Printable Version of Topic

Posted by: Sunspot May 20 2004, 01:23 PM

Has anyone any idea what kind of imaging system is being proposed for the MSL? Would it be the same or simiilar to the MER's, or as improved as the MER pancams are to those that were used on Pathfinder?

With a nuclear power source for the rover, and hopefully the mars telecoms orbiter in place the data rate and data volume could be phenomenal

Posted by: djellison Dec 20 2004, 02:12 PM

Going on the recent news ( i LOVE looking at really old forum posts ) the answer involves any and all of the following words

Order of Magnitude

Can of Whoop Ass

Mike Malin

VIDEOS

Where did all my bandwidth go



Doug

Posted by: RNeuhaus Jul 8 2005, 01:39 AM

The new bandwith that the MRO will transmit is between 1 to 30 MBits/sec depending upon to the Mars' atmosphere transparency between MRO and Earth. The improved transmission bandwith is based on new technology of utilizing the bandwith closer to the light.

Rodolfo

Posted by: BruceMoomaw Jul 8 2005, 02:12 AM

There was a lot on this at the recent LPSC:

http://www.lpi.usra.edu/meetings/lpsc2005/pdf/1214.pdf
http://www.lpi.usra.edu/meetings/lpsc2005/pdf/1170.pdf
http://www.lpi.usra.edu/meetings/lpsc2005/pdf/1580.pdf

Posted by: James Sorenson Sep 1 2008, 11:00 PM

Forgive me for opening up an old post, but I am a little confused on how Mastcam can aquire both multi-spectral, aswell as natural color imaging all in the same package. How would this all work out when there is bayer filter limited to the visible spectrum and IR filters that move into the infrared part of the spectrum?. How could any narrowband information passing through the filter wheel be salvaged when the light also passes through a broadband bayer filter?. I emailed the MSSS guys about this but have not got a response back .

Thanks

Posted by: djellison Sep 1 2008, 11:15 PM

http://www.unmannedspaceflight.com/index.php?showtopic=4251&st=20&p=91486&hl=filters&#entry91486

Posted by: James Sorenson Sep 1 2008, 11:27 PM

Thanks Doug .

Posted by: PhilCo126 Nov 8 2008, 11:42 AM

remarkable ChemCam: http://www.youtube.com/watch?v=fEZ5dEi4oPo&feature=related

Posted by: BrianL Nov 8 2008, 05:00 PM

It's also reassuring that this rover will be able to defend itself if attacked.

Posted by: jekbradbury Nov 8 2008, 07:41 PM

One issue that just popped into my head (not sure if it's been mentioned before): MSL will survive for a long time, so there will probably be significant dust buildup on MastCam (and on the other cameras), just like with MER. Phoenix showed that a magnetic ring around a surface will prevent that surface from being covered with dust. Has any effort been put into such a system on MSL's camera lenses? Would it even be possible?

Posted by: mcaplinger Nov 8 2008, 08:08 PM

I'm unaware of any quantitative analysis of how much dust buildup there has been on Pancam; maybe others know more. I had thought the dust buildup was mostly on the hazcams, which are much closer to the ground.

As for Mastcam, the current fixed-focal-length systems have the lenses set quite far back inside the sunshades (if you look at http://www.msss.com/msl/mastcam/index.html the front element is at about the front of the lens barrel, color-coded red), so the path for dust is quite long. MAHLI, of course, has a movable cover.

Posted by: dvandorn Nov 8 2008, 08:21 PM

The MER Navcam lenses (esp. Oppy's) have dust accumulations ranging from light to very heavy -- Oppy's are quite occluded on the right sides of the images. However, this isn't a result of general dust accumulation, it happened almost entirely during the global dust storm (which also deposited dust onto the "protected" optics of the mini-TES instruments on both MERs, enough that it might possibly render the devices useless).

I think there's a difference between designing optics to remain "as clean as possible" during normal accumulation cycles and designing dust-storm-proof optics. I'm positive that MSL's cameras are good for the former, but doubt they're designed to handle the latter.

-the other Doug

Posted by: Doc Nov 8 2008, 08:33 PM

Quite right. Aren't they concerned enough to make a mechanism for removing dust from the lenses?
If I recall well, the Beagle 2 lander had a 'wiper' for each eye of its stereo imaging payload. Why on earth won't anyone consider such a thing?

Posted by: mcaplinger Nov 8 2008, 08:39 PM

Because the cameras aren't on Earth. Designing a mechanism that can survive for many years under martian conditions and not fail in some way that degrades imaging is not worth it, given the low chance of dust contamination as I have already described.

The cameras are normally stowed pointing down. Was there some period on MER when they were left pointing up?

Posted by: djellison Nov 8 2008, 09:25 PM

What makes you assume no one has considered it. I would bet the farm that it's not only been considered, but looked at carefully, the risks studied, and a call made.

Doug

Posted by: djellison Nov 8 2008, 09:31 PM

Spirit's cameras all remain in great health

Opportunity got a smattering in both Navcams and Pancam's. If you notice a lot of Pancam imaging with Opportunity is done with only one half, or 3/4s of the CCD, as one side got quite a smattering.

http://marsrovers.jpl.nasa.gov/gallery/all/opportunity_p1671.html is an example of the technique.

http://marsrovers.jpl.nasa.gov/gallery/all/1/p/1607/1P270849438EFF90CSP2629L4M1.JPG is a typical example of the obstruction


Doug

Posted by: mhoward Nov 8 2008, 09:46 PM

The contamination happened during the dust storm, didn't it? So it's not exactly 'buildup' so much as a 'smattering' as you put it.

Regardless, the contamination of Opportunity's optics remains distressing - I'm sure to the science team as well as the sightseers.

Posted by: mcaplinger Nov 8 2008, 11:34 PM

http://www.planetary.org/news/2007/0930_Mars_Exploration_Rovers_Update_Spirit.html

It looks like the effects of this contamination could easily be taken out with a proper flat field, and I'm a little surprised that they haven't done this.

That said, owing to the configuration of the Mastcam I think it will be far more resilient to this sort of problem.

There was never any serious consideration given to adding a cover mechanism to the Mastcam, and nothing I see in the Pancam data would justify the added cost, complexity, and risk. I wouldn't even know how to begin to implement a "wiper" that would work reliably under martian conditions, and some kind of roll of clear material that could be driven past the lens would be problematic for all kinds of reasons.

Posted by: ugordan Nov 9 2008, 09:40 PM

If you look closely at the comparison mosaic on that page, notice that while the corner darkening was removed with a new flatfield, the frames suffer from lowered contrast there. The dust particles are no longer just a nuisance in the way, they diffuse/diffract the light. The effect can be likened to Titan's haze but on a much different scale. A flatfield simply cannot remove that. Had the dust been deposited on the actual CCD the situation would probably have been different.

Posted by: elakdawalla Nov 9 2008, 10:20 PM

This is a question I've been wondering about too. Were the cameras left unstowed at some point, or did Oppy just happen to get smacked in the face with an unusually large pile of dust during the brief period when it was trying to do imaging one day?

--Emily

Posted by: mcaplinger Nov 10 2008, 01:05 AM

If by "flat field" you mean something that merely multiplies each pixel by a single number, I agree. But these sorts of stray/scattered light problems can be addressed with more sophisticated techniques, for example the way the NEAR images were recovered after the front elements were covered with monoprop burn products. See "Inflight Calibration of the NEAR Multispectral Imager II. Results from Eros Approach and Orbit" by Murchie et al in Icarus 155, 1. I'm not sure the MER images are bad enough to require that level of processing though.

Posted by: Deimos Nov 10 2008, 01:53 AM

First, a 'stowed' camera is still at risk on MER. There are no in-flight dust shields for Navcam, Pancam. Looking down avoids sedimentation of dust, but not turbulence. I think turbulence maybe was involved. Second, I don't think the camera bar has been left stowed at nadir in quite some time. Stowing the cameras now places them looking sorta down, sorta out, I think. The actuators were not designed for a 5 year mission, let alone one with Opportunity's current ambition. If they fail, they should fail where Navcam is still useful rather than contemplating the rover's navel, umm, deck.

Posted by: DDAVIS Nov 10 2008, 09:07 PM

I wouldn't even know how to begin to implement a "wiper" that would work reliably under martian conditions...

I recall reading that Viking had two methods to avoid dusty optics, a physical barrier which covered the 'slit' when the camera was in its 'rest' position as well as compressed air which could be used to blow off the optics. I would think the latter could be adapted to a wide variety of camera designs.

Don

Posted by: Phil Stooke Nov 10 2008, 10:41 PM

"I would think the latter could be adapted to a wide variety of camera designs."

I think we need it for Spirit's solar panels now...

Phil

Posted by: mcaplinger Nov 10 2008, 10:59 PM

I'd be willing to bet that the tank to contain the gas would weigh more than an entire Mastcam camera head. Not really practical given the mass constraints.

Posted by: DDAVIS Nov 11 2008, 12:18 AM

the tank to contain the gas would weigh more than...

Perhaps, but I wonder what the Viking unit weighed? How much more compact cound a similar modern system be built? I imagine something like a soda charger cylinder filled with very clean compressed gas, perhaps giving puff(s) of gas upon demand. A magnetic 'ring' around the lens could also help. However this is dealt with, cameras for long term Mars surface missions will require more attention to the dust hazard as part of the design.


Don

Posted by: djellison Nov 11 2008, 12:28 AM

Now that's a solution I've not seen mentioned before, and it might make a lot of sense.

Posted by: Fran Ontanaya Nov 11 2008, 12:34 AM

What about holding the air in the bodywork itself?

Posted by: djellison Nov 11 2008, 08:27 AM

Making the entire electronics box a pressurised body? Given that it has to travel thru a vaccum? How do you engineer in the insruments and their openings to the sample collection mechanisms. That's an epic engineering challenge right there.

Posted by: AndyG Nov 11 2008, 10:33 AM

For Mars the body wouldn't have to be at insane levels of pressure - just "enough" over the external ~1 kPa to provide a squirt of gas where and when you need it. I don't think the high pressure 20 MPa sort of structure you see on air tanks is at all necessary here.

Just to put this into perspective: the r/c model sub community regularly work with pressures up to ~50kPa and manage with commercial linkages/prop shafts, etc., penetrating the core electronics box without too many problems.

That said, http://www.amazon.co.uk/Giottos-GTAA-1901-Large-Rocket/dp/B000RGM762/ref=sr_1_4/275-9776193-6764349?ie=UTF8&s=electronics&qid=1226399472&sr=8-4. And it's even rocket-shaped!

Andy

Posted by: djellison Nov 11 2008, 03:03 PM

You still have to render the entire box air tight. That means seals of every access panel, seals for every cable in, every cable out. That pressure is going to go up and down with thermal cycles. It might change the thermal properties of the box itself. Then you've got to have valves, flexible pipe work out the box, up the past and to the front of the lenses. Pipework that's going to bend with every camera movement. With all due respect, a $1.X billion dollar rover is just a tiny bit more complex than an R/C sub.

AND - at the end of all that - if you've got a CRT monitor or an old TV - go and blow on it. The dust doesn't move very much, if at all. I know - wind has cleared arrays etc before, but we don't know the mechanism by which that occurs, nor do we know the mechanism by which it sticks.

If we really really want to have dust-removal - the best way is with the thin-film technique I blogged about at Valencia '06. Tiny power, low mass, low complexity. Turn them on for a few seconds - dust gone. Gas tanks, wipers, rolls of film - they're all very heavy, complicated, and have their own failure modes.

DOug

Posted by: AndyG Nov 11 2008, 03:40 PM

Unfair! Any CRT being bombarded by electrons (and thereby charged) in a dry, dusty atmosphere is not quite in the same league as a glass lens sat in front of a passive CCD.

Personally I like the thin film concept (much better than abrasive wipers - though even one of those would be helping Spirit's solar panels right now), but I also like the home-engineering "what if...?" kind of discussions here. Hmmm...Blame it on a childhood filled with Meccano.

Andy

Posted by: Jeff7 Nov 12 2008, 08:24 AM

I wonder about maybe some kind of electrostatic "wiper" - maybe something to move across the lens while holding a charge. Might the dust adhere to something like that, or perhaps be repelled by it? Either way, it could get rid of dust. Possibly maybe. I'm afraid I've not got much experience with electrostatics.

Posted by: tedstryk Nov 12 2008, 08:45 PM

There is also the issue of having to make sure that whatever pressurized gas you used didn't cause a residue to build up on the lens. The cure could be worse than the disease.

Posted by: mcaplinger Nov 12 2008, 09:05 PM

I don't recall this reference, could you point me at it again?

This is all getting pretty off-topic for MSL; the hardware is built and doesn't include any kind of active dust removal.

Posted by: Del Palmer Nov 12 2008, 10:47 PM

The thin-film technique is described here:

http://science.nasa.gov/headlines/y2006/19apr_dustbuster.htm

Posted by: djellison Nov 12 2008, 10:56 PM

Bingo. That's it. The wires are so fine, I bet for a future design, one could put something like that over the front of a Mastcam like Hood. The wires would be totally out of focus and not noticeable ( like the heating element in a heated front windscreen).

It certainly beats wipers, rollers, blowers etc.

Doug

Posted by: mcaplinger Nov 12 2008, 10:59 PM

Cool, thanks! But I don't think this would work for a camera without a good bit of technology development, because it'd be very difficult to make the transparent material optically flat.

Posted by: djellison Dec 4 2008, 11:55 AM

Just noticed

http://mars.jpl.nasa.gov/msl/spotlight/images/20081119a/20081119_1.gif

It has four rear hazcams - two each side of the RTG. I wonder if it's the same at the front.

Posted by: Doc Dec 4 2008, 02:54 PM

Perhaps, but isn't the reason because the RTG would block the view of 1 set of hazcams if mounted in the middle? Or maybe because the rover is just so big; hazards increase with size?

Posted by: djellison Dec 4 2008, 03:01 PM

I think it's just that you can't put them where you would want to - because the RTG is there. Can't

Posted by: Oersted Jan 4 2009, 12:30 AM

What about a stationary brush placed somewhere reachable by the mastcam? The brush could be wiped by the back (or whatever) of the camera housing to brush off dust, and then the brush itself could be used to clean the lenses. We might be on Mars for 10 years (oh joy!), so some sort of cleaning mechanism would seem necessary...

On another subject, I was thinking to myself whether there is some sort of mastcam lens protection during landing. Caps or just turning the lenses towards an enclosed space? A lot of dust will swirl up.

Also, wouldn't it be great if the navcams were turned on during landing, taking a movie? Not just great, but helpful for analysing EDL. If they have caps on them during landing, maybe the caps could be made of transparent material, so some filming would still be possible despite the dust.

Posted by: djellison Jan 4 2009, 12:56 AM

No hardware changes from the current design before launch.

We have MARDI doing a movie during descent, we don't need Navcams doing it - and they'll be pointing into the rover deck at that time anyway - along with MastCam I presume.

Doug

Posted by: mcaplinger Jan 4 2009, 01:12 AM

As I said earlier in this thread, the Mastcam lens is many inches inside its sunshade and could not be reached by any plausible brush.

As to the navcams imaging during EDL: of course Doug is right, the navcams and Mastcam are pointed down towards the deck when the RSM is stowed. I guess the hazcams are looking at the ground during descent but the hazcam frame rate is quite slow (remember DIMES from MER) so only a few images could be taken, whereas MARDI will be running at about 4 FPS.

Posted by: ElkGroveDan Jan 4 2009, 02:11 AM

Does this MARDI have a microphone, Mike?

Posted by: mcaplinger Jan 4 2009, 02:40 AM

Nope, no microphone on the MSL MARDI. You should note that while it shares a name with the instrument on PHX, it's a completely different electronic design and doesn't use the cellphone processor that made adding a microphone to the PHX design fairly straightforward.

Posted by: ElkGroveDan Jan 4 2009, 03:09 AM

You can't simply plug one in through the spare USB port?

Posted by: mcaplinger Nov 17 2009, 04:38 PM

Interesting comment about the Mastcams from Mike Ravine of MSSS over at http://nasawatch.com/archives/2009/11/avatar-a-stunni.html#comments

Posted by: djellison Nov 17 2009, 05:33 PM

Not entirely accurate is it? If the fixed focus stays - then one eye will be about 3.75x the res of Pancam, the other one, about 1.25x Pancam. But the more I think about it - the more the zoom descope hurts.

Posted by: mcaplinger Nov 17 2009, 05:42 PM

Doug, you're right about the resolution. What we are really losing with the zoom is the ability to capture the entire surroundings in color with a small number of frames. As is, we will have only hazcam/navcam monochrome images to do that. And of course there's the 10 fps capability, which is really not that interesting with a narrow FOV.

Posted by: djellison Nov 17 2009, 05:47 PM

Damn - 10fps wide angle movies. Driving. DD's. It's enough to bring a grown space fan to tears.

Posted by: hendric Nov 18 2009, 05:59 PM

Would it be possible in a future rover to use single shot color imagers for the nav and haz cams? They're not scientific instruments per se, so there wouldn't be much of a penalty.

Posted by: Deimos Nov 19 2009, 05:40 PM

The "penalty" is technical rather than scientific. Single shot color uses patterned filters on the sensor. The spatial content of the scene is undersampled, and later interpolated. Stereo processing for range maps and navigation would then look at the results of the interpolater. It is not obvious how much degradation there would be, especially if the green channel or maybe the sum of all 3 channels got used in the stereo/navigation processing. But it might be a tough sell without data. (As I recall, using a bayer pattern detector under mastcam's science filters was a tough sell to some of the camera team, and that's a less touchy operation.) For the record, I think it is do-able; but it's not the same to think it can be done reliably versus to show you'll never put the rover in jeopardy due to such a decision.

Posted by: ugordan Nov 19 2009, 05:45 PM

On Mars where there's not much color variation in the first place, wouldn't you effectively be able to recover full resolution by simply multiplying different bayer "channels" by a corresponding factor to make a sort of white-balanced, "grayscale" image? Effectively getting back a panchromatic image as you'd normally have.

Posted by: mcaplinger Nov 19 2009, 06:50 PM

The balancing operation ugordon describes is something we've been playing with for Mastcam. But note that since Mastcam's Bayer filters are transparent in the near-IR, when we use narrowband filters in those wavelengths the images are effectively grayscale and there is no spatial penalty (we just turn the interpolator off in that case). Of course you wouldn't want to have a filter wheel on your engineering cameras.

Posted by: hendric Nov 19 2009, 08:13 PM

If you can get a custom filter pattern applied, you could do RGB_clear_ quads across the chip instead of the RGBG Bayer.

Also, the hazcams on MER were the same die as the navcams and pancams, with the hazcams downsampled. In my idea, use an off-the-shelf single shot color die for navcams and hazcams, with the hazcam SW only looking at one downsampled color channel. The pancam could use the same die without the Bayer pattern, to reduce software complexity and allow full spectral range.

Would there be value to get color info from Navcams or Hazcams? Maybe, maybe not. It might be possible to see a difference in the surface colors that might not be as obvious or distinctive in B&W, ie there's a bluish patch over there, we should check it out vs that's just a typical grey patch, let's move on. The rear hazcams could provide their color data to another process that looks for outliers, ie the white silica that Spirit digs up.

Anyways, sounds like a good PhD project for someone someday.

Posted by: Deimos Nov 20 2009, 04:00 PM

The kind of ideas here are why I think it can be done. It just remains to be shown that in 99-th percentile cases, it still works well enough to do no harm. Each adds complexity to the data stream--a color camera can generate jpegs without the rover (or camera) cpu being involved. Multiple paths (color download vs. reprocessing the colors to retrieve monochrome) add complexity. Again, something that seems easy to deal with. The software would have to include data from all parts of the pattern through a standard interpolation scheme or through some sort of multiplicative balancing (which is sensitive to small errors in assumed color). Blurring the image could help--aliasing is much worse than oversampling for stereo processing. So, having RGB-Clear and just using the clear would be bad, but using RGBG or RGBC and summing them into a lower resolution product would be fine; slightly blurring the latter would be optimal.

Ultimately, I think this is a good way to go. But I see no urgency for the path, since the value of the color is undercut by the assumption that it works partly due to the lack of color variation... But as long as it still works in cases where the color is useful, and I think it can, then this will become the path. Mastcam is paving the way for this, even with the unfortunate descopes. Others will follow.

Posted by: algorimancer Nov 20 2009, 08:44 PM

Using a Sigma FOVEON sensor, http://en.wikipedia.org/wiki/Foveon_X3_sensor, where each pixel records all colors simultaneously, would avoid the interpolation problem.

Posted by: briv1016 Mar 26 2010, 04:26 PM

According to Emily's twitter account, James Cameron is teaming up with Malin Space Science Systems to try and bring back the original zoom feature to Mastcam.

http://twitter.com/elakdawalla/status/11058862622

Posted by: djellison Mar 26 2010, 05:29 PM

I really really really really really really really really really really really really really really really really really really really really really really really really really really really hope they can pull this off. I was hoping Cameron might help out with some of his Smurf money to rescope Mastcam so he could then get a 3D HD film out of it

Posted by: Stu Mar 26 2010, 07:40 PM

Wow... that would propel MSL into a whole new catergory of cool, wouldn't it? Imagine the impact it would have if people could travel to Mars, virtually, in 3D, not just at an IMAX theatre but in their own local cinema a la AVATAR, etc. It could be the greatest leap forwards for Outreach, and for increasing public awareness of the worth of space exploration, for a generation.

Pleeeeeeeeeeeeeeeeeeeeeeeeeeeeease let that happen.

Posted by: punkboi Mar 26 2010, 08:04 PM

I read elsewhere that Panavision (the long-time supplier of movie cameras for Hollywood films) was to provide lens for the Mastcam. Don't know if that's still the case.

Posted by: briv1016 Apr 7 2010, 01:54 AM

Fixed focus Mastcam has been delivered. MSSS "just started" building the zoomed version.


http://marsprogram.jpl.nasa.gov/msl/news/whatsnew/index.cfm?FuseAction=ShowNews&NewsID=990

http://www.msss.com/press_releases/mast_delivery/

Posted by: Explorer1 Apr 7 2010, 05:50 AM

That's the camera's first image, correct? Or were there some from before?

Posted by: centsworth_II Apr 7 2010, 06:48 AM

There is a Mastcam mosaic http://www.lpi.usra.edu/meetings/lpsc2010/pdf/1123.pdf

Posted by: djellison Apr 7 2010, 07:35 AM

It's AN image taken AFTER shipping to JPL. There will of course have been many many test images taken previously to that.

Posted by: lyford Apr 8 2010, 01:38 AM

That swiss army knife photobombs all the MSSS promo shots!

Posted by: tharrison Apr 26 2010, 04:43 AM

Definitely, it would be a huge leap for "public engagement" (NASA's new term for public outreach) to have a 3D high-definition movie that's REALLY from Mars—not just a computer animation like you see right now in documentaries on the Discovery Channel, and to have it in a real theatre rather than just in science center IMAX theatres like Hubble 3D. I am incredibly excited for this possibility. It's also personally exciting because I'll be one of the folks operating and targeting the Mastcam, so it would be really neat to take my family to a film using Mastcam footage so they can see it.

Posted by: algorimancer Jul 2 2010, 01:08 PM

Any word of progress with zoom-cam? Likely to be completed in time?