It seems that every year I'm compelled to try to get an image of the Horse Head.  It's a detailed dark nebula with amazing complexity and stunning colors that's been imaged tens of thousands of times; but in my opinion, it is not an easy object.  First, there's the light from the nearby star, Alnitak, one of the bright stars in Orion's Belt.  With my former C14, the stray light from that star played havoc with stray light in the system.  I had to point in just the right place to make it work.  Since I've had the CDK20, I've tried the Horse Head multiple times, but I've never been happy with the quality of my data, so I never processed any of that stuff.  I must have 200+ hours of data sitting on various disks that just frustrated me.  Unless the raw data is good, I have little interest in spending time trying to process an image out of it.  This winter, I had the scope dialed in and the seeing cooperated a bit better so I gathered a lot LRGB + HSO data.  By setting a threshold of 2.1", the yield was a very acceptable ~50%  This image contains nearly 54 hours of LRGB data out of the maybe 200 total hours that I gathered.  The scope did produce a long faint diffraction spike running through the data from Alnitak but that was pretty straightforward to remove.

My first idea was to simply process a pure LRGB image and then to look at combining some Ha into the mix.  I also considered doing a full NB image just to see how it compared, but I put off both of those ideas for when I might have more time.  The LRGB data was pretty good with FWHM ranging from 1.4" to 2.1" and I started processing it with the idea that I'd quickly pull a nice image out of it all; but, not so fast!  This image is the first image that I've ever released showing 100% of what the full sensor sees--and as you can see, the field is relatively small compared to the entire nebula complex.  There are no regions within my FOV that represent an obvious "black point" and most of the field is flooded with red light from the hydrogen emission region above he Horse Head.  So, I did the obvious thing and tried to calibrate the colors in a LRGB combination using the SPCC tool in PI.

The first run produced a violet cast to the emission region, which clearly wasn't right.  Some of the bright stars also showed fairly obvious purple halos and it was easy to see that the result contained way too much blue.  I messed around with manual calibration and after a lot of fiddling, I arrived at something that looks closer to what I expected to see; but, whenever I have to do that much manipulation, I'm left feeling like I've missed something.  I went through 4 different versions experimenting with a lot of ideas including separating the stars from the background, fiddling with different color calibration schemes and messing with other methods to better show the details in the dark regions--and all of them produced different, unacceptable artifacts.  In the end, the biggest problem was that I never liked any of the colors!

This thing sat on my drive for a couple of weeks between iterations but I finally had time this week to sit down and give it my full attention. Here's how I finally cracked the color problem.  First, I created a straight RBG image and then experimented with how SPCC worked using different white references.  The obvious problem with this region is that light from most of the stars in the field pass through interstellar dust, which shifts the peak wavelength into the red due to the scattering of shorter wavelengths by dust particles.  Remember that the Harvard stellar classification starts at the very blue end (for the hottest stars) with type 'O' stars and works through to cooler temperatures with B, A, F, G, K, M, R, N, S type stars.  My original SPCC effort used the "Average Spiral Galaxy" white reference, which is great for calibrating most galaxies, but in this case, it's the wrong color reference...as my results demonstrated.  Juan and his team were fully aware of this issue when they produced the SPCC tool and they provided a long list of stellar types to use for white reference so I just had to find the right one--or at least one that would work.  I quickly discovered that using cooler stars for the white reference added more blue into the result.  Reds became violet boarding on violet and even purple around the edges.  Using hotter stars caused a stronger green mix that shifted reds to become more orange.  I've looked at a lot of images here on AB and from other imaging sites and I've noticed that all of these variations show up.  It is very hard to find one "standard" color pallet for the Horse Head among the hundreds of images that are easy to find!  I wound up using B5III stars (pretty hot stars) as the reference for this image.  I won't put it out there as the most correct balance but it produces strong crimson clouds behind the Horse Head along with a nice blend of blue and "dirt-colored, brownish" dust in the dark regions.  It also reveals how the blue reflection nebula is buried behind dust along with interesting filaments of reddish/orange emission regions that appear to sit in front of the blue reflection nebula.

Once I had the colors tuned to my satisfaction, I then processed the L channel in parallel with the RGB image.  Once I had everything where I wanted it, I combined the stretched RGB with the Lum data in PhotoShop.  I could do it in PI but this is one of those things that is just so easy in PS that I got lazy and did it there.  I alway agonize over saturation but I think that this object does well with bright colors.  This image has been subsampled by 0.5x from the original data.  There isn't much visual difference between the two and this prevents the image from being way too big to easily zoom in on here on AB.

As usual, C&C are welcome so let me know what you think of the colors...or anything else for that matter!

John