[ Updated : Scroll down for people who allready read the entire document ]
For a more scientific/mathematical view on "True colors on mars" please read :
I hope this post will help adress the countless "conspiracy style" webpages and postings I encounter each day, and that have cought the attention of the ill-informed and ignorant alike. The people who start these theories often did quite a bit of researching and experimenting before coming to the statements, so they are just from their point of view. Everyone has his area of expertise, and image capturing / postprocessing / color calibration is not one that is very easy, nor a very widespread one for that reason.
I write this little bit of clarification from the point of view of an experienced image postprocessor/photographer, and not that of a scientist, there have been enough fact-finders that use the power of math and science to explain it, but that does not seem to be enough.
My experiences on this area came from working for Canon, where I was tought a lot about the inner workings of CCD's and imaging devices such as scanners, and from a personal point of view, as being a photographer.
To simplify matters, I have made two sections - theory, explaining some of the knowledge needed to get to the conclusions I made in a very simplified yet still quite lengthy form, and some simple photos to show it after that. Skip the theory if you're not the type of person to read long stories.
Start of theory/explaining
All we are talking about here is "Color calibration" aka "whitebalancing".
These two words basically discribe the nightmare of all digital camera manifacturers, ever since the first was tested to capture a photo.
CCD's (CMOS etc, any electronic recording device) and film suffer from an issue that they cannot "change" the way they capture light.
The result of this is that with changing types of light (e.g. tungsten, lightbulbs, sunny daylight, cloudy daylight etc) the final image is rendered incorrectly (since the camera or film is programmed/designed to record at a certain color temperature by default)
Most analogue films are "daylight balanced", which means they render pretty accurate colors durning regular daylight, being anything lit by the sun. Try them in artificial light, and the results can vary from a light colorcast to green, purple, yellow etc.
The same thing goes for CCD's - the digital counterparts of film.
However the data they capture we can call "RAW", can be color profiled and processed in-camera, adapting it to whatever the camera is set to.
This color profiling is by no means a perfect art - There are countless variables that will mess up the presets for whitebalancing, and the only near-perfect way to calculate the type of light you are working in is to use a "graycard" - Fill the camera frame with a special type of standardized gray-card held in the same light you are shooting, set colorbalance to that, and voila..pretty good color rendition (If said recording device can actually adapt to shooting at such low or high color temperatures - measured in Kelvin)
The CCD's in the spirit and opportunity are, albeit more sophisticated, still cursed with the same basic problems that the cheapest webcam has to cope with, and this is the source for most "false color" theories I encounter on the net.
The images that get returned from the Rover's are actually 3 black and white captures - since the CCD cannot distinguish color! These 3 photos were taken through colorfilters - usually this would be one green, one red, one blue shot, and then merged together to form the final color image.
However the red filter is replaced by a filter that is more tolerant to near-infrared, and hence manually merging the RAW data from nasa WILL yield incorrect colors, unless you know how much to tweak the "Infra"RED image to balance for its extended infrared light capture. Yet people still try merging the 3 B&W images at home, and come up with skewed colors.
This process of color calibration is not an automated task - The sundial with it's 4 colored specs and gray-toned circles helps the people down here to understand what certain colors look like under given light up there, with the exception that the specs of color contain a pigment that looks quite different from regular paints when looked at through a near-infrared filter, again part of NASA's method to determine perfect whitebalancing. (or, according to the conspiracy theorists, the perfect way to mislead us so obviously!)
I do think that NASA did have this upon themselves, since the colors of the initial images released on the web differed quiet a bit, from pale to darkred - a nightmare for an image postprocessor like myself, and I cursed the unexperienced / timerushed person at NASA who allowed the images to be released in this state, at the time. In retrospect it possibly is due to pure timeconstraints and "public pressure" - SOMETHING had to get out, and it had to get out NOW, not in 1 hour, or two.
End of theory/explaining
Now, the photos.
Let's start with one of the other big sources of misunderstanding and misguidance :
Recent NASA press conference with color photo of Mars
As you see the photo is devoid of most red's
1) "Photo-of-a-photo" :
Whenever you capture a photo of something, your camera must be balanced to the lightsource at that location (99% of all modern day journalists use digital SLR cameras for their photographs). Judging the faces, the camera was properly whitebalanced to whatever light was used there. Updated : The photo itself isn't even properly toned - the faces are pale, and devoid of any human red/brown colors, this makes it even more appearent that the background shot was of a different whitebalance to begin with.
THIS MEANS IT CANNOT BE WHITEBALANCED AT THE SAME TIME FOR THE PROJECTION ON THE BACKGROUND, WHICH IS MADE OF LIGHT WITH A DIFFERENT COLOR TEMPERATURE! .
result : properly colored faces, anything not made of or shed upon by the same light will look WRONG/DISCOLORED.
2) Projectors have a tendency to wrongly calibrate/convert input to start with...if anyone here has ever used projectors for photos/images you will know it is by no means easy/possible to get things looking accurate.
MORE FACTS : What color does the MARS surface really have?
Just take a look at the photos made by the Hubble telescope, and you see the colors of mars as seen out of space - This is the color we have to compare to that photographed by the Spirit/Opportunity in proper daylight (sunset/sunrise will give skewed colors, either warmer or cooler).
Some people came up with photos from mars that had a blueish cast over them, and the source was...NASA. I did some research on this and came to the unsurprising result that the Hubble has as much problems color calibrating itself as any other CCD made by mankind. However, we can see a unique opportunity to compare older and more recent, technologically more advanced postprocessing / colorcalibration, due to several imaging-specific sattelite-based missions sent to mars after 96.
EXAMPLES : (As collected through simple google image searches "Hubble Mars 2001" for example)
1991 : Early hubble image I stumbled upon :
1995 : Calibrated with the knowledge of mars at that time, with sattelite data from dating back to the 70's (remember : nearly 2 decades were without sigificant sattelites sent to mars!)
1997 : The Global surveyor was sent to Mars, the first proper sattelite since many, many years. All images from NASA have a distinct change in color (more reddish) since that year (98-99 etc)
2001 : Mars Odyssey was sent out, specifically to capture the surface of mars (imaging mission). A host of new techniques and electronics to establish better color calibration of nearby mars for NASA.
2003 : A recent photo made by the Hubble...not unlike that of 2001, and very much like the colors we have witnessed on Spirit Photos - To my knowledge the spirit is located in the lighter colored areas, not on the poles (which have, generally, a much darker soil)
And for pure reference, A photo from the recent ESA (Mars Explorer) images, independandtly from NASA :
The sky in ONE of the NASA images (a panorama) _was_ doctored ...this was easy to check up, since the entire sky color was of exactly the same RGB value - and a rather silly way of trying to make it look better (problem with panorama's is that in parts shot into sun direction the sky tends to bleach out tremendously, giving you uneven skycolors which looks "crap")
The colors of the JPG's on the JPL website have been inconsistend to say the least - some were too dark, some were reddish, some yellowish. Please take a look at the properly calibrated TIFF's before judging. Also over time the images are corrected, where we can make the simple conclusion NASA is not trying to cover up, just fixing what they did wrong. Why not spend 1 day more per image to get them properly colored if this all would be a huge cover-up-and-desceive mission?
I dont feel I have put enough information in this posting, since there are a lot of more subtle things I could say to show people that there is no evil doctoring going on with the images and or color calibration, but the main conclusions are noted above.
[ UPDATED SECTION ]
1) Changed "photo of a photo" to update my thoughts about the photo itself - after looking at it for a little longer, it seems as the entire photo doesnt even have a proper whitebalance set (either in-camera or afterwards for processing it for the web). The red channel in that photo is quite badly represented, with the faces of the people being a bit pale and devoid of color - this magnifies the allready yellowish effect from the projection.
To clarify this for even the most stubborn people out there, I decided I'd just show a little "real life" situation - My monitor (9300 Kelvin) and a regular tungsten 150 watt bulb (about 2700 Kelvin). My camera had a tough job capturing the extremes, and especially the 9300K (very cool) monitor setup I have was giving him some trouble. Camera was a Fuji S2 12 mpixel digital SLR, for thos who care to know (not much unlike the cameras used by journalists, although these are mostly nikon/canon)
2) This image was created for "everyone" so apologise for my rather basic and non-technical usage of words and color/shapes ;-)
Most of my thoughts are embedded in the image, albeit shortened and without much detail. I noticed new discussions, where people were complaining about the atmosphere of mars being "blue" when viewed from outer space (Hubble) and "hence you can never have a red sky on the NASA/JPL ground images" .