Hi folks,
Nope, it's not something I captured unless you count logging on to the
Hubble Legacy Archive and downloading a number of images of the Ant Nebula taken through various filters. My intention was to have a bit of fun but also have a first go at combining narrowband data rather than, as I have up to now, just combining H-alpha and RGB data. Here's the result (you can read all about the object
here):
The chosen images (a subset of the narrowband images
here) all had their stars removed so I could have free rein on colour choices. A slightly unconventional palette was used but I started off straightforwardly enough with H-alpha (656 nm) represented by a mix of 100% red and about 25% blue. I could have used H-beta imagery (487 nm) directly for the blue but after I downloaded it I could see no structural differences and the H-alpha data looked cleaner. NII (658 nm) was added next as 100% green but via a curve which strongly de-emphasised the shadows and mid-tones. OIII (502 nm) came next, pure blue but with just a slight de-emphasis of the mid-tones. SII (673 nm) was the final addition so far as nebulosity was concerned, this by mixing red and green in the proportions of 4:1. Finally some stars were added from the SIII (631 nm) narrowband image as this had minimal nebulosity which was easily removed but a good clean star background. Each contributing image was cleaned and enhanced as best I could before being added to the stack and I also took the opportunity to use the available higher resolution H-alpha and NII images (WFPC2-PC).
As I say, this is my first attempt at this sort of compositing and it's fantastic to have free access to the Hubble Space Telescope data for practice. In an ideal world our eyes would have separate photo-receptors for more than three colours which would have made colour choices easier. That said, any attempt to go photo-realistic would have been doomed to failure due to H-alpha/NII proximity as well as the fainter emission lines being drowned by the brighter ones, not to mention my inability to decide from the Hubble data exactly what the relative brightness actually are! In the end, having started off with a natural palette for H-alpha (plus simulated H-beta) my choices of both hue and intensity were driven entirely by aesthetics, including strenuous efforts to keep the H-alpha emission towards the red end of the spectrum despite a push from my NII green colour choice driving the end result towards yellow, plus a strong desire to show structure. I hope you enjoy the result: comment and criticism is welcome.
Bob.
Login
Register
FAQ
Search
