Home » Color in the Deep Sky -- even under NJ Skies?
Color in the Deep Sky -- even under NJ Skies?
The beautiful intense colors in astrophotos of the Messier and other deep sky objects, with glowing reds and blue-greens emitted by vast ionized hydrogen and oxygen gas clouds among the stars, have become familiar sights on NASA and Hubble internet sites and in those big glossy coffee table books. Galaxies also emit intense light energy in the violet, blue, and red bands due to numerous regions where young stars are being born (the H-II regions). But the deep sky color pallet is actually quite subtle and almost never visible in a telescope eyepiece, due to the limits of human eye physiology and the faintness of the light captured by the telescope. So we rely on CCD imaging techniques which fortunately have advanced remarkably over the past 15 years or so.
Experts, often working in high altitude desert locations with very dark skies, usually approach color astro-imaging using the “LRGB” method. Here many subframes are taken over the course of several hours with a high sensitivity monochrome CCD through a telescope using a sequence of color filters (Luminance, Red, Green, Blue = LRGB). The individual filtered subframes are then calibrated and combined and balanced in the computer to create the final color images. But here in central Jersey our less-than-pristine skies introduce serious challenges even with top-notch equipment. Moon and skyglow backgrounds along with light pollution gradients conspire with changing weather, humidity, and seeing conditions over the course of a single New Jersey night, often introducing nearly intractable noise and light gradients in the LRGB-derived images.
But there is hope for northeast US astronomers! Whether you are considering astrophotography, or are already on the learning curve, you don’t necessarily have to buy that remote property or telescope time-share in the desert southwest. After more than 12 years using the LRGB imaging method, I began to consider whether a “one-shot color” astronomical CCD camera would potentially produce better results under New Jersey skies. The much reduced hardware complexity and lower weight and telescope balancing issues with one-shot color cameras add to their potential attractiveness over the LRGB technology. The one-shot color approach uses a CCD sensor with a Bayer-matrix of RGB filters directly on the sensor itself (like terrestrial DSLR cameras). Once acquired the images are “de-Bayered” to convert to color using software after many subframes are calibrated and stacked. The idea of using a one-shot color camera disputes much of the expert advice out there, which warns that these cameras are unsuited to light-polluted areas. I wanted to test whether this was true, and theorized that the ability to capture all colors simultaneously in each subframe could actually minimize the noise/gradient issues which change over the course of an imaging run here in New Jersey. Further, it is proposed that the final resolution could equal that of the LRGB method if the image scale is appropriately selected (that is, CCD pixel dimensions must be carefully matched to telescope focal length).
Below are final images of the same object, the spiral galaxy M106, created using each method with the cameras indicated. The sky and moon conditions were similar and total exposure time was the same for each, 4.5 hours. The CCD chips of the two cameras have similar pixel sizes but different total pixel number and sensor area, so their fields of view differ, and the resulting images were not cropped. While other celestial objects may compare differently, it’s pretty clear that for a target with low surface brightness like a spiral galaxy (magnitude spread out over a wide angular area), the beautiful blues and reds are more intensely captured and better balanced by the LRGB than the one-shot color camera. While that may not be true from a desert mountain site, it does appear so here in central Jersey. More comparisons will be made in the future to see how magnitude, surface brightness, and type of celestial object affects results.