Hello Olivier,
if I understand you correctly, you would like to make an image in true color from your spectral images corresponding to different spectral lines. This is quite a bit more ambitious than the false color images, where the different colors represent different elements, e.g. red for hydrogen, green for oxygen and blue for nitrogen or whatever you like.
In general, if you have a spectrum for every pixel in your image you can calculate the coordinates in a suitable color space by calculating a weighted integral with a color matching function, which describes the eye response to the spectral energy for the different color sensors of the eye for an average person (I am very sorry to hear about your color vision problem). In a second step these color coordinates are transformed into RGB values, depending on the screen on which you would like to display your images, in general a typical computer or TV display. A good description of this process you find e.g. here by John Walker:
https://www.fourmilab.ch/documents/specrend/
(John Walker is not the Whisky drinker but the founder of Autodesk, as a hobby he writes programs to all kinds of things such as calendars, planetariums etc.)
It describes well how the full color range of a spectrum is reduced onto the color space posssible by the colors of a screen.
In the case of spectra of a planetary nebula, the whole process can be reduced to a weighted sum over the different line spectra (at least if you can ignore the continuum of the central star). The two steps of determining color coordinates and conversion to RGB described above and the avoidance of negative coordinates can be done first and then be applied to the different spectra. The formula converts energy at wavelength lambda (in nm as used by physicists) into RGB values, therefore it assumes correctly calibrated spectra (instrument response corrected is sufficient).
I have worked before in calculating color coordinates (for optical coatings such as AR coatings on lenses and solar cells etc.) so I include an EXCEL sheet with the values you need (some additional tables useful for color calculations are included as well, but you only need the 1931 color observer):
In this file you find for each wavelength (you have to interpolate if necessary) the values of the color matching functions x,y,z(lambda), the transformation to R,G,B by multiplication with the transformation matrix and finally the needed R',G' and B' desaturated (nonnegative) values according to the descriptiion by John Walker.
So as an example for the line H-beta:
Interpolated values:
H-beta G' B'
486.1 0.218 0.290
You would have to multiply the instrument response calibrated H-beta image by 0.218 and add it to the green image and multiply it by 0.290 and add it to the blue image, which gives in the final RGB image the typical blue green color of the H-beta.
This you would have to repeat for the other lines. Fortunately you have to calculate the weights only once, for H-alpha:
H-alpha R' B'
656.3 0.348 0.026
with practically only contributing to the red image and reflecting the low sensitivity of the eye at 656 nm.
Important: do not lose the intensity calibration on the way, some images corresponding to weak lines should be almost black in comparison.
(In a more precise calculation you would calculate the correct RGB values for the image first and then apply the desaturation to the possible color space, but this is more complicated to calculate, requiring pixel by pixel arithmetic of the color coordinates, so first I would try the simpler method. Anyway, once you have a true color image, you can still "improve" it by stretching the color contrast and gamma)
I am sure you can write a small procedure to do the weighted sum of the individual images corresponding to the different wavelengths, in IRIS or Audela, then you should obtain beautiful and true color images of NGC 6543
Please ask if something is not clear and show us your result.
It looks that with your new software you achieve remarkable results, I assume you have written a script to move the slit across the nebula.
Here in Switzerland the weather has not been so favorable for much observing, so I was only able to observe NGC 6572 a few days before the anniversary:
http://www.astronomie.ch/forum/viewtopi ... 63&p=30197
The call to observe was drenched in rain
Regards, Martin
PS:
I noticed that there was a mistake in the calculation of R'G'B' in the original post. I have corrected it in the meantime and hope that everything is ok now. I have to admit that I have not tested it with any spectra.
Martin
], I have hard time to find the color balance. At the same time, I guess one can find the balance parameters in a more scientifical way... any idea?