Sandy, yes, you are right in those assumptions. A spectral transmission chart is made for the filter comparing wavelength vs. absorbance (or transmission - it doesn't make much difference since transmittance and absorbance are directly related to each other).
Originally Posted by sanking
The wavelength at which the maximum tranmittance occurs is noted. This will probably be close to the "dominant wavelength" of the filter, or its "hue" as seen with our eyes. (Note, the dominant wavelength will shift a few nanometers depending on the color temperature of the light source, the maximum transmittance will not change with color temp. The dominant wavelength also relates more to how the eye sees colors on the color wheel than what its maximum transmittance is.)
The chart can then be "normalized" by setting the tranmittance at that wavelength to be 100%. All the other transmittance values for each wavelength will then be adjusted by the same ratio as was used to set the 100% point and we can make a new chart.
So now you have a chart that has been "normalized", with the wavelength of maximum tranmittance at 100%. To find the Tmax/2 point, look for where the plot of the filter crosses the 50% tranmittance mark. For a band-pass filter like we are describing here, we hopefully have a filter that "cuts-off" sharply and is fairly symmetrical. Think of a steep, sharply sided Bell-curve used for grading papers or in statistics, the shape should be similar.
The #93 and #94 Blue Wrattens both do this, but the #92 Red only cuts-off on one side! It has a transmittance of only 0.40% at 620 nm, a dominant wavelength of about 645 nm where it transmits about 60% but it has a peak transmittance of 88% at 700nm. But it does not drop off on the Infrared side. They must be assuming that since all of the filter in this series must be used with a #310A IR cut-off filter, the IR-blocking filter will take care of that side (it starts to filter out at around 680nm.)