PDA

View Full Version : The colors of the rocks on Mars.



RGClark
2004-Apr-02, 12:39 AM
There were several color images of the rocks at the Spirit and Opportunity sites at the latest conference, yet no one commented on the blue color of the subsurface images of the rocks.
Strange, since one of the participants was Jim Bell lead scientist for the Pancam imager.

News conference archived here:

http://mars.systemsfirst.com

Bob Clark

RGClark
2004-Apr-03, 01:30 PM
The news release below describes the subsurface as grey. What's curious is that during the news conference several images were displayed that showed a blue subsurface for the rocks. If this is only a problem of calibrating the exposure times why wasn't this done for the displayed images? You may not be able to put all the available corrective algorithms to the images but at least you could combine them using the correct brightness corrections according to exposure times.
The lead scientist for the Pancam imager Jim Bell was present at the conference so he must have been aware of the questions about the colors shown on the rocks.

Spirit Finds Multi-Layer Hints of Past Water at Mars' Gusev Site
April 1, 2004
http://www.jpl.nasa.gov/releases/2004/93.cfm

Also, a member of this bbs who attended the Lunar and Planetary Science Conference in March reported that Steve Squyres said the colors of the subsurface of the rocks was "bluish".

Here are the captions that accompany the released JPL images of Mazatzal:


http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/Mazatzal_Postbrush-380.jpg
This image was taken by the Mars Exploration Rover Spirit's panoramic camera during the rover's grinding of the rock dubbed "Mazatzal" with its rock abrasion tool. The picture shows the rock after two targets dubbed "New York" (left) and "Illinois" were brushed on sol 81. The exposed, dark surface is a second coating beneath a top white veneer. This approximate true-color image was created using the panoramic camera's red, green and blue filters.
http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/captions/image-7.html


http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/Mazatzal_Postrat1-380.jpg
This image was taken by the Mars Exploration Rover Spirit's panoramic camera during the rover's grinding of the rock dubbed "Mazatzal" with its rock abrasion tool. The picture shows the rock after the rover drilled 3.8 millimeters (.15 inches) into the target dubbed "New York" on Sol 82. The dark grey coating seen after brushing remains on the right side of the hole, while the left side is the underlying basaltic rock. This approximate true-color image was created using the panoramic camera's red, green and blue filters.
http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/captions/image-8.html

Maybe by "grey" they mean "any dark color seen on Mars that is not red or brown."


Bob Clark

RGClark
2004-Apr-03, 01:47 PM
Here's the caption to the image of Mazatzal taken by the Microscopic Imager:

http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/NewYork_Brushed-380.jpg
This enhanced-color image taken by the microscopic imager on the Mars Exploration Rover Spirit on sol 79 shows the rock dubbed "Mazatzal" after a portion of its surface was brushed clean by the rover's rock abrasion tool. The reddish material on the right side of the image is the original dust coating. The darker, grayer surface on the left side was exposed after brushing. The crack in the rock may have once contained fluids from which minerals precipitated along its walls. The color in this image was created by combining pictures taken with the microscopic imager's orange-tinted dust cover in both its open and closed positions.
http://www.jpl.nasa.gov/mer2004/rover-images/apr-01-2004/captions/image-11.html

Once again "grey" appears to mean "any dark color seen on Mars that is not red or brown."


Bob Clark

TrAI
2004-Apr-03, 03:33 PM
There have been several threads on this board about the apparent bluishness of the rocks in the color composite pictures, though it is a while since the last I saw.

Anyway,the bluishness is caused by the way the rover imaging system will stretch the grayscale of each imaging to use all the levels available. The features with the most blue will be 100% white in the blue channel grayscale, so if you try to make a color image of the L4, L5 and L6 channel, the features will be too blue. It can be corrected, but you'll need the extra data from the imaging system. Yes, Gray rocks would exhibit this phenomena, as it is more reflective in the blue end of the spectrum than the red and yellow rocks...

NASA have released some corrected pictures, but I guess the scientists of the missions have more important things to do with their time than color correcting every image they release, that kind of thing will probably be done after the missions are over, for now it is likely more important to use the images to plan the gathering of scientific data while the rovers still is operation. For scientific use the stretched grayscales are good, they contain more data than a picture where the channels are adjusted to keep the relative color ratios of the channels correct...

somerandomguy
2004-Apr-03, 09:41 PM
It kinda seems to me that EVERYTHING looks blue next to the reddish dust ... looking at the whole image it's almost like sand and surf at Key West, but if you cover up the right side of that Mazatzal image and just look at the left side, it looks bluish-gray.

RGClark
2004-Apr-03, 11:55 PM
It's not just a case of looking blue. If you use a image processing program that can do color histograming you find the blue areas really do reflect most strongly in the blue part of the spectrum.


Bob Clark


It kinda seems to me that EVERYTHING looks blue next to the reddish dust ... looking at the whole image it's almost like sand and surf at Key West, but if you cover up the right side of that Mazatzal image and just look at the left side, it looks bluish-gray.

RGClark
2004-Apr-04, 12:01 AM
The phrase "approximate true color" suggests they did do color balancing according to the exposure times.
Here's what an uncalibrated image looks like:

http://www.lyle.org/~markoff/processed/2P134081436EFF2238P2530L234567M1.JPG


There have been several threads on this board about the apparent bluishness of the rocks in the color composite pictures, though it is a while since the last I saw.

Anyway,the bluishness is caused by the way the rover imaging system will stretch the grayscale of each imaging to use all the levels available. The features with the most blue will be 100% white in the blue channel grayscale, so if you try to make a color image of the L4, L5 and L6 channel, the features will be too blue. It can be corrected, but you'll need the extra data from the imaging system. Yes, Gray rocks would exhibit this phenomena, as it is more reflective in the blue end of the spectrum than the red and yellow rocks...

NASA have released some corrected pictures, but I guess the scientists of the missions have more important things to do with their time than color correcting every image they release, that kind of thing will probably be done after the missions are over, for now it is likely more important to use the images to plan the gathering of scientific data while the rovers still is operation. For scientific use the stretched grayscales are good, they contain more data than a picture where the channels are adjusted to keep the relative color ratios of the channels correct...

Tripp
2004-Apr-04, 05:40 AM
Here is a good demonstration of what is going on and leading to these "blueberries".

Here is what the MER Imaging team at JPL has had to say about the subject:


"We target the exposure times to give every image roughly the same average density value, regardless of filter. So because of that, it's dangerous to just blindly use the raw images for an RGB composite. The reason it works for some scenes, like the calibration target or spacecraft parts, is that if there is enough 'gray' stuff in the scene (metallic spacecraft parts, for example) then that is what sets the exposure time. When looking at Mars, though, which is quite red, the blue and green exposure times are significantly longer and so the resulting 'raw' RGB composite will look way too blue. Cool, but still 'too blue.' At the very least they should scale the values by the exposure times."


The below images demonstrate the effect of these density weighed exposures on color. Take note of the color changes of the two rocks evident in all images and how the blue hue is a result of other colors in the images, particularly red density.

On the left, is the scene shot in color, with its red, green and blue channels displayed next to it. The right images were composites created by shooting three individual black and white images through red, green and blue filters. much as MER does it. Each of these shots were averaged, so that the exposures of the scenes have about the same approximate density.

http://www.highmars.org/niac/education/mer/images/rocks01a.jpghttp://www.highmars.org/niac/education/mer/images/rocks01b.jpg

http://www.highmars.org/niac/education/mer/images/rocks02a.jpghttp://www.highmars.org/niac/education/mer/images/rocks02b.jpg

http://www.highmars.org/niac/education/mer/images/rocks03a.jpghttp://www.highmars.org/niac/education/mer/images/rocks03b.jpg

Source:
Using Color to See More - Looks Can Be Deceiving (http://www.highmars.org/niac/education/mer/mer00e.html)

Other Relevant Color Information: (specifically addresses red filter concerns) (http://www.atsnn.com/story/30048.html)

slinted
2004-Apr-04, 07:39 AM
When looking at Mars, though, which is quite red, the blue and green exposure times are significantly longer and so the resulting 'raw' RGB composite will look way too blue. Cool, but still 'too blue.' At the very least they should scale the values by the exposure times."

I'm responsible for the image listed above in RGClark post, and I would absolutely love to follow the direction suggested by the JPL Imaging team and work exposure times into my caluculation of the colors of Mars...too bad that information has never been released and won't become available until the PDS dataset is released 3 months from now, or later.

As stated above, they are uncalibrated, but I'll be the first to say I'd be more than happy to attempt calibration, if there were something to work from. Some people are having more luck than I working from context clues, or expected colors, but my own attempts to create color images are stymied by the lack of hard data as to the exposure times and contrast stretching of the raw images released by NASA. The last thing I want to do in my processing is to tweak settings until it 'looks right'. Until then, we have to rely on the calibration done by the JPL team (which I'm sure is great, but they have released only 1/50th of the actual L2 L5 L6/7 images as color, and maybe 1/15th of the L2L3L4L5L6L7 images), or uncalibrated images as the one shown above.

There is no 'average' exposure time difference, and any attempts to use such gives just about as much distortion as the uncalibrated pictures. The variation is great between one target and the next, and even between 2 images of the same target in the exposure of the blue frames (L6 L7).

A good example of the variation in the relative exposures times would be these two images of the same target taken on the same sol of Opportunity's 'Berry Bowl' target:
http://www.lyle.org/~markoff/processed/thumbs/1P132004235ESF05A6P2552L234567M1_thumb.jpg
http://www.lyle.org/~markoff/processed/thumbs/1P132020763ESF05A6P2552L234567M1_thumb.jpg

These 2 images were processed, uncalibrated, exactly the same way, and as you can see, there is great variation in the colors produced.

I agree that my images, and almost all the others produced by those not at JPL or ASU contain heavy distortion because of the lack of information, and look forward greatly to working with the calibrated images once the PDS set is released.

TrAI
2004-Apr-04, 10:11 AM
The phrase "approximate true color" suggests they did do color balancing according to the exposure times.
Here's what an uncalibrated image looks like:


I am sorry, I was responding to the opening post of this thread, but forgot to quote it. I was commenting the general blueness seen in many of the press conference images. Approximate true color does indicate that they have done some calibration. I have looked at the image you were referring to in your 2. post on this thread, and I agree, it does seem bluish in the picture. It seems the surface is quite shiny, you get reflections at the angles pointing more to the sun, and it is these reflections that are bluish, the darker areas are more grayish...

There are a few possibilities, the rock have bluish specular reflections(I think that is the right english term, at least), or the high intensity of the highlight made it the bluest thing around and the picture is not completely color corrected. Also I think it may partly be that it is reflecting the sky around the sun, which have a bluish color because the dust in the atmosphere scatters away the reddish light, similar things happen on Earth, where you will have bluish reflections of the sky in for example water, or reddish-yellow during sun set/rise...

Anyway, when they said gray, they were probably referring to the surface color, not the specular reflection color.

The picture from the microscopic imager is probably a false color image, the MI does not have RGB color filters, just the yellow dust cover...

RGClark
2004-Apr-04, 10:54 AM
BTW, here is a gorgeous image from your site of the 'Bounce' rock.
I think this image happens to be color accurate because the JPL logo has the right red color.
Some speculation on the shiny areas: perhaps it could be pyrite, also known as "fools gold" because of its shiny appearance. Pyrite is iron sulfide and has been speculated as a possible mineral on Mars since both iron and sulfur appear in abundance on Mars.
Another speculation: the area immediately around the rock that is indented has a wetted look at least to me. If so, then this rock hit by the Opportunity air bags and the "Magic Carpet" area at the Spirit site hit by the air bags both took on a wetted look.
The "Magic Carpet" led to speculation of mud on Mars. I heard Steve Squyres give a suggested explanation that the air bags inflation canisters might give off some liquid water on inflation or deflation. This maybe so. Then this wouldn't give evidence for existing water on Mars but might give evidence for how long liquid water could persist near surface on Mars.

Bob Clark

http://www.lyle.org/~markoff/processed/1P133952981ESF08AQP2574L234567M1.JPG



When looking at Mars, though, which is quite red, the blue and green exposure times are significantly longer and so the resulting 'raw' RGB composite will look way too blue. Cool, but still 'too blue.' At the very least they should scale the values by the exposure times."

I'm responsible for the image listed above in RGClark post, and I would absolutely love to follow the direction suggested by the JPL Imaging team and work exposure times into my caluculation of the colors of Mars...too bad that information has never been released and won't become available until the PDS dataset is released 3 months from now, or later.

As stated above, they are uncalibrated, but I'll be the first to say I'd be more than happy to attempt calibration, if there were something to work from. Some people are having more luck than I working from context clues, or expected colors, but my own attempts to create color images are stymied by the lack of hard data as to the exposure times and contrast stretching of the raw images released by NASA. The last thing I want to do in my processing is to tweak settings until it 'looks right'. Until then, we have to rely on the calibration done by the JPL team (which I'm sure is great, but they have released only 1/50th of the actual L2 L5 L6/7 images as color, and maybe 1/15th of the L2L3L4L5L6L7 images), or uncalibrated images as the one shown above.

There is no 'average' exposure time difference, and any attempts to use such gives just about as much distortion as the uncalibrated pictures. The variation is great between one target and the next, and even between 2 images of the same target in the exposure of the blue frames (L6 L7).

A good example of the variation in the relative exposures times would be these two images of the same target taken on the same sol of Opportunity's 'Berry Bowl' target:
http://www.lyle.org/~markoff/processed/thumbs/1P132004235ESF05A6P2552L234567M1_thumb.jpg
http://www.lyle.org/~markoff/processed/thumbs/1P132020763ESF05A6P2552L234567M1_thumb.jpg

These 2 images were processed, uncalibrated, exactly the same way, and as you can see, there is great variation in the colors produced.

I agree that my images, and almost all the others produced by those not at JPL or ASU contain heavy distortion because of the lack of information, and look forward greatly to working with the calibrated images once the PDS set is released.

RGClark
2004-Apr-04, 11:22 AM
Here's another amazing image from the www.lyle.org site.
That the RATted area really is that white is suggested by the raw visible light filter images: this area appears bright in all of them:

Panoramic Camera :: Sol 068 (32 images)
http://marsrovers.nasa.gov/gallery/all/opportunity_p068.html

Bob Clark

http://www.lyle.org/~markoff/processed/1P134224664EFF08ASP2581L234567M1.JPG

PeteB
2004-Apr-04, 12:28 PM
Also, a member of this bbs who attended the Lunar and Planetary Science Conference in March reported that Steve Squyres said the colors of the subsurface of the rocks was "bluish".
Bob Clark


Who was that? Just curious, I went to LPSC too.

I heard Squyres talk and he may have said rocks are bluish - I'm not doubting it, I just don't specifically recall it. In any event, saying rocks are "bluish" doesn't necessarily mean that they look that way visually. That phrasing, depending on the context, can mean that, spectrally, the rocks are skewed in that direction. A visually gray rock could be described as reddish if the spectrum is toward that end or bluish if in the other direction.

Maybe you have said this before but why is the existence of "blue" rocks on Mars seemingly so important to you?

01101001
2004-Apr-04, 04:24 PM
Maybe you have said this before but why is the existence of "blue" rocks on Mars seemingly so important to you?

Really. This reminds me of the living-room armchair my parents had. The 2 women of the family called it the blue chair, and the 3 men -- only one with diagnosed colorblindness, of the red-green variety -- called it the gray chair; there was never a meeting of the minds.

What does it really matter how the bluish-gray rocks are described in words? Rest assured NASA has the spetroscopic data, and in scientific analyses can state exactly how much brightness a certain rock may provide through each filter.

And, please, could RGClark (and others?) who insert large images into their messages instead provide clickable links so that people who want to avoid spending the time to fetch images, can do so? Thanks.

RGClark
2004-Apr-04, 04:30 PM
That's easy enough to find: just do a bbs search on "Squyres" and "bluish":

Me and Mars Results - Tonight!
Irishman
Bad Fellow
Joined: 23 Oct 2001
Posts: 480
Posted: Wed Mar 17, 2004 8:30 pm
--------------------------------------------------------------------------------
Summary of what I remember.
...
Then it was Dr. Steve Squyres's turn to talk about the Mars Exploration Rovers, Spirit and Opportunity. He is a superb speaker and gave a very exciting, very fun talk. He is dynamic, and strolled around the front of the stage with his wireless microphone while pointing at the slides on screen. He included bits of humor, including the animation of the rover spinning the RAT into a rock and the rover doing the spinning.
He described the mission from launch, and that because they had been working so hard to get the rovers delivered for launch, they didn't learn how to drive them until the 7 months of travel (using the engineering unit/trainers). He talked about the landing sequence, and how the rovers deploy, showing a great animation of the sequence. The petals deploy, then the rover has to deploy its solar arrays, deploy the mast for the cameras, deploy the radio antenna, then it has to jack itself up to extend the legs, and finally prop itself to spin the front to legs over to deploy those wheels.
He gave an overview of the rover design, with the cameras and spectrometer up top, and the arm with instruments on the end. He mentioned how the arm coincidentally has the exact lengths of his own arm from joint to joint.
Then he showed the pictures, many of which have been shown. He started with Gusev crater. They went there because they think it might have been a lake bed at one time, from the surface features from above. He showed the tracks and moving around and looking at rocks, like the Adirondack. He said that is probably a bluish rock with dust on it - yes, that one is bluish. They used the RAT on it and determined it is basalt - lava rock. They're looking for hematite, because hematite is often formed by water.
...
http://www.badastronomy.com/phpBB/viewtopic.php?t=12076

There are several explanations for a blue subsurface on rocks, some biological, some purely mineralogical. Either way this would result in an important addition to our understanding of the (bio)geochemistry of Mars.
But first you have to acknowledge the blue colors are real.

Some possible mineralogical explanations are given near the end of this discussion on the Habitablezone.com bbs, notably by the posters "jms" and "dtb":

Space Sciences
Glassy material at the Spirit site.
Posted by Robert Clark on 3/28/2004 8:47:07 AM
http://habitablezone.com/space/messages/320917.html
(note: the other "Robert" in this discussion is the board moderator Robert Shepherd.)


Bob Clark





Also, a member of this bbs who attended the Lunar and Planetary Science Conference in March reported that Steve Squyres said the colors of the subsurface of the rocks was "bluish".
Bob Clark


Who was that? Just curious, I went to LPSC too.

I heard Squyres talk and he may have said rocks are bluish - I'm not doubting it, I just don't specifically recall it. In any event, saying rocks are "bluish" doesn't necessarily mean that they look that way visually. That phrasing, depending on the context, can mean that, spectrally, the rocks are skewed in that direction. A visually gray rock could be described as reddish if the spectrum is toward that end or bluish if in the other direction.

Maybe you have said this before but why is the existence of "blue" rocks on Mars seemingly so important to you?

RGClark
2004-Apr-04, 04:57 PM
And, please, could RGClark (and others?) who insert large images into their messages instead provide clickable links so that people who want to avoid spending the time to fetch images, can do so? Thanks.

Good point. I'll either provide smaller versions of the images or just give a link.
BTW, can I call you 105 for short? :-?

Bob Clark

01101001
2004-Apr-05, 04:35 AM
There are several explanations for a blue subsurface on rocks, some biological, some purely mineralogical. Either way this would result in an important addition to our understanding of the (bio)geochemistry of Mars.

So it seems you don't just want to know whether or not there is a hint of blue in the grayish rocks; you want to know exactly how much blue there is? You want the spectroscopic data?



But first you have to acknowledge the blue colors are real.


I would think that first, investigators would want to write their papers to explain their conclusions. The first thing they have to do is release the full science data to the Planetary Data System on the agreed-upon schedule, specified in MER Project Archive Generation, Validation and Transfer Plan (http://wufs.wustl.edu/missions/mer/docs/mer_archive.pdf). For Spirit, it looks like the data that documents any blueness in rocks, and its degree, from the first 30 sols will be transferred August 3, 2004.

Nirgal
2004-Apr-05, 11:41 AM
That the RATted area really is that white is suggested by the raw visible light filter images: this area appears bright in all of them:


This illustrates the main problem we have without the additional data from JPL:

Because with each of the color channles being contrast streched
(and or expsure time adapted) *individually* this is not necessarily so.
The original emission could still be much lower in the, say, blue channel
than the red one and only after exposure/contrast stretching the brightest part in the image would look "white" in the "raw-composite"

So, unfortunately, I'm afraid we can not really draw many conclusions
about the true colors of Mars without the hard data (exposure time and
contrast stretching factor of each individual channel)

Although JPL does already include a lot of meta-data encoded
in the image filename (Filter-Number, Timestamp etc.) the exposure time
is not among the published data :-(

Irishman
2004-Apr-05, 04:14 PM
Yes, I'm the one who made the comments about Squyres and the blue rock. Actually, I didn't attend the day conference at the LPSC itself, but rather the special presentation that was given that Tuesday evening at the University of Houston Clear Lake. So that could explain the difference in what we heard.

Specifically regarding the rock Adirondack, Squyres said it was a bluish rock. This was in contrast to the blueberries, which he described as visually more gray but spectrally blue. I think the images on the JPL site showing the yellow coating with the dark interior are basically what he's going by. The interior of the rock is dark and blueish, and the dust from the RATing looks fairly blue. I seem to recall Adirondack is basalt, and basalt can be fairly blue, so this is not extrordinary.