28
$\begingroup$

Like many others, I have marveled at the images made available from the Hubble Space Telescope over the years. But, I have always had a curiosity about the color shown in these images. An example is shown below. Are the colors we see, such as the yellows, blues, and so on the true colors or are they applied by some kind of colorization method to enhance the image quality for realism.

Hubble Image

$\endgroup$
  • $\begingroup$ A good source for this image will tell you more about it: hubblesite.org/newscenter/archive/releases/2009/05/image/a - for example that it was taken with the Wide Field and Planetary Camera 2 (WFPC2) which could use 48 filters, 3 of which were used here. hubblesite.org/the_telescope/nuts_.and._bolts/instruments/wfpc2 $\endgroup$ – JulianHzg Aug 3 '15 at 15:32
  • 2
    $\begingroup$ There was a fantastic von Kármán lecture (about an hour long, but it's great) about how calling things "true" or "false" color is an iffy question. The presenter also discusses what goes into the creation of these images. $\endgroup$ – Nick T Aug 3 '15 at 19:23
  • $\begingroup$ In addition to the answers you've received, we also do this in amateur astrophotography; it's called "narrow band imaging". The idea is to capture light at a specific wavelength (using a filter) as a B/W image. Then map that image to any color you like, and combine with others. The more wavelengths you capture, the more colors you can map to. Some images use two wavelengths that are "true" red (in that they are in the red/infrared part of the spectrum), but one mapped to blue and the other to green (as an example) to make the image more appealing. OIII, SII and Ha are the most common filters. $\endgroup$ – coblr Aug 3 '15 at 21:38
  • $\begingroup$ If you look at the fastfacts tab for the image you will see that this was photographed with the filters for OIII, Ha, and NII and then mapped into the colors for 658nm (NII) to red, 656nm (Ha) to green, and 502nm (OIII) to blue. There's only a 2nm difference between the NII and Ha lines while they're being actually remapped to colors that are fairly high contrast. $\endgroup$ – user20936 Aug 3 '15 at 22:27
31
$\begingroup$

Sort of.

As Space.com writes,

The raw Hubble images, as beamed down from the telescope itself, are black and white. But each image is captured using three different filters: red, green and blue. The Hubble imaging team combines those three images into one, in a Technicolor process pioneered in the 1930s. (The same process occurs in digital SLRs, except that in your camera, it's automatic.)

Why are the original images in black and white? Because if Hubble's eye saw in color, the light detector would have to have red, green and blue elements crammed into the same area, taking away crucial resolving capability. Without those different elements, Hubble can capture images with much more detail.

As an interesting aside, the Wide Field Camera 3 sees in wavelengths other than visible light, as do the Cosmic Origins Spectrograph and the Space Telescope Imaging Spectrograph.

NASA goes into a litte detail about the process here, as well as some of the rationale behind choosing some colors. Some of the reasons for using artificial colors include showcasing elements whose emission lines are out of the visible spectrum, and showing features that are too dim at visible wavelengths. Remember, CCD detectors usually don't see the same things that humans do, and Hubble can see outside the visible spectrum.

$\endgroup$
  • $\begingroup$ That explanation for "why are the original images in black and white" seems a bit odd. CCDs are inherently greyscale, and the as-yet-unattained holy grail of scientific imaging is for every pixel to report not just intensity but wavelength for every photon. As the previous paragraph alludes to, all digital cameras are also greyscale, but the camera automatically combines the filtered images according to the manufacturer's preset approximation for human vision. $\endgroup$ – user10851 Aug 3 '15 at 21:20
  • $\begingroup$ @ChrisWhite I added in some more information; there might be more in the last linked article that could help. $\endgroup$ – HDE 226868 Aug 3 '15 at 21:28
  • 3
    $\begingroup$ Isn't this also basically how our eyes work? Light intensity is always a one-axis system. You can only pull out colour by simultaneously capturing an image with varying wavelength filters (e.g. the rods & cones in our eyes). The difference is in the definition of "simultaneously" being used, but I still think we can safely call this "true-colour". $\endgroup$ – Lightness Races in Orbit Aug 4 '15 at 10:57
  • 1
    $\begingroup$ "than they would appear to the human eye" Arguably, we have no idea how anything appears to the human eye other than to our own. :) $\endgroup$ – Lightness Races in Orbit Aug 4 '15 at 10:58
  • 1
    $\begingroup$ @shrx Thanks; I was able to replace one of them, but the other seems gone for good. $\endgroup$ – HDE 226868 Jul 20 '18 at 16:08
13
$\begingroup$

The classic color mapping for Hubble is described in Flase-color astrophotography explained.

What you have is (in the Hubble palette):

Line   Freq      True     False
Ha    (656.3 nm) Red   -- Green
S-II  (672.4 nm) Red   -- Red
O-III (500.7 nm) Green -- Blue   

An example of this for true color from John Nassr at Stardust Observatory at Coming to Life Again…. Hubble’s 25th Anniversary Honored with New Hi Definition Photo of Pillars of Creation in the Eagle Nebula:

And false color from wikimedia commons (the pallet that of the Hubble)

You will note that the true color image is dominated by red - which is the Hydrogen alpha, and Sulfur-II lines (which show up as reds, greens and yellows in the false color).

The colors in astrophotography are specific remappings of frequencies into other frequencies so that our eyes are capable of perceiving the subtle differences between different emissions from elements. Otherwise, we have difficulty seeing the structure that is there and indicated by different elemental densities in the nebulae.

I will point out that the H/S/O palette given above isn't the only one. The image in the question has a link to the fastfacts tab which states:

The image is a composite of separate exposures made by the WFPC2 instrument on the Hubble Space Telescope. Three filters were used to sample narrow wavelength ranges. The color results from assigning different hues (colors) to each monochromatic image. In this case, the assigned colors are:

F658N ([N II])  red  
F656N (H-alpha) green  
F502N ([O III]) blue 

In that image, the doubly ionized nitrogen is used as red rather than the doubly ionized sulfur in the Eagle Nebulae picture.

$\endgroup$
6
$\begingroup$

I find myself now answering my own question but only because the comment feature is not suited to this "comment". I have selected the answer by @HDE 226868 as my answer and primarily due to the linked Space.com reference. Very good answer to my question.

In particular, I also thought this quote from the same page as being important as these reasons (below quote) for adding color were more of what I originally thought they were doing:

The Hubble team uses color in three ways. First, for objects that would otherwise be too faint for the human eye to see,the team adds color to make the objects visible. Second, the team uses color to depict details that the human eye can't see, like astronomical features only visible in infrared or ultraviolet light. Third, color can highlight delicate features that would be otherwise lost.

$\endgroup$
  • 3
    $\begingroup$ Similar issues come up in medical imaging. A CRT often has 4096 gray levels, way more than the eye can resolve. It can contain detailed information about density variations in bone and lung. This is handled with brightness and contrast. If you adjust so you can see bone details, lungs appear pure white. If you adjust so you can see lung details, bone appears black. Color would add more options to show both. Besides, NASA likes to put pretty color pictures on its website. It keeps the public interested. $\endgroup$ – mmesser314 Aug 3 '15 at 18:28

protected by Qmechanic Aug 4 '15 at 15:04

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.