Slashdot Mirror

← Back to Stories (view on slashdot.org)

The Real Reason why Spirit Only Sees Red

Posted by CmdrTaco on from the now-thats-interesting dept.

use_compress writes To produce a color photograph, the rover's panoramic camera takes three black-and-white images of a scene, once with a red filter, once with a green filter and once with a blue filter. Each is then tinted with the color of the filter, and the three are combined into a color image. In assembling the Spirit photographs, however, the scientists used an image taken with an infrared filter, not the red filter (NYTimes, Free Registration Required). Some blue pigments like the cobalt in the rover color chip also emit this longer-wavelength light, which is not visible to the human eye."

13 of 273 comments (clear)

  1. Was in New Scientist a week or so ago by Space+cowboy · · Score: 5, Insightful

    The reason being that the science gets better results using th e IR filter than if the red filter were used... At the moment, despite great public interest, the science is more important... that IS what it's there for....

    Simon

    --
    Physicists get Hadrons!
    1. Re:Was in New Scientist a week or so ago by mcbevin · · Score: 5, Informative

      If you read the whole article, you'd see that they actually used both the infrared AND the red filter for the pictures. So they had their infrared for their science as well as the red for the photos to show the public. However they mucked up in producing the photos for the public, using the infrared instead of the red. Nothing to do with science vs public interest, rather a simple mistake.

  2. Cue a thousand alien-watcher website updates.. by Channard · · Score: 5, Funny

    'Aha! So that's why they don't see little green men...' - at last, the dream of aliens living on Mars is alive again.

  3. obligatory registration free link... by corsetboy · · Score: 5, Informative

    How the Red Planet Came Down With the Pink Blues

  4. prepare the servers for a nytimes-dotting... by mobiux · · Score: 5, Funny

    They mention slashdot.org by name.

    Could this be some sort of revenge?

    1. Re:prepare the servers for a nytimes-dotting... by Asprin · · Score: 5, Funny


      Quick! Put up the free registration page!

      --
      "Lawyers are for sucks."
      - Doug McKenzie
    2. Re:prepare the servers for a nytimes-dotting... by lukewarmfusion · · Score: 5, Funny

      Then NYTimes will have to link to the Google cache, someone will copy and paste the unformatted text on their site anyway, and we'll see a plagiarizing reporter trying to karma whore in order to get his job back.

      Of course you know.... this means war.

  5. Why b/w & filter? by Lolaine · · Score: 5, Interesting

    Can anyone explain why 3 separate B/W images are taken? If it is because of bandwidth... 3 grayscale images weights (more or less) like one color image ... so why B/W and filters?

    --
    ------- The last Sig. got fired.
    1. Re:Why b/w & filter? by herko_cl · · Score: 5, Informative

      Sorry to say this, but the parent in NOT "Informative". The only sensors with 3 photosites per pixel are Foveon's. The vast majority of digital cameras has ONE photosite per pixel, and a Bayer mask (RGB filter) layered on top of it. Pixel color in the final image is then interpolated from the measured intensity of the three adjacent photosites. Yes, this means that digital cameras have higher Luma resoultion than Chroma. No, it does not matter much, because the eye is much more attracted to Luminance detail.
      Almost all of the manufactured sensors are black and white; only Foveon's are 3-color, and they're expensive for the resoultion and the first generation software had color clipping problems (overexposed areas of images went abruptly to white). This has apparently been fixed.
      A monochrome sensor with external filters is much more flexible than the single-duty Foveon, so I guess that's why they chose it. Also, NASA doesn't usually buy space-faring hardware off-the-shelf two weeks before launch, and this full-color sensor simply did not exist a couple of years ago.

      --
      No .sig for you! ONE YEAR!
    2. Re:Why b/w & filter? by vofka · · Score: 5, Informative

      A broadcast television camera (which is really pretty low-resolution, unless it's a true HTDV camera) has three CCD sensors mounted to a prisim block that splits the image into the three component colors for television (RGB). The use of three CCDs for television is necessitated by the fact that the desired result is a color image without waiting to assemble a color composite from three black and whites. Broadcast television results in images that are pretty close to 640x480 (again, prety low res).

      Close, but not quite...

      A Broadcast Quality camera is usually capable of recording a substantially higher resolution of image than is eventually broadcast. This allows for much better editing facilities later on - ie. Cropping and resizing of the recorded images without loss of detail in the later broadcast. Final Broadcast (in the UK at least) is around 760x575 pixels (actual broadcast lines are 625, but several are taken by the Vertical Blanking Pulse, the Frame Field Markers and Teletext data) - but the camera definately records a much higher resolution than that.

      For comparison, a standard Hi-8 Domestic Hand Camera records around 540 picture lines (about 720x540), and the picture quality from this kind of camera is much lower than that needed by the broadcast editing suites to work effectively - just watch any "home video" programme (such as "You've been Framed!") for proof!

      Also, expensive professional broadcast cameras use "Dichromatic Mirrors", not prisms to do colour seperation. Prismatic seperation would lead to too much signal loss and colour bleed accross the image. The first mirror directs the Red image to the appropriate sensor, and also allows enough light of all wavelengths to pass to the next mirror, where the Green image is diverted to the appropriate sensor, and again, light of all wavelengths passes to the final sensor in the camera. Blue is never explicitly seperated from the incoming image, but is instead inferred from the intensity data from the three individual sensors.

      I can be very certain of both of these facts because my dad was a Video Electronics Engineer for the BBC for a number of years...

      --
      Disclaimer: I meant what I thought, not what I wrote! What? You can't read my Mind? Oh dear!
  6. Re:Why don't they... by Carl+T · · Score: 5, Interesting
    [...] just use a 4 megapixel digital camera that anyone can buy from Compusa

    Quite possibly because it wouldn't survive the conditions on Mars. Or on the way there. Try deep-freezing your digital camera, then put it in a vacuum chamber, then in a really dusty sandbox, and finally subject it to a potentially lethal (for a human) dose of radiation, and see if it still works. Oh, and don't forget simulating the landing; heat it, vibrate it, and toss it on the ground.

    Disclaimer: I wasn't there. I don't know exactly how the poor thing was treated. I'm not a member of the PETC (People for the Ethical Treatment of Cameras).

    --

    This signature is not in the public domain.
  7. But what is this thing? by tjmcgee · · Score: 5, Interesting

    The martian crab http://homepage.mac.com/thomasmcgee/ I know, I know, go ahead, mod me off topic. The truth is out there. Would anyone like to start a petition that requests NASA to try to get one more photo of this thing before they drive away?

  8. Re:Why not just give NEW pictures! by daina · · Score: 5, Interesting
    It is a misconception that you can use Photoshop or some other image processing program to produce a "true colour" or RGB image when one channel represents infrared data. Here's why:

    If you use an infrared filter like the L2 filter on Sprit's Pancam, you get data that represents only things which reflect or emit light in that particular region of the spectrum. Anything that emits light ONLY in the red will be absent from the data set. It is possible for something that appears as a fairly monochromatic red to be entirely invisible. How can you use Photoshop to put back something that is invisible? You cannot.

    You can adjust an individual colour in the image using a reference image taken with the appropriate filters, and that colour will then appear correct. Other colours, however, will remain distorted.

    Worse, you cannot possibly know the emission/reflectivity spectrum of things on Mars, so any image you produce that appears to show the sundial colour chips correctly may distort terribly the Mars components of the image. It is not really very interesting to see a colour corrected photo of the sundial, is it? We could have achieved that without sending the rover all the way to Mars.

    Nope, using a relatively narrow-band-pass infrared filter like the L2 simply leaves out information about the red part of the spectrum, and extrapolation only goes so far in recreating that data. Non-linear data - discontinuities within the missing portion of the spectrum - are simply gone, never to be retrieved.

    Also, NASA is lying. Perhaps 'lying' is too strong a word, but they are either deceiving us or they are operating under a serious misconception.

    "We just made a mistake," said Dr. James F. Bell III, the lead scientist for the camera. "It's really just a mess-up." Well, NASA claims to be releasing the raw data from Spirit on its web site, but the raw data does not contain any image sets for the panoramas taken with the L4, L5, L6 filters. They have almost never used the L4 filter.

    So either the "mess up" is that they have forgotten to use the L4 filter from day one (unlikely, since each photograph taken presents another opportunity to switch to the L4) or that they have L4 images but they are not releasing them, in which case they really are not releasing the raw data.

    The argument about the L2 being better for science is bogus. There's no way that NASA scientists are doing serious mineral analysis with a pretty, stitched-together wide view panorama. That's just rubbish. they would be looking at detail images, and possibly comparing between detail-level images. The panoramas are strictly for public consumption, and maybe office posters at JPL.

    It's probably not a conspiracy, but it is a mystery.