Starship Asterisk*

What causes the green, primarily in the upper left and lower right corners? Thank you.

I didn't process this specific image, but I am versed in Hubble image processing so I'll answer...

The short answer is that this is not a true visual color image.

Most Hubble images are captured using colored filters set up for scientific study, rather than for visual image representation (though the space telescope has visual filters as well).

You can see some of the specifics on the preparation of this image here:
http://hubblesite.org/newscenter/archiv ... fastfacts/

When people process the Hubble data into imagery, they assign colors to the data captured with various science filters, and those colors may not truly represent the color of light that was captured. So, for example, they may choose to represent one set of filtered data as red, another as green, and a third as blue. Here's a quote from the above page:

This image is a composite of many separate exposures made by the ACS instrument on the Hubble Space Telescope using several different filters. Two filters were used to sample broad wavelength ranges. One filter was used to sample narrow wavelength emission. The color results from assigning different hues (colors) to each monochromatic image. In this case, the assigned colors are:

F435W (B) blue
F555W (V) green
F814W (I) red

Note that the filter names (e.g., F814W) are slightly at odds with the text, insofar as all of the names end in 'W' - implying wideband, yet the text specifically states one filter was a narrow-band filter. Also note that while F435W transmits blue light and F555W transmits green, F814W does NOT transmit red light - it is an infrared filter, implying the red channel data in the image is actually invisible visually, and true red light was not captured. This can explain why this image does not look like most visual light astroimages.

Also not stated is whether the data assigned to each of the color channels has been normalized per the various exposure times and filter transmissivity values, so it may be biased toward one or another color. Such normalization is not trivial, and is often skipped during Hubble image processing because of the way the data is stored and retrieved.

One of the things that has always bothered me about images presented from the Hubble data is that the folks doing the processing don't make it clear what color mapping was done right out in the main press release, and so uninformed viewers often assume the Hubble colors are accurate, visually, when they are not. In this case, the color info is on their web page, but you have to dig to find it (and it's accuracy is suspect).

-Noel