that the colour.SpectralDistribution is built on top of a continuous signal representation and thus evaluate that function on the wavelengths of interest.
• Alternatively, you can align the spectral distributions together before the operation:

In [1]: import colour

In [2]: a = colour.sd_ones(colour.SpectralShape(400, 700, 10))

In [3]: b = colour.sd_ones(colour.SpectralShape(400, 700, 20)) * 2

In [4]: (a * b).values
Out[4]:
array([  2.,  nan,   2.,  nan,   2.,  nan,   2.,  nan,   2.,  nan,   2.,
        nan,   2.,  nan,   2.,  nan,   2.,  nan,   2.,  nan,   2.,  nan,
         2.,  nan,   2.,  nan,   2.,  nan,   2.,  nan,   2.])

In [5]: (a * b[a.wavelengths]).values
Out[5]:
array([ 2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,
        2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,
        2.,  2.,  2.,  2.,  2.])

In [6]: (a * b.copy().align(a.shape)).values
Out[6]:
array([ 2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,
        2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,  2.,
        2.,  2.,  2.,  2.,  2.])

[ademmler]

Thx for your help!