It has been my wish to do a detailed RGB and LRGB combination study using binned and unbinned color data. A series of nights presented themselves lately in which I was able to add on early morning imaging of M33 the Triangulum Galaxy. The data set includes 1x1 L, R, G, B frames as well as 2x2 R, G, B frames.

It is a wide FOV image taken with a Takahashi FSQ-106EDX4, FLI ML16200, and Astrodon Gen2 LRGB filters. Overall I like how it all combined together. For the final LRGB image shown here I decided to only use the 1x1 frames, L for luminance, RGB for color. I made other versions using synthetic L and RGB from the binned data, but I liked this latest version.

M33 has been described in great detail elsewhere, so I will not add more description here. I shot this last year with a different telescope and camera combination, so those interested can look at my description there. I chose to combine the colors using PhotometricColorCalibration in PixInsight using the G2V option. With some saturation the galaxy arms came out blue as desired.

I've included two comparison images from my study. The first is a fairly obvious one that demonstrates the reduction in noise by combining more and more integration time into the L combination. I ordered the frames in decreasing value of SNR, combining the highest SNR first and then adding more to include increasing integration time. The SNR improves with more frames, but the increase rate decreases due to combination of frames of decreasing quality. In the final image I include 240 min total rather than the 300 min available to me. The same was done with the other channels. Note that there was no noise reduction here, and comparisons were done when the images were still linear.

The second study image is more qualitative and pictorial rather than it is quantitative. Here I show RGB combination of the unbinned color channels only, then RGB combination of the binned color channels, then LRGB combination using the unbinned color, then LRGB comination using the binned color. For both LRGB combinations, I used the same L channel. It is certainly obvious to see the difference in image quality between the RGB and binned RGB combinations, where the latter shows square or diamond shaped blocky shapes. The noise is lower, of course, owing to the outcome of binning, but at the expense of reduced resolution.

But the main thrust of my study was to observe the effect on LRGB combination. Close inspection will reveal differences between using the unbinned and binned color data, but the overall effect is quite subtle. Note that I did not perform any noise reduction at this point, although I did have to stretch the data in order to perform the LRGB combination. This may have introduced differences that make it difficult to make a fully quantitative comparison. What I notice in the binned color case is that the colors have more speckled texture than unbinned case, particularly in the lower intensity regions of the galaxy and background.

This was a fun and interesting study, and I hope it helps those interested to consider the effects of binning colors for LRGB combinations. Ultimately, this choice allows for more integration time to be devoted to acquiring the detailed luminance image.