backtop


Print 77 comment(s) - last by johnsmith9875.. on Aug 14 at 4:12 PM


  (Source: mckaylaisnotimpressed.tumblr.com)

Mike Ravine, Advanced Projects Manager for Malin Space Science Systems  (Source: torynfarr/flickr)

A two-image panorama shot with the rover's 1024 x 1024 mono navigation cameras  (Source: NASA)
Reasons included the amount of data produced, the fact that they had to meet the needs of different cameras, and the team's familiarity with these sensors

NASA recently accomplished a huge feat by landing its Mars rover Curiosity on the Red Planet, but one of the questions on the minds of many was why such a sophisticated machine used 2 MP cameras.
 
Mike Ravine, a project manager from Malin Space Science Systems, was happy to answer that in an interview with Digital Photography Review. He said the main reasons for using 2 MP sensors in the cameras were the amount of data produced, the fact that they had to meet the needs of different cameras, and the team's familiarity with these sensors.
 
"There's a popular belief that projects like this are going to be very advanced, but there are things that mitigate against that," said Ravine. "These designs were proposed in 2004, and you don't get to propose one specification then go off and develop something else. Two MP with 8 GB of flash didn't sound too bad in 2004. But it doesn't compare well to what you get in an iPhone today."
 
The amount of data produced is a large reason for using 2 MP cameras. There just isn't enough bandwidth for anything more powerful because the cameras must share with other instruments. Curiosity sends data back to Earth via the UHF transmitter, which transmits to two spacecraft orbiting Mars. The data is then sent back to Earth, and this system only allows for 250 megabits per day to be shared amongst various instruments.
 
The 2 MP camera sensors also were the tools of choice for the use of four different cameras, including the MAHLI, MARDI and two Mastcams. Having four different sensors for each camera would be expensive and more difficult to maintain rather than having one type of sensor all across the board. 
 
The team's familiarity with the sensors was crucial, too. The team knew the behavior of Truesense imaging chips and Kodak's KAI-2020 chip, so it makes sense that they'd work with what they know. 
 
"We know how to clock them and drive them," said Ravine. "They're a very easy CCD to drive."
 
Other issues, like the low pixel count, are not an issue either since the two Mastcams will create images from multiple exposures. 
 
NASA rover Curiosity landed successfully on Mars earlier this week after launching from Cape Canaveral, Florida on November 26, 2011. Curiosity is a one-ton, nuclear-powered, Mini Cooper-sized science laboratory that will explore the Martian surface for the next two years. 

Source: Digital Photography Review



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Still no reason why they couldn't put 8MP
By quiksilvr on 8/9/2012 1:46:02 PM , Rating: -1
You could easily have placed an 8MP camera and just make it take 2MP shots and send it. Then during less busy times or when the mission lifetime is past, take awesome 8MP shots.




By geddarkstorm on 8/9/2012 1:48:43 PM , Rating: 2
Their decision making is just such a Curiosity.


By Brandon Hill (blog) on 8/9/2012 1:50:48 PM , Rating: 5
Aren't these designed pretty much locked in YEARS ahead of launch? It's not like you can just walk into Frys Electronics and pick up a T3i and strap it to a rover. It has to be qualified/verified to work in the harsh conditions of space.

NASA tends to use gear they are familiar with for these missions.


RE: Still no reason why they couldn't put 8MP
By geddarkstorm on 8/9/2012 1:53:13 PM , Rating: 3
And rightly so. Considering the extreme environments, temperatures, radiation, and so forth; you've gotta go with some seriously hardened and robust stuff. I don't think there's any 8 MP cameras that are validated to handle that--and certainly not in '04. Besides, taking multiple exposures allows us to bump up the resolution count anyways.


RE: Still no reason why they couldn't put 8MP
By DennisB on 8/9/2012 2:20:50 PM , Rating: 2
Astronomy CCDs (cooled by NO2) have been in the Megapixel range since the early 80's so size is not the problem.
The problem is as mentioned the RAM, processors and UHF.
Not mentioned but related is the power consumption of higher power transmitters. Curiosity probably cannot have enough to spare. The instruments or getting scientific information have higher priority. That's the mission after all. Not to mention the power needed to move it.


RE: Still no reason why they couldn't put 8MP
By Solandri on 8/9/2012 6:46:53 PM , Rating: 1
So how much will it cost to deliver a replacement tank of liquid nitrogen to Curiosity on Mars?

The liquid nitrogen-cooled CCDs can be high res because the cryogenic cooling reduces the noise you typically get from high-MP small commercial CCDs. If you can't cool it, you want a bigger lower-res sensor to reduce noise, just like a 16 MP DSLR produces cleaner pixels than a 8 MP point and shoot.


By m51 on 8/10/2012 11:03:59 AM , Rating: 3
Astronomy CCD's are cooled to reduce dark current in the CCD so very long exposures can be used.

Cooling does not affect the resolution usable.

Lower resolution sensors with larger pixel elements are used in Astronomy CCDs so there is more light gathering area per pixel and shorter exposures can be used.

These are all the result of needing to take images at extremely low light levels. This is not the case with the cameras on MSL.

The MSL cameras are already cooled since Mars is damn cold to begin with.


By bleach32 on 8/9/2012 2:34:47 PM , Rating: 2
Great response but I'm curious if you have any idea of just how MUCH JPL can "improve the resolution?" You give those guys a few pictures taken in generally the same direction, especially if you can provide any object ranging information or camera position change information for the shots and WOW!


By Samus on 8/10/2012 1:46:55 AM , Rating: 3
it's a navigation camera. who cares what the MP is as long as its sensitive enough they can see where its going.


RE: Still no reason why they couldn't put 8MP
By quiksilvr on 8/9/12, Rating: 0
RE: Still no reason why they couldn't put 8MP
By Hieyeck on 8/9/2012 1:57:23 PM , Rating: 4
What all of NASA gets to spend in a year, the US military pisses away in 30 minutes in a single conflict. You expected any better?


By delphinus100 on 8/9/2012 10:21:45 PM , Rating: 2
(sigh) 'The military' often has to pass serious amounts of data through a finite bandwith as well (and sometimes under attack).

Shannon's Law doesn't care about anyonme's budget or politics...


RE: Still no reason why they couldn't put 8MP
By mindless1 on 8/10/2012 1:03:27 PM , Rating: 3
Biased much? Military is essential to preserving our way of life. Space exploration won't be as beneficial to society until the day comes that we have to leave the planet and have somewhere else hospitable to travel to.

You might say the space program developed a lot of tech in our lives today. Certainly true but how does the program exist? By preserving our society and global peacekeeping efforts by the military.

Funny how someone can benefit from peace but not appreciate the cost to maintain it. Without the population willing to go to war (Americans are spoiled by their prosperity, and yes I am American) what remains is using technology and special ap tech costs are quite high, there is only one customer.


RE: Still no reason why they couldn't put 8MP
By BansheeX on 8/11/2012 2:41:18 PM , Rating: 1
America hasn't had to preserve itself since WWII, not to mention WWII was actually paid for with war bonds. Every ill-conceived adventure since then has been purely to line the pockets of the industries that Smedley Butler warned us about. We're going to destroy society all by ourselves with debt-financed welfare and warfare. Look at the trade deficits we're running each month, look at the Fed's inability to raise rates past 0 without bankrupting our biggest banks and insurers again. We're like Greece with a printing press right now. Interest rates are necessary to create savings from which real loans and capital come, but when your national debt is this high and short-term, 1.5 trillion to interest alone is hard to pay when tax revenue is 2 trillion and you're spending 4 trillion.


RE: Still no reason why they couldn't put 8MP
By mindless1 on 8/11/2012 8:03:06 PM , Rating: 2
... BECAUSE we have such a mighty military.
Yes there certainly is that industry/profit angle but at the same time, everyone profits from stable world wide markets instead of repairing destroyed cities and developing weapons of mass destruction.

The military budget is not destroying society. You can't just lump in essential government services with all the nonsense that money is wasted on.

Military spending in inflation-adjusted dollars hasn't changed all that much for over half a century, while the national debt skyrocketed since the early 1980's or especially in the last 4 years. It's the OTHER programs getting out of control that make the interest on that debt seem out of proportion, by not leaving enough in the budget to spend nearly the same we used to.

http://en.wikipedia.org/wiki/File:PerCapitaInflati...


By johnsmith9875 on 8/14/2012 4:12:00 PM , Rating: 1
Military spending is 100% discretionary. Obligations paid to our taxpayers are mandatory, thats why they're called obligations.

Republicans have redefined this using their Orwellian Newspeak Dictionary and have called them "Entitlements".


By johnsmith9875 on 8/14/2012 4:08:11 PM , Rating: 1
The military's job is preserving itself, and preserving socialist jobs that wealthy businessmen in the defense industry want to keep.


RE: Still no reason why they couldn't put 8MP
By kleinma on 8/9/2012 2:13:06 PM , Rating: 2
Maybe the thing is there isn't a whole lot to see there.
I would imagine that for what they want to get in pictures, 2MP will be good enough. Most of the work the rover will do is with its other science instruments. They are looking for microbial life evidence, probably not something a few MP in a camera CCD will make or break.


By BZDTemp on 8/9/2012 3:07:31 PM , Rating: 2
Exactly. If more MP would have mattered NASA would of course had done something to use more recent camera tech but they aren't there to take pretty pictures.


RE: Still no reason why they couldn't put 8MP
By maven81 on 8/9/2012 6:06:26 PM , Rating: 3
I think that's what it really comes down to. It's not like higher resolution sensors aren't in Mars orbit as we speak! The MRO for example (the probe that took the photo of the parachute) was launched in 2005, and uses custom sensors to sweep out panoramas over 20,000 pixels wide: http://en.wikipedia.org/wiki/HiRISE Clearly the tech is there, and the bandwidth and processing is there, but the goal wasn't to get the highest res images. The MRO is the imaging mission, this rover is an exploration mission.


By Solandri on 8/9/2012 6:42:19 PM , Rating: 4
MRO has a much more powerful transmitter, and being in orbit makes its transmissions much easier to pick up from Earth. The rovers actually uplink their data to orbiters like MRO, MO, and MGS (RIP) first, then the orbiters transmit it to Earth. But these communications windows are very short (usually a few minutes), so the amount of data they can uplink is limited despite much higher local bandwidth.
http://www.astrosurf.com/luxorion/qsl-mars-communi...

quote:
The MRO for example (the probe that took the photo of the parachute) was launched in 2005, and uses custom sensors to sweep out panoramas over 20,000 pixels wide

That isn't really a camera. It's a strip which is swept across the area being photographed (either by turning the telescope or movement of MRO itself). The other axis is generated by the sweep, and is limited by the computer memory onboard. It's incapable of generating an instantaneous 2D picture.

MRO does have a 6 MP camera on board. Do note that camera requirements for orbiters are much less stringent than landers since they don't have to deal with atmosphere or dust.


By ChronoReverse on 8/9/2012 6:44:25 PM , Rating: 3
From that link, HiRISE has 14 0.26MP CCD's that it takes multiple images and stitches them together (and only for red, blue/green is 4000 pixels wide).

So the 2MP sensor of the Curiousity has more megapixels =)


RE: Still no reason why they couldn't put 8MP
By estarkey7 on 8/9/2012 2:25:27 PM , Rating: 5
I've been working in New Product Development for the last twenty years.

If they constantly wanted to perform continuous improvements on every subsystem on Curiosity, guess where that rover would be today? Still in some lab at NASA because it would never get done.

At some point you have to shoot the engineer(s) and ship the product. When the product requirements were approved for this project, things that were necessary were included and things that weren't were cut out. Also, it's not like they had the luxury of adding a bunch of secondary and tertiary systems just in case they get a chance to use them - the damn thing already weighs a ton - which caused this elaborate landing sequence in the first place.


By JediJeb on 8/9/2012 3:38:17 PM , Rating: 2
quote:
If they constantly wanted to perform continuous improvements on every subsystem on Curiosity, guess where that rover would be today? Still in some lab at NASA because it would never get done.


Just like how people complained because Duke Nukem was continually pushed back. You set the specs in stone, then you complete the project. Back in 2004 a 4MP sensor was pretty much the top end that was in common use, and lower ones were the very well know ones. It is much better to have a 2MP camera that works well, than an 8MP camera that would glitch out the first time they took a picture.


By Manch on 8/10/2012 2:24:59 AM , Rating: 2
I remember a long time ago, NASA was looking for 8086 procs for the space shuttle when upgrading some of the onboard systems. They were asked teh question then too why they didnt use faster newer cpu's and more up to date componenets. They said basically the same thing. They need to harden them for use in space. Validation requirements to ensure no anomolies, etc. I think another reason was that the smaller the transistor, the harder it was to protect it. Damnit, now I need to go find that article.


By Reclaimer77 on 8/9/2012 2:01:43 PM , Rating: 2
Who cares? We've seen gillions of pictures of Mars at this point. A few more megapixels isn't going to reveal some discovery that we wouldn't have seen otherwise.

I think they explained very well why it would not have been an "easy" matter of putting an entirely new camera on it years after it's been designed and tested to work with another.


By ChronoReverse on 8/9/2012 2:19:51 PM , Rating: 2
In 2004 the 8MP sensors were JUST starting to be available in pro-level digital cameras.

Furthermore, this isn't as simply as duct-taping the camera to the rover. They mentioned how they're familiar with the sensors for instance in the article.


By Mitch101 on 8/9/2012 2:33:08 PM , Rating: 2
I was thinking the same. Also its not the number of megapixels its the quality of the sensor and lenses finally multiple images can and will be stitched together to make for large megapixel images anyhow.


RE: Still no reason why they couldn't put 8MP
By Flunk on 8/9/2012 2:22:26 PM , Rating: 2
Downsizing those shots would take more processing power. This thing is very power and processing constrained. If there is only enough bandwidth for 2MP, putting anything higher on it is a waste.

These things have very strict objectives and the operating requirements are designed to hit those objectives. Pretty pictures is not one of those objectives.


By MrBlastman on 8/9/2012 3:14:03 PM , Rating: 2
At only 120 watts, there really isn't a lot to work with. I'm quite happy with what they have here now and see no reason people should be fussing about the lack of 8 megapixels.


By Amiga500 on 8/9/2012 2:25:07 PM , Rating: 3
quote:
You could easily have placed an 8MP camera and just make it take 2MP shots and send it.


And it could just as easily have broken on launch from g loading... or on upper atmospheric boost from further g loading.

Or from radiation in the transit to mars.

Or from turbulence during the descent.

Or from a host of other reasons that cameras and cameraphones on earth simply don't have to worry about.

Qualifying something for spaceflight is not the work of a minute!


By TexMurphy on 8/9/2012 2:28:27 PM , Rating: 5
I am astounded that this is the comment everyone is replying to.

I would have been easy to replace such a camera? I just don't understand what makes you think you're qualified to say that it's so easy. "Hey man, just send the next probe to Neptune. It's easy, it's just a little further than Mars."

Seriously, I logged in for the first time in months just to say this, that's how retarded your comment is.


By ESmooth on 8/9/2012 2:38:53 PM , Rating: 2
Don't we all already know the MP's don't matter as much as the lens anyway? Its not like there's a competing Mars rover maker that is banking on selling itself by way of better features.


By GotThumbs on 8/9/2012 2:51:25 PM , Rating: 1
Your ignorance is overwhelming. The story just addressed the reasons...and your still too stupid to understand. You never heard of the KISS principle? Keep It Simple Stupid.

Fine, You just go ahead and build your own version...and send it to Mars. You show everyone that you know best and that the NASA scientists don't know anything.

What a Twit you are.


RE: Still no reason why they couldn't put 8MP
By talikarni on 8/9/12, Rating: 0
By Solandri on 8/9/2012 6:56:50 PM , Rating: 2
See my other post above. The downlink speed during landing was a gerry-rigged network of the lander to an orbiter to earth (Curiosity actually landed on a part of Mars facing away from Earth). Any time the transmitter is moving or its position isn't exactly known, the bandwidth will be much lower.

During the mission, Curiosity will transmit data to an orbiter, which will then relay it to Earth. For Spirit and Opportunity, these uplinks to an orbiter are at about 256 kbps - 2 Mbps. But are limited to a few minutes a day (when an orbiter is sufficiently high overhead).


RE: Still no reason why they couldn't put 8MP
By s12033722 on 8/9/2012 7:28:10 PM , Rating: 6
I am a digital camera design engineer. I've worked on quite a few scientific cameras similar to this. The Kodak (now Truesense) KAI series has had higher megapixel count imagers that would have been no problem to get rated for a mission like this, but the higher resolution would have come at the cost of power consumption, slower frame rate, larger size, and increased weight, all of which are much more important than resolution in this type of application. Resolution is really a fairly unimportant aspect of camera performance in these types of applications. Of much higher importance are how well the sensor converts light to electrons (quantum efficiency), low noise readout electronics, high full well capacity to allow good signal to noise ratios, linearity of the camera response, etc. Resolution is a very narrow, relatively useless aspect of camera performance that gets hyped up in the consumer space. The actual determining factor in overall system resolution is usually not the imager in the first place - it's the lens. If they knew their lens was going to limit them to the effective resolution of 2 MP, why put in a bigger sensor with all the tradeoffs that go with it just to over-sample the image blur?

Don't try to apply consumer-level understanding of cameras to a scientific instrument like this. There's so much more complexity in a real camera than what gets talked about at the consumer level that it's just useless.


RE: Still no reason why they couldn't put 8MP
By m51 on 8/10/2012 11:24:49 AM , Rating: 2
Excellent post.

I'd also point out that they did consider the need for higher resolution on MSL and addressed the problem by having two cameras on the mast with different focal length lenses, 34mm and 100mm. Longer lenses gives you even better resolution that increased pixel count. That coupled with photomosaic should more than do the job. I believe they will be downloading the full resolution Mastcam panorama in the next few days and it will be impressive.

Malin Space Science Systems had also built a pair of Mastcams with zoom lenses, but these were cancelled in March 2011 when it was deemed there was not enough time to fully test them out and integrate them on the vehicle before launch.


RE: Still no reason why they couldn't put 8MP
By EricMartello on 8/12/2012 5:45:46 AM , Rating: 2
The focal length of a lens does not affect resolution - it affects field of view and depth of field. The optical resolution of a lens is largely determined by the quality of the optics (glass) and the precision with which they are engineered and assembled.


By m51 on 8/14/2012 10:45:24 AM , Rating: 2
Focal length doesn't strictly affect resolution. However a longer focal length lens gives more amplification of the image. That allows you to resolve more detail at a distance with the compromise of reduced field of view. A telescope by any other name..

Reduced field of view can be accomodated with photo mosaics.


By EricMartello on 8/12/2012 5:40:36 AM , Rating: 2
Some guy claims to work for kodak...I guess we better believe him since this is the internet.

That being said I agree with most of what you said except you're downplaying the value if resolution a bit too much.

The better the lens' optical resolving capabilities, the more a high resolution sensor will benefit and a lower resolution sensor will experience diffraction...and even average lenses have higher optical resolution than 2 MP.

Quantum efficiency is more or less at its practical limit for current sensor technology, and higher sensor resolution does not necessarily translate to additional weight or power consumption.

The main benefit of a lower resolution sensor would be the ability to consolidate the an image processing ASIC onto the chip itself and thus simplify the overall design of the system...but a higher resolution sensor would yield more detail per shot. It really boils down to whether they want a lot of "ok" pics or fewer high-res pics...in exploration I'd opt for the most detail that's practical because we have missed potential discoveries in the past due to basic resolving limitations of our equipment.


"Mac OS X is like living in a farmhouse in the country with no locks, and Windows is living in a house with bars on the windows in the bad part of town." -- Charlie Miller














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki