Help - Search - Members - Calendar
Full Version: The sol 588 and 589 "strange bright lights" [sic]
Unmanned Spaceflight.com > Mars & Missions > MSL
Pages: 1, 2, 3
elakdawalla
In the past, unmannedspaceflight.com has provided a public service to debunk conspiracy claims by helping people locate and describe images related to the weird claim of the week. My favorite two examples of these were the "Sasquatch on Mars" and the "Puddles on Mars." In that spirit, I'm hereby providing links to images and other data related to this week's fun, originating with this Houston Chronicle story: "NASA photo captures strange bright light coming out of Mars", quoting a UFO enthusiast website. Alan Boyle followed up on it with this post that annoyed me, and one just now that has me intrigued, quoting Justin Maki as saying it's not a cosmic ray hit.

Phil Plait asked me about this and I took one look at it and said "cosmic ray hit." Here's the picture, right Navcam from sol 589:

If you compare the image to the left eye taken at exactly the same moment, there is no bright pixel -- this is diagnostic of an event that affected only one camera, so is most likely a cosmic ray hit:

Another thing that tells you it's likely a cosmic ray hit and not a bleeding pixel from something bright is the fact that pixel bleeding on Navcams happens in the horizontal, not vertical direction. Just check any Navcam image of the Sun, or this low-light image from sol 593 in which the sloping side of the RTG is overexposed and bleeding horizontally. By contrast, cosmic ray hits can be oriented in any direction, such as in this nighttime Navcam pic.

And I figured my debunking work was done, until someone pointed out to me that there's another right Navcam image, shot from a similar but not identical location, at the same time of day, pointed in roughly the same direction, that also contains a bright dot. Here's the picture, right Navcam from sol 588:

As with the sol 589 image, the bright dot is not in the Left Navcam frame taken simultaneously, although this time that fact is explained by the presence of a foreground butte blocking the field of view:


This dot is different from the other one. It is not extended vertically. It's just a dot, that overlaps more than one pixel. Still, I would be inclined to dismiss this as a cosmic ray hit (saturating pixels, in one eye and not the other) without extraordinary evidence to the contrary. There are interesting coincidences here that could lend themselves to an alternative explanation, such as a specular reflection from a bright object: both are on the horizon, seen in the same direction, at the same time of day. But there is another coincidence that has me skeptical: seen in right eye only of the Navcam. And the vertical extension of the bright pixel in the sol 589 image just doesn't make sense for a specular glint; that would extend horizontally, not vertically, while cosmic ray hits can make streaks in any direction. So I am still inclined toward cosmic ray hits and coincidence, but I'll admit to being less totally certain about that after seeing the sol 588 image than I was after seeing the sol 589 image alone. And now there's this from engineering camera lead Justin Maki, via Alan Boyle:

QUOTE
"Bright spots appear in single images taken by the Navigation Camera on NASA's Curiosity Mars rover on April 2 and April 3. Each is in an image taken by this stereo camera's right-eye camera [with links to the April 3 and April 2 pictures] but not in images taken within a second of each of those by the left-eye camera [again, with links to April 3 and April 2]. In the two right-eye images, the spot is in different locations of the image frame and, in both cases, at the ground surface level in front of a crater rim on the horizon.

"One possibility is that the light is the glint from a rock surface reflecting the sun. When these images were taken each day, the sun was in the same direction as the bright spot, west-northwest from the rover, and relatively low in the sky. The rover science team is also looking at the possibility that the bright spots could be sunlight reaching the camera's CCD directly through a vent hole in the camera housing, which has happened previously on other cameras on Curiosity and other Mars rovers when the geometry of the incoming sunlight relative to the camera is precisely aligned.

"We think it's either a vent-hole light leak or a glinty rock."


Anybody got anything else to add? Other images of this spot? Where is the spot on the map, exactly? I can (and have) drawn lines on Joe Knapp's map but I'm not convinced I understand the geometry precisely enough to want to say anything about where any putative reflective object would be.
elakdawalla
Also, here are the visualizations from Joe Knapp's site of the viewsheds of the two images.

Sol 588:



Sol 589:

mhoward
Only thing I have to add: Many people here will remember, earlier in the mission, we had a whole discussion of sparkles that (I'm going to say now) probably were glints off rocks - because it seems like there were too many to be readily explained by bits of hardware laying around, and also we've seen a lot of polished, shiny rocks. I can't find the discussion at the moment. All of those glints appeared in both cameras, though (which is why we thought they might actually be interesting).
fredk
I noticed the 588 blip at the time since it appears to sit on the distant ridge and doesn't obviously look like a cosmic ray hit (unlike the 589 blip, as you point out Emily). The lack of lnav to confirm it as a cosmic ray at first disappointed me. But then I remembered that (MSL especially) overlaps the navcams considerably. So the neighbouring rnav shows the same region as the 588 blip:
http://mars.jpl.nasa.gov/msl-raw-images/pr...NCAM00252M_.JPG
No surprise: no blip. My conclusion at the time: cosmic ray. I see nothing new here to change my mind.
djellison
My very crude take on triangulation from the two Navcam images - put me in a spot that I think is also visible in one of the Sol 580 MastCam mosaics.... perhaps the tall thin rock left of center, near the top, on that rock face, is the cause of the excitement.

Sol 580 MastCam
http://mars.jpl.nasa.gov/msl-raw-images/ms...044E01_DXXX.jpg

That area is also visible on Sol 572 in Navcam - and that same rock is just about visible.
http://mars.jpl.nasa.gov/msl-raw-images/pr...NCAM00250M_.JPG

freddo411
Here's another "glint" from the right navcam, on sol 589. Also in the top left corner of the image, somewhat on the apparent horizon. Note the camera is pointing in completely different direction. Left camera shows no glint.

http://curiosityrover.com/imgpoint.php?nam...0000NCAM00262M_
djellison
That glint is really nothing like the others in question - that really REALLY is a CR hit. It's nowhere near the local horizon. It's against the Gale crater wall some 25+km to the north.
ngunn
How simultaneously are the navcam pairs taken? If there is even a small delay then one camera might miss a very small glint that was 'twinkling'. That seems unlikely but I can't think of any other way a real feature in the landscape could be missed by the other camera. It's surely too distant for the difference in viewing angle to come into play.
djellison
Well - we have one pair where the feature is clearly hidden behind a nearby hill. The other pair is more curious - it may very well be the exact same situation - more local topography occluding it in one eye. I believe it's near instantaneous - the have the same spacecraft clock time time to the second.
fredk
Here's a stretched, 200% zoom of the two 588 frames:
Click to view attachment
No hint of anything in the first frame. To repeat, I see no sign of anything other than cosmic rays. These aren't the blips you're looking for. Move along.
neorobo
Depending on the smoothness of the surface, specular reflection is heavily dependent on the angle between the pixel and the surface normal of the object it's viewing. If the light reflected off the surface almost perfectly aligns with the camera, you will see the glint, otherwise you won't, even with a very small change in viewing angle between the two cameras.

It could be that one camera pixel or pixels is very closely aligned with the reflection angle, whereas the other camera is off slightly and little of the specular reflection travels to it comparatively. I'm not convinced that's it, as over several km the difference in viewing angle will be very small and would require a very very smooth surface. I'm not sure about the pixel bleeding that Emily is talking about either, I'll have to look into that. Just something to think about though.
djellison
I was absolutely 100% "It's a CR hit" when I saw them. I've done a complete 180. 589 could be a CR hit. 588 isn't. It hides behind a hill behind the two eyes. It also happens to triangulate well with the Sol 580 MastCam and 589 Navcam feature to a tall, thin shiny rock.
fredk
QUOTE (djellison @ Apr 8 2014, 09:00 PM) *
It hides behind a hill behind the two eyes.

Huh? Do you mean that it's visible in 588 rnav but not 588 lnav? That's also consistent with a cosmic ray hit on the rnav and not the lnav! It's certainly not evidence that 588 is not a CR hit. And the fact that the other 588 rnav frame shows absolutely nothing also points to a CR.
mhoward
One 'gleam' of hope for resolving this (ha ha): The area of the Sol 589 Navcam image was covered by Mastcam-100 on sol 590; only the thumbnail is available on the website at the moment.
djellison
The Sol 588 observation ( visible in Right, not in Left ) is entirely consistent with an actual object being obscured by the perspective shift between the two eyes applied to the northern side of the nearby topography.

The Sol 589 observation does not have a similar topography to explain it's one-eye appearance (although a small rock on the nearby topography might explain it )

However, if one triangulates between the two observations, one finds a point on a small ridge line. That point is also visible in Sol 580 MastCam imagery that shows a tall, thing, bright rock at the exact same point ( see my first post on this thread )

This means either....


1) 2 CR hits happened to appear on two images on the same camera on two sols at different pixel locations that happens to be geometrically consistent with a tall thin bright rock see 8 sols earlier ( which is QUITE a coincidence )

2) It's an actual thing.

I'd expect M100/M35 and ChemCam imagery of the same spot to be acquired soon that should quite easily settle the matter. The object, if it's real, is approx 160 meters away.



elakdawalla
Where would it be in the sol 593 late afternoon Navcams? Behind the butte?

Edit:
QUOTE (freddo411 @ Apr 8 2014, 01:06 PM) *
Here's another "glint" from the right navcam, on sol 589. Also in the top left corner of the image, somewhat on the apparent horizon. Note the camera is pointing in completely different direction. Left camera shows no glint.

http://curiosityrover.com/imgpoint.php?nam...0000NCAM00262M_

Although the "light leak" notion of Justin's is my least favorite explanation, I'll note that the bright pixel in the image that freddo411 points out is also in the upper leftish area of a right Navcam image so could be investigated as light-leak-related. But if it were that easy to make a bright pixel with a light leak, we'd see them in a lot more images, I'd think.
djellison
The rock I think we're seeing glinting is also visible in Sol 582 MastCam as well.

MR - Sol 582
http://mars.jpl.nasa.gov/msl-raw-images/ms...325E01_DXXX.jpg

MR - Sol 580
http://mars.jpl.nasa.gov/msl-raw-images/ms...044E01_DXXX.jpg

elakdawalla
Boy, would I like to see less-JPEGgy versions of those images...
PDP8E
QUOTE (elakdawalla @ Apr 8 2014, 07:33 PM) *
Boy, would I like to see less-JPEGgy versions of those images...

Me too! Here is a lightly rinsed 3x ... really looks like a CR ... your mileage may vary

Doug, there is that rock off to the left on the ridge,in the image you posted:
MR - Sol 582 (2 posts up)

Click to view attachment
fredk
The 589 blip is very clearly a CR. Note that the bright pixels extend above the far outcrop and are set against the distant Gale rim - see the arrowed part here:
Click to view attachment
As Emily pointed out in the first post, navcam bleeds horizontally. (This doesn't look like bleeding anyway!) How else could we be seeing bright pixels above the outcrop in question? CRs often form streaks that look very similar to this.

This view also shows that the bright blip/streak is above the outcrop with flattish rock Doug identified above.

Therefore there is nothing to triangulate 588 with. It seems like a stretch to imagine that one of the 588 frames caught a glint off a distant rock and the other 30 seconds earlier didn't.

The other problem is that we have never seen mirror-like specular reflections from large surfaces on Mars. Even if a surface was somehow smooth and flat enough, dust tends to cover surfaces.

If it looks like a CR, and quacks like a CR, and occam's razor dictates it's a CR, then...
atomoid
excellent sleuthing, especially on locating the candidate rock face! its still hard to see what exactly is the nature of that rock..? thats it of course!!!!!
It still seems so unlikely for a glint showing up only in the Right eye, but not in the left eye that's just a few inches away yet does appear the previous day in a separate location, though the different shape of the glint is consistent with that, if there were to be two faces of the rock that happened to be configured perfectly for sending reflections to these two locations..
zoomed non-interp 400% shows it blends into image as if its being resolved through the lens.. sol588 on left sol589 on right
Click to view attachment
MarsInMyLifetime
What percentage of the cruise stage may have survived entry and dropped short of the "flown" EDL path? I'd expect any such parts to be small and well ablated, not visibly shiny, but this spot does seem to be in the possible strew field for that stage. I'm not willing to push this idea very hard--I lean towards the CR explanation as well, but Mars EDL being as untidy as it is...
PDP8E
The glint on sol 588 does not appear to be a CR !

Here is a straight forward bi-cubic spline 6x image from the original...
the bright spot is concentrated on a central pixel on a ridge ...

The wall of Gale is also in the image, anyone that can triangulate (from the original 588 Nav R B ) may get closer to the 'rock'

Click to view attachment
DLC
Is it ice?
The vertical, light-colored feature seen in the mastcam 580 image looks to me to be a good candidate for something like the well-known Mars gullies, caught in the act.
Consider:
It appears to be precisely vertical, and lighter in color than anything else in the frame, consistent with ice or flowing water.
It extends from the rim-edge of a portion of what appears to be the crater wall.
It appears to be in the same vicinity as something that seems to be producing unusually pronounced specular reflections.
It looks an awful lot like a waterfall or frozen waterfall, and we know that seasonal gullies are seen on Mars. Have any been seen at this latitude or in this vicinity?
It seems that the JPL folks are just saying specular reflection, at this point. That's consistent, and appropriate.
djellison
QUOTE (DLC @ Apr 8 2014, 07:36 PM) *
Is it ice?


No. It's also visible in two MastCam images from 6 and 8 days earlier - and it's clearly a well wind polished rock.

A large expanse of ice like that would have sublimated away long, long ago.
fredk
QUOTE (PDP8E @ Apr 9 2014, 03:04 AM) *
The glint on sol 588 does not appear to be a CR !

On what basis do you say that? CR's can take on a wide range of appearances - have a look at some low-light navcams or better nighttime navcams.

Seriously, the speculation is getting a bit out of hand here. There are better places for this kind of discussion...
nprev
Is this (possible) thing anywhere near the planned traverse path? The mainstream media is going nuts about it; might be worth a side trip just to let the hot air out of the loonies.
elakdawalla
No, it's not near the planned path, and no, it's not worth a side trip. It's either a cosmic ray hit, or something in RNav optics, or a shiny rock. None of those things would warrant taking Curiosity away from the planned path.
nprev
Agree scientifically, of course. Too bad, though; great opportunity to place some well-deserved egg on the right faces.
djellison
QUOTE (fredk @ Apr 8 2014, 09:01 PM) *
There are better places for this kind of discussion...


Such as JPL perhaps where today Justin Maki, engineering camera team lead, has said a shiny rock 160m from Curiosity is a perfectly plausible explanation, as well as a possible navcam light leak or CR hits.
I appreciate your emphasis on keeping the woo out of UMSF, but your repeated demands to shut down the discussion is at odds with mainstream thinking on the issue.
vikingmars
QUOTE (nprev @ Apr 9 2014, 06:52 AM) *
Agree scientifically, of course. Too bad, though; great opportunity to place some well-deserved egg on the right faces.

Could some "candidates" be seen from orbit on HiRISE images ? Or are they too small ?
xflare
QUOTE (fredk @ Apr 9 2014, 05:01 AM) *
Seriously, the speculation is getting a bit out of hand here. There are better places for this kind of discussion...


Everyone likes solving mysteries biggrin.gif biggrin.gif

The candidate rock even looks quite bright in that mastcam image.
monitorlizard
Something that shiny makes me think of iron-nickel meteorites, which can be essentially pure metal. We already know there are meteorites on Mars. If it isn't too old, it
would not have oxidized the surface yet (which would take much longer on Mars than Earth).
Ant103
I can hardly think it's a reflexion from a distant object. The parallax between the two Navcam is too close to have a very bright spot in one, an none in the other.
Let me explain : On Earth, when you are looking to a city in sunlight, sometimes there is a specular reflexion coming from a window or a large panel of bright metal. This reflexion don't disappear if you just moving less than a meter. Or the object that is reflecting the sun have to be a few centimeters large, with a very plane surface (typically : a mirror). I doubt that on Mars such a rock exist : very plane and very reflective in the same time. They can be flat, but not as flat as mirror. Because the surface aspect is certainly not a perfect plane, the light beam could not be so tight. And just because of that, if it was actually a reflexion from a rock, it should have been saw by the TWO camera, not one.

For me it's just a cosmic ray hit, maliciously place by coincidence on a particularly place, but it's just cosmic rays.
ngunn
They're shiny [iron meteorites] but rarely have perfectly flat faces. I'd prefer very large single crystals, possibly feldspar, in the unlikely event that these are real landscape features.
ustrax
It's baaack...
http://i16.photobucket.com/albums/b14/ustrax3/nru4.jpg
jmknapp
Some basic geometry of the situation using SPICE:

Raw image: NRB_449790582EDR_F0310000NCAM00262M_

Map view

time of shot: sol 589 03:08:07 P.M. LMST (2014 APR 03 10:00:03 UTC) et = 449791270

NRB boresight direction (azimuth, elevation): 298.81 -16.86

Direction of bright pixels: 286.06 0.67

Sun position: 300.33 31.79

So the sun position relative to the line from the camera to the bright pixels is 14.27 northward and 31.12 up.

Click to view attachment

Kind of interesting (coincidence?) that the pixels are within a degree of horizontal.

EDIT: Also, if it is the same object in both images, here's my attempt at triangulation:

Click to view attachment
Gerald
Do the two lines of sight intersect in 3d space?
centsworth_II
It seems that the intersection of those two of those lines is in mid air above the bowl of that crater.
djellison
That triangulation matches, by my estimation, the tall, thin, bright, shiny feature seen in these Mastcam images a week earlier.

http://mars.jpl.nasa.gov/msl-raw-images/ms...325E01_DXXX.jpg

http://mars.jpl.nasa.gov/msl-raw-images/ms...044E01_DXXX.jpg

It will also be in this image once down linked at full res
http://marsmobile.jpl.nasa.gov/msl/multime..._DXXX&s=590

john_s
Tall and thin, indeed. Bright and shiny? Not so sure. It looks to me like an rock of unusual shape but ordinary color and brightness, that just happens to have a vertical face oriented to catch the light in that shot.

John
fredk
QUOTE (ustrax @ Apr 9 2014, 09:48 AM) *

Hah! laugh.gif

At least the Beacon didn't appear and disappear like... like... you know it's coming... a cosmic ray! wink.gif
marswiggle
After individually rotating and resizing each of those R navcam images, cropping the pair, enlarging 1.5 x, vertically stretching 2 x, and sharpening a bit for good measure, I think I can discern the ridge at 150 m (late edit: probably farther off, 200 m or so) distance in both images as an arched outline at the center. This comparison also works as a crude cross-eyed pair. The bright spots do not seem to be located at or near the ridge, instead being unrelated to it and to each other as well. Don't know how much this helps though.
elakdawalla
I'm still having a hard time thinking of the sol 589 one as anything but a cosmic ray hit; it extends several pixels above the horizon, onto the distant hills.

It did occur to me while awake last night (as these things do) that you could make a singularly bright vertical rock face by exposing the planar surface of a gypsum vein. Still, I would expect more than one such feature to show up, and I would expect it to be visible to both eyes, and it should not have that vertical extension. So I just don't like it.
marsophile
If the vein was inside a narrow crack between two rocks, then a glint from it would only be visible within a very narrow viewing angle.
jmknapp
If the sun is 30 degrees above the horizon and the reflected ray is even with the horizon, wouldn't that imply that the reflecting surface/point is oriented 15 degrees from the horizon? I.e., relatively horizontal.
Gerald
If it's just one surface, yes, but it if it's actually a reflexion it could also be a reflection of e.g. two planes in different angles.

[About possible reflecting surfaces on Mars] A cleavage of feldspar might cause plane reflecting surfaces, too.

[About the position estimate] A relative pointing error of 1 could move the intersection of the two lines in Joe's graphic to the hill in front. But after trying to estimate the error ellipse of Joe's cross bearing by using the parallax of the bright pixels relative to background rim features as a partly independent measurement, I got a distance of 176 m relative to the Sol 588 position; that's very close to the position in Joe's graphic. This may still be wrong, or there might have been two independent reflexions. Nevertheless, I'm now prefering coincidental CR hits as most likely, as well.
Click to view attachment
Click to view attachment
(The second graphics is Joe's graphics with some additional annotations.)

Camera artifacts (straylight, hot pixels) should follow some reproducible rule, which I can't yet recognize.
jmknapp
My triangulation could easily be 1 degree off. The direction of the boresight (center of image) is pretty reliable (as reliable as the SPICE kernel anyway) but to get the direction of the bright pixels I just counted pixels from the center assuming a uniform field of view of 45.33 and that the rover was level (which it fortunately was to within about a degree). So any lens distortion could throw it off too. It could be done more precisely if warranted, BUT how about this argument: if it was some kind of specular reflection that was so sensitive to direction that one NAVCAM saw it but the other didn't, and it was 176m away as Gerald figures above, then a 42 cm shift in baseline (0.13) is the difference between seeing and not seeing. The sun for its part moves 0.13 in about 45 seconds. So then what are the odds that the thing would be seen twice on different sols from different positions? Pretty low right? Therefore by Ockham's Razor it's not a reflection.
testguru
QUOTE (jmknapp @ Apr 9 2014, 10:07 AM) *
My triangulation could easily be 1 degree off. The direction of the boresight (center of image) is pretty reliable (as reliable as the SPICE kernel anyway) but to get the direction of the bright pixels I just counted pixels from the center assuming a uniform field of view of 45.33 and that the rover was level (which it fortunately was to within about a degree). So any lens distortion could throw it off too. It could be done more precisely if warranted, BUT how about this argument: if it was some kind of specular reflection that was so sensitive to direction that one NAVCAM saw it but the other didn't, and it was 176m away as Gerald figures above, then a 42 cm shift in baseline (0.13) is the difference between seeing and not seeing. The sun for its part moves 0.13 in about 45 seconds. So then what are the odds that the thing would be seen twice on different sols from different positions? Pretty low right? Therefore by Ockham's Razor it's not a reflection.

Is there any chance to do time lapse photography of the area in question? Zooming in to the area with 64X64 or 256X256 pixel coverage should not overburden the uplink bandwidth I would think
and a resultant "movie" with ~ 1 sec intervals should catch any reflections. Also doing this at night would be interesting.

I have wished for a meteor watch movie at night but don't know if the rover has this capability or not. In October it would be really interesting to see if a meteor storm happens.
Phil Stooke
The experience with Spirit was that meteor searches revealed only cosmic rays. Even things thought at first to be meteors couldn't be confirmed in the end to be anything but cosmic rays. And presumably a search would be done with the Navcams, for the wider field of view, so basically the same cameras. So it may not be useful, and certainly would be a low priority... except, as you say, in October, which is a special time.

Phil

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2018 Invision Power Services, Inc.