Slashdot Mirror


How Spirit Takes Pictures

Some Clown writes "MSNBC has a great article on the details of the camera system on the Mars Rover titled How Sprit makes great photos. Apparently the high resolution images are all done with a 1-megapixel camera. All the money is in the CCD and Lens. The hardcore digital photographers in the crowd will probably find the article to be only a teaser on the technical specs, but the rest of us in the unwashed masses should find it interesting."

20 of 311 comments (clear)

  1. I was honestly surprised. by ActionPlant · · Score: 1, Insightful

    The Spirit being from NASA, I was assuming it had at least 6MP cameras. This really is pretty cool. Perhaps I'll dust my old 1MP camera off and see if I can do anything similar. If nothing else, they've proven that it's not completely worthless yet. Pretty nifty.

    Damon,

    --
    http://actionPlant.com
    1. Re:I was honestly surprised. by Mod+Me+God · · Score: 5, Insightful

      It is interesting, but I suppose it is explained because this sort of tech has to be (physically) tested a couple+ years in advance (and 2 years is in the final test/construction phase already, initial planning was years before), so strapping the latest untested camera in to a planned and tested system would potentially introduce a weak point in the system. Yeah, pretty nifty, kinda shows the potential 'old' hardware has when used to its full potential.

      --
      --

      FreeNET user? Comfortable with the adverse selection?
    2. Re:I was honestly surprised. by Fr33z0r · · Score: 3, Insightful

      I'd imagine it's more to do with radiation hardening, they could have, for a paltry sum, put a top-of the range con/prosumer camera up there and if it broke, no big deal, but it wouldn't survive outside the protection of Earth's atmosphere, the smaller alectronics would be fried by radiation from the sun in no time. Radiation hardening is great, but a difficult, complicated, "fiddly" procedure, so basically all the tech on the shuttles are a lot less advanced than the stuff we take for granted here on Earth. They wouldn't survive space if they weren't.

    3. Re:I was honestly surprised. by FrostedWheat · · Score: 2, Insightful

      I'd say the biggest limit on the size of the images is the bandwidth to Earth. The links are getting faster but still nothing compared to a typical DSL line.

      One thing I've been wondering, and maybe someone out there knows more. What kind of image compression are they using?

    4. Re:I was honestly surprised. by joggle · · Score: 4, Insightful
      As far as mission-critical micro-electronics (like CPUs or CCDs), bigger is better for radiation-resistance. That's why many spacecraft still use 286 - 486 chips because greater speed isn't needed, are cheaper to radiation-harden and are less complicated (harder to break) than new chips. In the case of CCDs, it is mentioned in the article that lenses need to be created with greater precision for high-resolution CCDs than lower resolution ones. I'm sure this wouldn't be a problem for NASA, so I would guess they went with the lower resolution CCD due to the larger size of each sensor and because it met the mission requirements. They don't mention this in the article, but the rover is very bandwidth-limited, so it wouldn't be possible to send back any more information than it already is anyways.

      They mentioned that the design process of the Huble's CCDs at a resolution of 800 x 800 contributed to the current mass production of consumer CCD cameras, so I don't think they are afraid of pushing the envelope if it is needed to meet mission requirements.

    5. Re:I was honestly surprised. by Ianoo · · Score: 2, Insightful

      Whatever it is, I'm pretty sure it's lossless. The pictures (no matter how snazzy they look) are intended for scientific research. There wouldn't be much point spending millions on optics to ruin pictures with decompression artefacts.

      As for the actual encoding -- considering the article states that the cameras don't work like normal cameras and instead red, green, blue components are built up separately -- I'd say it's something NASA cooked up just for these probes.

    6. Re:I was honestly surprised. by joggle · · Score: 2, Insightful

      I know it is a lossless format, so it obviously isn't JPEG. I would guess that they are using some form of RLE (run-length encoding), perhaps combined with a Huffman table for the shading values. I wonder what the pixel-depth is.

  2. Interesting, but.. by xankar · · Score: 5, Insightful

    High quality images are good for PR, but what I really want to know is how it extracts information from the environment, how this information is being used, and whether or not we found anything we didn't expect to find.

    --
    ~To choose doubt as a philosophy of life is akin to choosing immobility as a means of transportation. -Yann Martel
  3. $400,000,000? by poppageek · · Score: 4, Insightful

    I am not trying to be negative, I think what they are doing is great and long overdue. Can't wait till we have Rovers on other planets. But why did it cost $400 million? I've read about what Rover is and how it was built and what it does. I am sure it was expensive to build but $400 million? Does that include the cost of getting it there?

    1. Re:$400,000,000? by pvt_medic · · Score: 2, Insightful

      its not only the cost of building such an item but the process of getting it certified for such task. You not only have to build it, but then have teams of review committies look over it again and again, so you dont have something stupid like a conversion error.

      --
      30% Troll, 50% Underrated, 10% Interesting
      Score:5, Troll
    2. Re:$400,000,000? by Erwos · · Score: 2, Insightful

      That's including the R&D costs. Ergo, the next rover will be far cheaper, because they've just got to build another one, not figure out how to make it in the first place.

      -Erwos

      --
      Plausible conjecture should not be misrepresented as proof positive.
    3. Re:$400,000,000? by ad0gg · · Score: 5, Insightful

      Expensive compare to what? B-1 Bomber is 1.2 billion a piece, F22 is $122 million. Communication sattelites can range from $100 million and up. And the R&D costs can spread accross multiple units, rovers only had two units to spread across.

      --

      Have you ever been to a turkish prison?

    4. Re:$400,000,000? by Bigfishbowl · · Score: 3, Insightful

      Well, look at it this way . . . if you break down the cost across the entire population of the US, that amounts to about $1.37 each (compared to the $97 billion = $297.95 each for Iraq). So for less then the price of a beer at a bar, I get to see Mars. Works for me.

  4. Re:Pictures by nucal · · Score: 5, Insightful

    Actually, it would be fantastic if all Spirit's pictures had a finger in the bottom corner!

  5. Lens by dgerman · · Score: 4, Insightful

    "NASA's Spirit Rover is providing a lesson to aspiring digital photographers: Spend your money on the lens, not the pixels."

    Every good photographer will tell you the same. It still amazes me that people are willing to drop Can$.5k for a digital camera, but think you are nuts for spending the same money in a lens.

    Too bad the digital cameras all come with Zooms. At the same price, a zoom lens will tend to be worse than a fixed lens. An old camera, the yashica t4 super won a great reputation for its superb fixed lens (35 mm Carl Zeiss).

    I have one, and I love it. It takes the best pics I have ever seen in a P&S.

  6. Spirit camera in effect 3+ megapixel by mc6809e · · Score: 2, Insightful

    The article implies that the camera is monochrome and that filters are used to capture each color.

    So, adding the images together, 1 megapixel green + 1 mp red + 1 mp blue = 3 megapixels.

  7. Keep in mind... by MrScience · · Score: 2, Insightful

    Most digi-cams say that they are 3MP, but keep in mind that for any given pixel requires four elements (RGGB) to create. I believe the Spirit camera is only sensitive to light, and has interchangable filters (so it must make three passes to get full color) -- effectively tripling the "element count" of the sensor.

    --

    You quitting proves that the karma kap worked. The most annoying of the whores shut up. --CmdrTaco

  8. you don't need gazillion megapixels ... by porky_pig_jr · · Score: 2, Insightful

    to produce decent digital pictures. In fact, having very large number of pixels introduced lots of noise. The latest Sony camera - 8M pixels - is a good example. The camera simply isn't good. High level of noise, and color abberations. They've crammed too many pixels in CCD with the area too small.

    OTOH, the high quality lenses and high quality post-processing of captured image are important factors in getting decent digital pictures. Yet 'unwashed masses' only understand one thing - the magic pixel number.

  9. Re:Here's what surprises me... by vicparedes · · Score: 2, Insightful

    Maybe because spending millions of dollars on a space mission entails rigorously testing every piece of equipment which could take years before the launch? And as the article pointed out, this is not your off-the-shelf digital camera so an extra-thousand bucks will probably get you nowhere near the quality you'd otherwise expect from off-the-shelf accessories.

  10. Re:Nonsense... by Skipio · · Score: 2, Insightful

    "Assuming the 7MB image is the raw output of the CCD, that gives 56 bits per pixel brightness. That is, each grayscale bit has 56 bits worth of information.

    Not that I'm saying that's what is actually going on, just that you shouldn't expect a multi-million dollar camera to stick with 8-bits per pixel. In order to get as much information as possible (including being able to use various filters to their full effect) you would want as many bits per pixel as possible. Probably one reason the CCD elements on the camera are bigger than consumer models - more light, more differentiation between different states, and more information gained.
    "

    Duh, the maximum bit depths one would ever use would be 16bits per pixel. An ADC that could output 56bits would be too expensive and absolutely useless as the signal to noise ratio on small CCDs is just too low for an ADC of better resolution than 16bits to be of any use. You'd not gain any more signal - just more noise.