vdB 142 - Elephant's Trunk

Taken over 2 nights for a total of 6 hours in 15 minute subs. As you can see the framing is rather poor. The first set was much better but the alignment on the second one was wrong because I used the wrong star as my slew target. For the 3rd attempt I actually plate solved the first image and issued a goto to those coordinates. The framing was great but guiding went off-track after the first image so that entire session was worthless. All three sessions were started after midnight, leaving the setup alone. Guiding with an OAG at F/10 is a skill I'm slowly mastering but there's still some ways to go. I waited for the first image to come in which was fine but after that all frames were unguided and therefore bad. There is still some time in the season to retry this. I think the image shows the promise of imaging this area at 2800mm FL. The next session the camera will be oriented correctly too. A wide field image of IC1396 which contains vdB 142 which I shot back in 2009. The van den Bergh catalog contains interesting but somewhat obscure targets involving reflection nebulae. These are clouds of material that don't emit light but rather reflect it. In this image the red color is generated mostly by hydrogen atoms emitting red light (Ha or H-alpha) after having been energized by radiation from nearby stars. Look carefully near the top of the image and you'll see small white/green areas where the light from bright stars reflects off inactive gas and other matter. These are the reflections of vdB 142.

(click for 1:1 version)

NGC 6946 - Fireworks Galaxy

Taken over 2 nights for a total of 4 hours 15m in 15 minute subs. I actually took this before the M16 image I published earlier. I was fiddling with my OAG setup and the moon was out for quite a few of the subs that were stacked. Still the result was worth processing if only so I have a benchmark to see how I can improve. Imaging at 2800 mm (HD 11 at F/10) is a challenge especially considering how poor the seeing is here. I would watch the guide star in PHD and it would move several pixels without any real reason. Guiding this setup is a challenge and occasionally the guide star will go on walkabout and ruin the sub. I'm not sure yet what causes this. The Celestron/Baader OAG doesn't have a very sturdy guide cam holder and it requires some tweaking to make it rigid.

(click for 1:1 version)

M16 - Eagle Nebula

Taken over 2 nights for a total of just 2 hours and 20 minutes in 10 minute subs. M16 is very low on the horizon to me with trees that block the view. Add light pollution from Norwalk and NYC in that direction and the opportunities to shoot this gorgeous target are few. Because the second set of images wasn't aimed right at the same spot as the first I lost the better framing of the first set. Still this came out pretty nice.

Update:

Based on feedback I've redone processing starting with HDRWT. It had pushed slight dimples in the big bright stars in the center. I created a coarse mask using the Range Selection tool and then used Atrous Wavelets to remove the first 2 layers to soften the edges. Using this inverted mask (zero to protect the brightest areas, white elsewhere) the HDRWT process did not dimple the stars. Keep in mind that this image will only show as intended in a browser with color management enabled like Firefox on Windows or Safari on Mac.

New version:

(click for 1:1 version)

The problem

When visiting astro photo related forums and while participating in various Yahoo discussion groups I often see questions related to getting the right colors or color balance in images. Generally this revolves around finding the right factors for either debayering or for combining color channels into a color image. Sometimes G2V stars or even color cards are suggested as sources of calibration data. These techniques can deliver decent results but are at best cumbersome and really don't get you all the way there. You still need to adjust color balance later during processing. So why go through all this trouble when a few simple assumptions can get you more than 95% of the way there?

The solution

PixInsight provides a number tools (processes in PI nomenclature) to make getting the right color easy. Here I show two of these tools: BackgroundNeutrlization (BN) and ColorCalibration (CC). This brief article does not explain these processes in detail but merely shows what can be done with even the default settings and not much effort at all. The image is a stack of 5 subs at 15m each. I intend to capture far more subs and fix other issues before processing it but it made a decent example. Please ignore any defects and noise you see. It's about color. The image is a 32b TIFF directly from DeepSkyStacker,  made visible by applying a ScreenTransferFunction (STF) in PI. STF does not modify the data but allows us to see 32b data on an 8b screen. BN and CC work best when applied on linear images so you are recommened to apply these early on during processing.

Here the 32b TIFF from DeepSkyStacker is loaded in PixInsight. It appears dark because of the huge dynamic range being compressed into 8b color.

Here an automatic Screen Transfer Function was applied. This does not change the data, just how it's presented on the screen. I drew a small preview in a dark area of the image that I expect to be just background without any nebulosity. The preview was chosen as the reference image for the Background Neutralization (BN) process.

After applying BN the colors look quite a bit better already.

Here I've added Preview02 around an area with stars that appear white. I then started the Color Calibration (CC) process and provided the previews as reference images and applied it. The change is subtle but worthwhile and low effort.

Notes

This was a really quick demonstration of how powerful these tools are even when used in default mode. With more careful adjustment of the parameters you can achieve even better results. After stretching and other processing you may decide to manually adjust the color balance a bit to your liking. BN and CC are not meant to replace artistic license but to give you tools that help get you in the right direction. No more excuses for purple backgrounds and colored stars. No struggling with G2V stars and trying to compensate for LP filters and spectrum response of your camera. Forget about atmospheric extinction. Use the image to calibrate itself using simple assumptions: the sky is gray (not black) and stars on average are white. This technique is a bit problematic when there is no visible background. In that case other techniques can be employed.