Extracting H-Alpha from DSLR image

[kkoppert]

I have a Canon DSLR which has been modified to allow H=a and I/R through. I have taken an image of GUM 39 using a 7nm H-Alpha filter. The result is a very red image with noise in green and blus. I want  to extract the H-a signal (as much as possible). Do I simply use Çhannel Extraction to extract the red signal or should I use Debayering first? I guess the same process would work for OIII and Hb?

[dzso.bacsi]

I don't know the right answer, but this may help you:
http://pixinsight.com/forum/index.php?topic=5748.msg39771#msg39771

[Phil Leigh]

after normal bias/dark/flat calibration, Debayer using the Superpixel mode (not VNG) - this prevents the blue and green "noise" from polluting the nice clean Ha data. Then use rgb channel separation to discard the b and g channels - then register and integrate the "red" (mono Ha) images.

The superpixel mode reduces the image size in pixels by 75% for obvious reasons.

With Oiii you can do the same thing but this time you need to discard the red and keep the blue and the green (oiii overlaps both)

Sii is the same as Ha (keep the red)