Mars Rover Picture Analysis Discussion

Skeptic,

good points.

the reason I put this link up is really two things:

1. It was a voyager image from 1979
2. It was an image taken near one of the poles

thus, a differing area and timeframe where NASA could have been doing things differently?

Is the image legit? I have no idea.

but the sky hue vs the ground is not out of the bounds of human experience on earth, but valid for mars is annother issue.

Originally posted by ArchAngel
In the new image THE SUNDIAL IS CUT OUT!

Wonder why.............

Blessings,
Mikromarius

I enjoyed reading Kano's description of the origin of the problems with the color images released from Mars Spirit.
Perhaps someone can confirm some other color images I've produced doing a RGB color filter composition.
The raw filter images are taken from this page:

Panoramic Camera :: Sol 012 (63 images)
marsrovers.jpl.nasa.gov...

The ones I used were the ones near the top that show the JPL logo in them. I used L4, L5, and L6. My question is that I got some bluish areas on some of the rocks by just using a straight equal combination of the 3 filter images, not brightening any of them up over the others. I got the same result using L2 in place of L4.
I know however that since these separate filter images are not brightness normalized in the same fashion you won't get reliable results by combining them equally.
The applied normalization data is not available yet though. However, since the image shows that JPL logo, I thought we might be able to use that to provide color calibration.
NASA just released a color image of this area:

A Puzzling Patch
Jan. 20, 2004
www.jpl.nasa.gov...

Their image does not show the blue areas on the rocks. There is a difference in the color of the JPL logo in their image from mine however.

Does anyone know of any color images of the rover that shows what the actual color of that JPL logo is, say while on Earth?


Thanks,

Bob Clark

Originally posted by Kano
The ones NASA is releasing, basically.

Your analysis doesn't justify producing an image where any blue or green colors on the surface would be similarly distorted as they are on the calibration sundial.


Bob Clark

Originally posted by rgclark
Your analysis doesn't justify producing an image where any blue or green colors on the surface would be similarly distorted as they are on the calibration sundial.

Thats because the reason the sundial colors are distorted (as mentioned) is the pigments are designed to give different brightness patterns when viewed with the different filters. The actual blue and green channels from the sundial are still there, and can be seen if you combine just the L5 and L6 filters. So even if there was something blue/green on the surface that was extra bright in the near-IR range (very unlikely really) it would still be noticed.

Dr. Bell was kind enough to send me a graph of the spectra of the pigments used. I will post it after I get home from work tonight.

Three simple questions for Kano or someone else with knowledge about this:

#1 Why don't they simply send RGB images through space instead of several monocrome plates in the first place?

#2 I would also like to know how it is even possible to make blue become somewhat neon red without manipulating the plates extencively or using color filters in order to hide something?

#3 Why do they still rely on radio waves (which by the way shouldn't even work in zero gravity theoretically ) when we have laser technology which would make them capable of sending high quality RGB colored images back to Earth in notime?

As it looks to me, the blue hue has either been turned about 180 degrees rendering it in a slight sephia tone. Either that or the pictures have all been shot using a red-purple filter, then later tweaked in order to make the images somewhat believable. Or both. Atleast that is what my evidence points towards when calibrating the pictures to the original, now cutaway sundial in the panorama shot.

Blessings,
Mikromarius

[Edited on 20-1-2004 by mikromarius]

i moved this from the other thread, because i think it's a tell-tale sign of subterfuge.

Originally posted by darklanser
I tried to get some other color references for the Spirit craft. This is a photo of the balloons taken on Earth. Does it look a bit modified like the images on Mars. It has way too much red. Take a look and decide for yourself.

way too red looking. looks like "mars". everything in the room looks red. it looks filtered. why would this earth picture have the same hues as mars, unless they were making sure the baloons look to be the same color on 'mars' as they do on earth?

kano, i don't think you've been getting enough praise, so here's mine....
AMAZING WORK! YOU'RE BRILLIANT. even before true color calibration.

here's a question, though. could that not be arizona, or afghanistan? could they not fake these images. can not the individual channels be doctored before publishing them to the web? i'm not asking you to say they are or were. just to surmise that it's possible. you see, i'm a skeptic.

The blue is purple/red.. the green is nowhere to be seen..
Nice explaning kano but that still doesn't explain those horror methods of sending colors back to earth.. NASA heard of hdri?? (floating point format.. for bimbos - that means you can have a color like -- red : 98384 green : 12,567 blue : 0,001)
i want to have a hdri pic from mars for further diagnostics

nasa sux.. the more blue sky image from mars looks much more natural to me like those others....
too bad i'm very bad from space stuff.. does mars have atmosphere?? because theroeticaly if we know what stuff is in the atmosphere we would know what colour it has when lit by sun.

Originally posted by Kano

Originally posted by rgclark
Your analysis doesn't justify producing an image where any blue or green colors on the surface would be similarly distorted as they are on the calibration sundial.

Thats because the reason the sundial colors are distorted (as mentioned) is the pigments are designed to give different brightness patterns when viewed with the different filters. The actual blue and green channels from the sundial are still there, and can be seen if you combine just the L5 and L6 filters. So even if there was something blue/green on the surface that was extra bright in the near-IR range (very unlikely really) it would still be noticed.

Dr. Bell was kind enough to send me a graph of the spectra of the pigments used. I will post it after I get home from work tonight.

Thanks, but you know the question is whether the Martian sky really is blue. What would really be convincing re your argument is if someone took RGB filter images of Earth's blue sky only replacing the Red with a near-IR filter and found for example its color did not change, i.e., stayed blue. If on the other hand its color changed to reddish like Mars sky appears in the released NASA image then that would raise the possibility that using this near-IR filter to produce a visual light image really did change the color of the Martian sky and perhaps the surface as well.
I"ve heard there are digital cameras that can take images in the near infrared. So perhaps this is something someone can try.
Also, there really is no reason by this time NASA could finally release an image using the correct L4, L5, L6 filters that shows the calibration sundial and the Martian sky and surface.
Perhaps you could bring this point up with Drs. Adler and Bell when you talk with them.


Bob Clark

Originally posted by rgclark
Thanks, but you know the question is whether the Martian sky really is blue. What would really be convincing re your argument is if someone took RGB filter images of Earth's blue sky only replacing the Red with a near-IR filter and found for example its color did not change, i.e., stayed blue. If on the other hand its color changed to reddish like Mars sky appears in the released NASA image then that would raise the possibility that using this near-IR filter to produce a visual light image really did change the color of the Martian sky and perhaps the surface as well.
I"ve heard there are digital cameras that can take images in the near infrared. So perhaps this is something someone can try.

Indeed, but I imagine we would come up with the same problem of calibrating the exposure levels when trying to combine a near-IR image with the blue and green channel. Perhaps another way would be to find a spectral graph of the blue sky from earth. We could compare the brightness at 600nm with the brightness at 750, if one is greater than the other, then we will know by how much shifting the red channel would affect the color of the sky.

As far as showing the sundial, ground and sky in one shot, I don't think the angles allow for this, but the sundial does have its little mirrors on the sides, which show the martian sky.

Originally posted by mikromarius
#1 Why don't they simply send RGB images through space instead of several monocrome plates in the first place?

#2 I would also like to know how it is even possible to make blue become somewhat neon red without manipulating the plates extencively or using color filters in order to hide something?

#3 Why do they still rely on radio waves (which by the way shouldn't even work in zero gravity theoretically

) when we have laser technology which would make them capable of sending high quality RGB colored images back to Earth in notime?
[Edited on 20-1-2004 by mikromarius]

1. Its effectively the same thing. Plus theres no advantage to be had doing the processing on site when it can be done just as easily, and far better on earth.

2. Read the article, also see the new graphs sent by Dr. Bell.

3. Why wouldn't radio waves work in zero g? I don't think laser communications are advanced enough for this sort of application yet. Plus the NASA Deep Space network is already in place, why fix it if it aint broke.

today 21 january I am watching on this moment to live broadcast of mars update.. and they talk about the color of mars and that they don't know the exact color yet... this broadcast will be in a rerun until next update...



[Edited on 21-1-2004 by MarkLuitzen]

Originally posted by Kano

As far as showing the sundial, ground and sky in one shot, I don't think the angles allow for this, but the sundial does have its little mirrors on the sides, which show the martian sky.

Well the original panoramic image that created the controversy did show the surface and sky in the same shot with the sundial:

marsrovers.jpl.nasa.gov...
[large image 4.1 mbytes]
My guess is that they can still get a similar shot now that the rover is up and moving.


Bob Clark

Originally posted by rgclark
Well the original panoramic image that created the controversy did show the surface and sky in the same shot with the sundial:

Ah sorry I was thinking of the individual pictures, the panorama's are mosaics, and yes I would suppose a mosaic with L4 L5 L6 channels will occur before the end of the mission.

The panorama shots do however take up quite a bit of the daily bandwidth. (Judging by the raw data folders). So I suspect such panoramas will be less and less frequent as Spirit gets its teeth into the work at hand.

(Or its 'whirling blades of doom' as it were )

Thanks for the awesome discussion and explanation, Kano. I stumbled on this debate on Martian Soil yesterday and have been searching for answers.

Originally posted by rgclark
I've heard there are digital cameras that can take images in the near infrared. So perhaps this is something someone can try.

Most of digicams are sensitive in near IR. Interestingly, lots of the more expensive pro and pro-sumer cameras aren't as sensitive since they have much better internal filters that the cheaper cameras.

Everything you wanted to know about shooting digital IR (and then some) can be found here at DP FWIW's digital IR page.

Would be interesting indeed to see equivalent shots of a known quantity. Nice suggestion!

Tungsten
www.poweredbysteam.com...

Kano, Mars Opportunity is scheduled to land Friday.
Let's hope this time the first released panoramic image will show the calibration sundial and the Martian sky and surface using the correct visual light filters.


Bob Clark

I agree Bob, although I wouldn't expect the firsrt panorama from opportunity to not use the L2.

Thunderstorm Delays Rover's Work on Mars.
"On another issue, mission members said the Red Planet, as seen in gorgeous panoramic shots from the rover, may not be really red.
The photos of bright red dust, rusty rocks and salmon sky were color-balanced to approximate what a person might see standing on the martian surface, but it may be weeks before scientists perform the calculations to show true color, said Ray Arvidson of Washington University, a scientist on the project.
Mars was sort of a medium chocolate brown in photos taken by the Viking landers in 1976, he said."
abcnews.go.com...

Really, I don't think it's that hard. Just use the same color combination you used to get the image of the sundial that showed the color tags with the correct colors.
Do the same thing except have both the calibration sundial and the Martian sky and surface in the picture at the same time.


Bob Clark

Yes but they are talking about calculating the exact colors. We can see for ourselves that there won't be a dramatic change. But using the equalized channels in the sundial shot still removes any tint given by the atmosphere, which wont be much, but is still relevant when we are looking for precise colors.

there's still a red balloon elephant in the room.