Quote:
Originally Posted by Hagar
|
Go Douggie!
Yeah I completely forgot about this article. Great read.
Although it is true that if your camera has "big pixels" like the QHY8 (7.8um) debayering will have an impact on your resolution but I can live with it. As Terry pointed out I'm happy with doing "pretty pictures" (or trying to anyway). That's all I'm doing. The spot healing brush is my best friend and there is
absolutely no scientific value in pictures I try to present.
There is not doubt that a mono QHY9 Lum will look better than a QHY8 debayer to grayscale so Martin has a very valid point and I do believe coming from a DSLR it must have been a revelation when he saw his first Lum. Here's an example of one sub 1:1 with a bayer matrix then the red channel only, then the two greens combined, then the blue. The bayer matrix is as follow:
RG
GB
So the two greens get combined and the resulting green shows better details than the 1x red or the 1x blue that are interpolated back to the full resolution of the sensor (3000x2100), which is why they appear slightly blurry. Also at this image scale (close to 3asp) you can clearly see the red, green and blue light deviate and don't end-up in the same pixels. Not an issue at higher sampling (as Dietmar mentioned).
Having said that the picture is zoomed 200% so the pixels shown in the bayer matrix appears x2 magnified on your screen. Apply a slight deconv, reduce the original from 3000px x 2100px down to let's say 1280px or even 1500px wide pic for presentation and you'd have a nice shot and be hard pressed to see a major difference IMO. So it does the job basically at a fraction of the time.