To all,
After StarAlignment, I notice the reference image has the most rejected pixels. This makes sense to me since the reference image does not get "smoothed" by a bicubic spline operation during the StarAlignment phase (i.e. it remains in its RAW form; whereas all the other images are smoothed with bicubic spline). Do you think the reference image has a detrimental effect on the overall signal/noise ratio due to the greater number of pixel rejections? Generally, what I've seen is the number of pixels rejected in the reference image is double or more of the other images. Any thoughts on how to alleviate this potential inconsistency? Has anyone developed a good work flow that lessens the effect (e.g. ATrousWaveletTransform)?
Wade
