iraf.net

Hi Fitz,I'm Vega. My currently work is to measure the radial velocity of variable stars. I found that fxcor task has been widely used to measure the 'rv' of celestial objects. The objects that I working on are the stars which have emission lines like Be stars and I use a standard star of radial velocity which had a very deep H-alpha absorption line. I have been trying to use this package but facing some problems:
1. The correlation value is very small below 0.1. No matter how I change the value of filter parameters the correlation height still below 0.1. The funny things is that sometimes the value of Vh, Vo seems to be good or reasonable. I know that the correlation height should be around 0.8 to justify that the correlation is good.
I have set both the object and template have the same dispersion value using dispcor, both have w1=6527.438, w2=6596.07, dw=0.999, nw=688. With these information how can I set the values of cuton, cutoff, fullon and fulloff so that will increase the correlation height? How ddoes the pixel number related with wavenumbers?
2. Why does the emission line in object's spectrum disappeared either in window display or in fourier mode after and even before the filtering. The spectrum shows in the window display also very much different with the actual object's spectrum. I only can see the H-alpha absorption line in the template spectra but the H-alpha emission line of the object seems disappeared. Why is that happened?
3. How can I justify the radial velocity being measured is true if the correlation height less than 0.1?Thank you very much in advanced for the helps.Regards,
Vega

Hi Fitz,Thank you for the help. I have put my first post in the forum. I want to upload the images but the ftp to the address failed. Is there other option to upload the files?
Beside H-alpha I also try on other lines like MgII 4471. I was surprised that I can make the correlation with height at least 0.7 and sometime I get 0.85. There are changing in the value of Vh of the line which were expected. I have been thinking why I cannot make it on H-alpha. One thing that I realised is that the profile of the MgII of the object and template are about the same but not exactly, however for H-alpha, it has a double peaks profile and the template only a normal absorption line. Is this the reason why I cannot make a good correlation? If it is the reason how to solve this?Since I managed to measure the radial velocity of MgII, I want to plot the line in radial velocity scale using gnuplot. I have tried under splot and using 'v' to toggle the velocity scale but the pixel value still in wavelength scale. How can I change the pixel value to the measured radial velocity scale?
Hope for the help.Thank you.Regards,
Vega