The Real Face on Mars

Recently in the news, we received from Mars new pictures of the supposed "Face" taken April 5,1998. It turns out to be a pile of rocks, as everybody with half a brain already knew.

On April 9th, I happened to catch the Art Bell show, with guest Geologist Jim Berkland, and a quick visit with James Hoagland. The main thesis of the evening was trying to debunk the new Martian photos. Their arguments: 1. The picture was taken at 1024 resolution instead of 2048 resolution like most of the other pictures of Mars. 2. The raw picture looks very dark with big streaks across it, to quote Art Bell, "They've got some good image software to get the above images from this?"

I'm not a spokesman for NASA, but the 1024 resolution was almost certainly used because of the lack of good sunlight at the time the picture was taken. When photographers do not have enough light to shoot a picture in, they can increase the aperture, slow down the shutter, or use film with lower resolution, I am sure this is what NASA did.

As to the lousy gray resolution of the picture, the original raw photos of the "face" from the Viking mission, were only slightly better. I can prove this with the histogram of both pictures, but I will not bother here.

As to the good image software needed to clean up the ugly raw data, I can do it easily with Paint Shop Pro 5 (Shareware $99, a free 30 day trial version can be downloaded from http://www.jasc.com ). I have much better programs that I could use (Corel Photo Paint 7 or Adobe Photo Shop which is what NASA used), but I want to prove that you do not need a complicated or expensive program to perform this kind of enhancement.

The ugly raw data can be downloaded from http://mars.jpl.nasa.gov/mgs/target/CYD1/ I went ahead and downloaded the 9 megabyte high resolution TIFF version. Also note that at the bottom of this web page, it describes how NASA did their own enhancement.

The portion of the raw data that contains the "face" is (0,4444) to (1024,5644). Copy and paste this portion into its own file.

I then took a sample of the raw data that is largely featureless (0,1200) to (1024,2400). Copy and paste this portion into its own file, as well. Then make a negative (colors/negative image) of it.

Then, combined the two (Image/Arithmetic set to Addition and Divisor = 2). what you end up with is a gray blur with no vertical scan lines. This is almost exactly how it was done by NASA.

Then, increase the contrast (Colors/Adjust/Contrast increase to 90%), increase the gray scale range (Colors/Histogram Function/Stretch), then get rid of the spotty look (Image/Normal Filters/Soften) and you end up with a pretty good picture. It is flipped the wrong way and distorted. The flipping is caused because the picture was take through a mirror. The distortion is the result of the picture being taken at a 60 degree angle instead of head on.

Next, we need to correct orientation and distortion. Orientation is easy (Image\Mirror). Distortion is a little more difficult. Use the crater in the lower left (after mirroring) as a guide, all craters are naturally circular in shape. To fix this: 1. (Image/Rotate Direction Right, Free, 40 degrees). 2. (Image/Resample, Turn off Maintain Aspect Ratio, set Custom Size to 1555x1100) 3. (Image/Rotate Direction Left, Free, 50 degrees) 4. Crop the image as desired. The result is very similar to the one NASA released.

Note that you can kind of make out a face on this picture (Jay Leno?), but on the full size version it is obviously just a mountain range.

And, if you think this is a lot of enhancement to get this image, the original Viking "face" raw data required just as much enhancement to be cleaned up.

Face on Mars update (April 1999): Despite all the evidence to the contrary, James Hoagland and company still think the recent face picture was faked by NASA. Furthermore, NASA has taken pictures of other controversial objects, like the pyramids on Cydonia and the "town center", all of which are very natural phenomenon.

Back to the Glossary
Back to Archives

The portion of the raw data that contains the "face" is (0,4444) to (1024,5644). Copy and paste this portion into its own file.
I then took a sample of the raw data that is largely featureless (0,1200) to (1024,2400). Copy and paste this portion into its own file, as well. Then make a negative (colors/negative image) of it.
Then, combined the two (Image/Arithmetic set to Addition and Divisor = 2). what you end up with is a gray blur with no vertical scan lines. This is almost exactly how it was done by NASA.
Then, increase the contrast (Colors/Adjust/Contrast increase to 90%), increase the gray scale range (Colors/Histogram Function/Stretch), then get rid of the spotty look (Image/Normal Filters/Soften) and you end up with a pretty good picture. It is flipped the wrong way and distorted. The flipping is caused because the picture was take through a mirror. The distortion is the result of the picture being taken at a 60 degree angle instead of head on.
Next, we need to correct orientation and distortion. Orientation is easy (Image\Mirror). Distortion is a little more difficult. Use the crater in the lower left (after mirroring) as a guide, all craters are naturally circular in shape. To fix this: 1. (Image/Rotate Direction Right, Free, 40 degrees). 2. (Image/Resample, Turn off Maintain Aspect Ratio, set Custom Size to 1555x1100) 3. (Image/Rotate Direction Left, Free, 50 degrees) 4. Crop the image as desired. The result is very similar to the one NASA released. Note that you can kind of make out a face on this picture (Jay Leno?), but on the full size version it is obviously just a mountain range.