As a way to improve my processing skills and to keep in practice during when I don't have any of my own data to practice with, I have recently started processing Hubble Space Telescope raw image data from the HST's publicly accessable archives. To view my attempts at processing this wonderful data please click on the links below. Although my processing is not as good as the professionals, I hope you will be able to see my skills improve over time.
Each of the Hubble Telescope Images on this page was assembled by downloading the publicly available original calibrated data from either the Hubble Legacy Archive or the MAST archive. Search tools on both sites allow you to find the object you are interested in and then, if the object has been imaged by Hubble, the data can be downloaded.
In general the Hubble telescope is not used to take "pretty pictures". The Hubble Telescope datasets that are used to create the images on this page are acquired for scientific purposes and the filters that are selected and used by the investigators are intended to allow the researchers to investigate certain features of the object being photographed, such as the chemical composition of a nebula by looking for emission line spectra charactistic of a certain element. However, sometimes those datasets can be combined in ways that yield amazingly beautiful false color images. That is what I am attempting to do here.
All of the images on this page started as two or three greyscale images. Sometimes the image I am interested in has been imaged using three or more filters. If so, then I can combine the three images as a false color (usually) image by assigning the filtered images to red, green and blue. It is quite common for the Hubble data to be combined in a way that does not represent what the object would look like "naturally" if viewed in space from a theoretical spacecraft.
Sometimes the data I am interested in only consists of two filtered greyscale images. I can either assign one filter to be the red component of the final image, another filter to be the blue component of the image and then have the software blend the red and blue component to create a synthetic green component. Another option is to assign one of the filters to red and the other filter to green and to blue. It all depends on how the various combinations look when combined. Its all subjective and depends on what seems to be the best combination to bring out the interesting visual details of the object.
The Stephan's Quintet image linked above is an example of a three filter image whose components were assembled using visible light and non-visible light components. The image was assembled using the Hubble wide band filters for the ultraviolet spectrum, the visible light spectrum up to 700 nanometers, and the red-infrared spectrum. The blue in the Stephan's Quintet image is represented by data captured through the 438w filter [~240 nanometers to ~310 nanometers]. Light in this wavelength is outside the typical human visible response spectrum, which typically is 390 nanometers to 700 nanometers. The green component of the Stephan's Quintet image is represented by data captured through the 606w filter [~475 nanometers to ~700 nanometers] which is essentially the entire range of the human visible response spectrum from blue, to green and red. The red component is represented by data captured through the 814w filter [~700 nanometers to ~1000 nenometers] which is deep red and into the infrared range. Using the filtered data in this way allows us to visualize the object in such a way as would not be possible using light exclusively from the visible spectrum.
Samples of the three filtered images are shown below. Each of the three filtered data images starts out as a greyscale image. Using the image processing software I assign each of the three greyscale images a color - either red, green or blue. When the software combines the color mapped images the result is a false color image and the combinations of the different intensities of red, green and blue create the "color" in the image.
Another example is the Messier 16 image linked above. The data used to create this image was captured through narrowband filters - 673n, 657n and 502n. These filters are designed to capture data at the emission lines for Sulfer II [S II], Hydrogen alpha + Nitrogen II [N II] and Oxygen III [O III] respectively. The emission lines for S II, H-alpha and N II are all in the red portion of the visible light spectrum. However, by assigning red to S II and green to the N II and H-alpha components, we can visualize these emissions as separate colors. The emission line for O III is in the green portion of the spectrum. Thus, the M16 image has been combined by mapping S II to red, H-Alpha + N II to green and O III to blue.
I use the PixInsight program to do the color assignments, color combination, and some basic color adjustments. I then export the image to Photoshop where I make the final tweaks.