In the simplest example, you would have, say R, V, and B data. You would first need to register these frames and trim them all to the same size. (Note that this is all separate from what you would do to extract the science from the data.) Suppose these trimmed frames are R.fits, V.fits, and B.fits. The natural thing is to map R to red, V to green, and B to blue. If you do
export R,V,B rgb ppm outtype="" outbands="zscale(b1),zscale(b2),zscale(b3)"this gives you a file called rgb.ppm, which is a color image in PPM format that you can read with 'xv' and convert to whatever other format you like. The "outbands" section describes how you are scaling the data. The example above takes each individual image and applies the same zscaling that is the default for the 'display' task. If I wanted to scale the red image with z1 and z2 values of 12. and 600., instead of the automatic values, I'd use the syntax "zscale(b1,12.,600.)". This scales the range between 12 and 600 linearly to red intensity values between 0 and 255. Everything below 12 is 0, everything above 600 is 255. (Note that b1, b2, and b3 are not variables; they have to have those names in the outbands expression.) If you want to do a logarithmic scaling of intensity, then you should use imarith or imexpr to make log-scaled images first, then use these images with export.
Now, since we don't have a B filter, it is more likely that you have V, R, and I (if you have 3-color data at all). In this case, the standard approach is to do a shifted-color rendering and map V to blue, R to green, and I to red. If you have only 2 colors, the usual trick is to fake the middle color with the average of the other 2. For example,
export I,V rgb ppm outtype="" outbands="zscale(b1),zscale(b1+b2),zscale(b2)"Depending on the counts in your data you may want to use a differently weighted average. Also note that in the above examples no attempt is made to match the resulting color to the true color indices of the objects. To make that kind of a color shift you would want to either explicitly scale your images or carefully manipulate the z1 and z2 parameters.
For the truly adventurous who want to experiment with software outside of IRAF, the CIAO package includes a task called dmimg2jpg that does exactly what it says. To initialize the CIAO package, execute the following line at a UNIX prompt or add this line to your .cshrc, log out, and log in again:
alias ciao source /iraf/ciao-2.2/bin/ciao.csh -d /iraf/ciao-2.2Then type "ciao" at a UNIX prompt. After doing this, all the CIAO tasks are available, in effect as UNIX commands. Typing "ahelp dmimg2jpg" will get you the documentation on the task. A word of advice: don't run CIAO and IRAF tasks in the same xterm or xgterm window.