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."
It is also interesting to see how it produces color photos. Instead of using a 3 color sensor, it uses a B&W camera with 3 colour filters that recombine into a colour image. This is calibrated by a colour wheel on the rover itself.
Neat stuff
I didn't think it would be 1MP camera either..
:)
How many people watched the Nova show about
Spirit? It was absolutely fascinating how many
obsticles have to be overcome and how much it
must cost (yep, that's where the $400 Mil comes
in, I suppose...). At the end, it really gives
you a good feeling when you finally see the rocket
liftoff.
No, I read it; obviously the sensors in my little 1MP cam won't be anything near what's in the pancam, but I can do something about the lens. Grab an eye doc friend who can get some decent prescription-ground lenses and go for the stereo effect. I'm not sure; one might have to write a small program to make it work, but it could be fun to see the results.
The NASA guys had to start somewhere. Their biggest advantage will be the sensors, but there's no reason we can't replicate the rest. If one wanted to go all-out, it might even be feasable to use an array from a high-mp camera and configure it to use multiple sensors to produce 1px.
Damon,
http://actionPlant.com
That's because they did it twice, the second rover lands a week from saturday.
And the 400M (each) includes all of the research, developement, construction, launch costs, operations, radio telescope time, etc until the end of the mission.
Nasa uses a 1MP camera but 3 filters. The 3 filters are combined to give you a color image.
Consumer 1MP have the same number of pixles buy each pixle has only 1 filter. The image is then interpolated to get a true color image.
Even if your optics and ccd were the same quality as Nasa's you would still only have 1/3 the resolution.
Serious question here... the article says "One megapixel is a million pixels set up in an array equal to 1,000 by 1,000."
Is this like hard drives using one GB = 1,000,000,000 bytes or is 1MP truely 1,000 x 1,000 and not 1,024 x 1,024?
------
"And may your days be long upon the earth."
remember how you can build a 1 gigapixel shot out of a bunch of lower resolution pictures.
30% Troll, 50% Underrated, 10% Interesting
Score:5, Troll
That's not all- the images are clearly composited, which is why they look so stunning(yes, the huge, low-noise ccd helps, as does a great lens). The very first image released(the 8mpixel one) had a very very obvious stitching error right smack down the middle, which is pretty bad, considering that with a robotic rig and known lens characteristics, you should be able to stitch the image exactly(most errors in stitching software comes when you didn't shoot the images perfectly overlapping, or at different angles, or you took a step forward/back, etc.) You can buy software off the shelf that does a better job than NASA's job.
"How much so a man can walk on mars?"
Please help metamoderate.
I read somewhere after the lander, um, landed that the pictures it were sending back were flawed. The arguement was that Mars, like Earth, was supposed to have blue skies. I can't say that this is a correct assesment but it seems plausible. I do recall watching C-SPAN last week or the week before when a group was talking about Spirit. One thing they talked about was a simple little 4-color chart that could be used to sync Spirit's camera color settings to once the rover landed. The plate the color chart was on also doubled as a sun dial (low tech at it's best!). Anyhow, I thought the blue sky idea was interesting. Is the red planet really red when you're standing on it's surface?
But I guess they still got the right kind of guys at some places... Not the Microsoft "let's throw more hardware" kind, but the old Woz-like variant which used to wander on the Earth until some 20 years ago (ok, I'm being bitter).
BTW, we could do many things alike if we had many modular, combinable parts.
Things which come already assembled in one immutable way are not useful to create, invent or innovate.
I understand that thermal noise can be a cause of noisiness in CCD images. Do the low temperatures on Mars (or in cold places on Earth, for that matter) have any significant effect on digital photo quality? Could the cold temperatures on Mars be taken advantage of to maximize the quality of images taken there?
Weeks of coding saves hours of planning.
Or, am I correct to assume that the atmosphere on Mars (or lack of it) does not propogate sound waves, so there would be nothing to listen to???
A lot of people see the prints from my digital SLR, a gracefully aging Nikon D1h and are astonished to learn it's from digital. Most then refuse to believe it's only a 2.7mp camera.
Near all my pictures at www.gavincato.com in the photography section are with the Nikon D1h.
The Nikon D1h has only a 2.7mp sensor, but the output is fantastic. The pixels are large, and the noise is pretty low. It's pretty much noiseless until you hit 800 ISO, and even at 1600 ISO it's significantly better than 1600 speed film.
NASA is very correct in saying the lens & sensor are important, for example most of my lenses are ludicrously expensive (often more than the camera body) and the majority of them are fixed length lenses and thus have incredible optics.
I've previously owned a Nikon D100 which had 6mp, but I found to my surprise that I preferred the output & prints from the D1h. I originally bought the D1h to complement the D100 (the D1h is a crazy fast camera designed for sports), not replace it, but after a while I ended up selling the D100.
The guys in the Canon camp have said the same thing, they much prefer the output of the 4mp Canon 1D vs the 6mp Canon 10D.
question. is your 1d full frame? that would easily explain why it's got better iamges than the 100d.
i'm one of those int he canon camp and i do have to concur with you on those findings. my canon has an APS sized sensor and can take very noise free iamges upto 400 ISO, at 800, it's still useable.. and i've done 1600 ISO shots, but i only use it if i have no other choice.
one thing people don't understand is the extra MPs only matter if you want to blow up your images. most people rarely print bigger than 4x6" and at the largest 8x10" 3MP resolution is more than good enough for an 8x10" print. after that, you want the best out of those MPs... NOT more MPs.
I believe sex is highly over rated... unless it involves me
His name was Sergei Mikhailovich Prokudin-Gorskii, and they're awesome. You have to remind yourself of the time period when you see them, or you'll instinctively think they're more modern:
http://www.ummagurau.com/art/russia/
I don't know which is more depressing, that 2/3 didn't care enough to vote, or that 1/2 of those that did are crazy.
What color was it before the picture was edited and if it wasn't "peach", why does NASA think we need to see a pink Martian sky? What happened to the blue sky that Viking showed us? Just wondering if anyone else has noticed this.
Shut up and eat your vegetables!!!
I am not surprised at all, and I am glad to see that NASA didn't fall for the marketing hype that the number of megapixels is the determining factor in the quality of a photographic image. Personally, I would prefer a 1 megapixel camera with an SLR (single reflex lens) to 5 megapixel camera.
I believe the secret to art is the process used in filtering down to the information that you really need in a picture.
The two areas that you miss with smaller megapixel cameras are textures and fractal patterns such as the shapes of leaves, forests and grassy fields. However, when you really need to study a pattern, you can zoom in (assuming you have a decent lens) and get the information you need.
I think 1 megapixel is the right image size for the job at hand. Of course, all the pictures from Mars with trees and grassy fields will be a bit fuzzy. Those of rocks, landscapes and strange green alien creatures will turn out fine.
It's not popular because you can't get a good library for free.
Most other formats are 'good enough', so why work with something new?
jpeg2000 is SLOW too. Even on high-end machines, it feels like jpg on a 386. It is impressive though. I have a highres picture I took which I took with a friends EOS Digital Rebel, I compressed with with jpeg2000 to under 200K and then visually compared them zoomed in. Was amazing, there were few noticable differences even when zoomed way in. (obviously wasn't using lossless option).
For lots more info on the Pancam, other instruments on the rovers, and tons more history, news, status updates, video, 3d photos, and more, check out:
(AXCH) 2004 Mars Exploration Rovers - News, Status, Technical Info, History.
The article is clueless technically. The panorama cam takes hundreds of 1 megapixel images and stitches them together to produce the panoramas that have been published on the web.
i t_ p011.html
Anyone can acheive a similar effect with their digital camera by taking enough pictures and stitching them together with software.
To get an idea of what the raw ccd images look like from the panoramic cam, check out the raw image gallery from JPL:
http://marsrovers.jpl.nasa.gov/gallery/all/spir