Best Approach in Creating a Synthetic Luminance?

[twade]

To all,

I have a RGB image that I would like to process using a synthetic luminance.  The  image has already been flattened.  First, I change the RGBWorkingSpace to 1:1:1.  The image is of the Rho Ophiuchus region so it is likely best to keep the ratios the same.  Next, I extract the Luminance channel.  At this point, I'm not quite sure what needs to be done.  How far do you stretch the luminance channel?  Do you also need to stretch the original RGB image the same amount?  What's the best way to combine the processed luminance channel back into the RGB image?  Any suggestions and/or examples would be greatly appreciated.

Thanks,

Wade

[Jack Harvey]

I have done this many times and find that you have to be careful stretching the lum and it varies from image to image for me.  WHat often works well is to extract the lum and do some deconvolution on that channel, work on the RGB and then either extract the R,G, and B and use the LRGB combine to recombine the four channels or you can try just using ChannelCombine  to add the Lum back to the RGB.