Hubble Telescope's Fantastic Images? Photoshop!

<p>This <a href="http://www.slate.com/id/2227828/">article</a> fascinated me, and I thought immediately of the Photo.net Discussion Boards as a forum for sharing it. Apparently, even the most scientifically responsible images involve some subjective interpretation, employing Photoshop. (But then, what was I expecting - analog?) I'd be interested in your reactions to their process of arriving at these stunning final colors.</p>

<p>I would give those shots a 4/5, but a 3/3 anonymously ;)<br>

For the post production work, a definate 6/7</p>

<p>Many of the shots are color enhanced for interpretive reasons, not just make them pretty. Things like certain element emissions in particular colors. Even going back to the days of the first Explorers and all, I think that NASA has generally been explicit about images not being "true color". But some of them are so really beautiful, all the same.</p>

<p>The <em>Astronomy Picture of the Day</em> is a wonderful site for astronomically related images (<a href="http://antwrp.gsfc.nasa.gov/apod/astropix.html">link</a> ). My desktop images are usually chosen randomly from files of these images.</p>

<p>Many of them are 7/7 for me.</p>

<p>As JDM indicated, it's really mostly science behind the manipulations. Most images like this, amateur and professional, are actually a series of images shot through narrow bandpass filters that isolate scientifically important wavelengths, like the wavelengths for hydrogen alpha and beta, oxygen III and sulfur, plus an unfiltered luminance image. The individuals images are shot in black and white, and multiple images are stacked to create the final, with colors assigned to each base image. So you're not even getting a continuous representation of the actual spectrum, just the scientifically interesting wavelengths. The colors are generally chosen to match the color of the wavelength, but it doesn't have to be that way, and frequently isn't in images used to study the surface composition of the moon and other solar system bodies.</p>

<p>The big debate in terms of accuracy would probably be over the brightness of the final image versus the object in real life, but again, the images are used primarily to study the objects, and you need to be able to see what you are studying.</p>

<p>Regardless, I find them great to look at - the upgrade Hubble images are awesome.</p>

<p>Visualization of scientific data is done all the time. Pseudo-color is one of the most powerful tools to let the human visual system interpret features that might be too subtle otherwise. Sometimes you do it just for aesthetic reasons, too...</p>

<p>My first major publication was a visualization of blood-flow from an MRI image-set for the Oct. 1990 cover of Science magazine.</p>