© Joel Hagen
In 1993, I had the pleasure of participating in a conference on Planetary
Data Visualization put on by a research institute in southern California.
I showed a series of computer experiments in which generalized graphics
software was used in creative ways. A paint program, for example, can be
used as a tool of investigation and analysis. If you are an educator, this
approach can be a way to involve students in active explorations of art
and science on the computer. Mars is the thread linking my experiments
here, but it could just as well be European history or desert ecology.
In this column, I describe the series of experiments I did for presentation
at the Planetary conference. Although I refer here to Amiga softwrae, most
of these ideas can be explored using any good paint or image processing
software. Photoshop is ideal for these experiments. Don't focus on the
techniques so much as on the concepts of using software in new ways.
Cassini did a series of drawings of Mars in 1666. I brought these into the Amiga using an Epson scanner linked to ADPro. You could also use a video camera and digitizer. In Deluxe Paint, I cut the individual drawings of the planet as brushes and registered them on a sequence of animation frames. This brings a new dimension to data three centuries old as the features of Cassini's Mars move across its face, tracing its rotation. (For the animations on this web page, I put each drawing or photo on a different layer in Photoshop to register them, them dropped the saved RGB document into Gifbuilder to create the GIF animations). The same animation technique can be applied to any sequence of still images. Early mars images from astronomy books yielded animations of a moving cloud photographed in 1926 and a full rotational cycle shot from Lowell in 1971
I turned my attention to historic maps and scanned one Schiaparelli
had drawn from observation in 1877 (map #1 below). Using DPaint techniques,
I turned that map into a rotating planet. The method involves grabbing
an animbrush from a section of a moving map, then using it to Wrap Fill
a sequence of circles. The result is the illusion of a spinning planet.
The Mercator map projection distorts into a good approximation of a hemisphere
as shown in the illustration. In preparing your map you may need to add
a blank area north and south of the maximum represented latitudes if they
are less than 90 degrees. I have used this technique to "restore"
the planet as it was seen by astronomers over many decades. 
In January of 1993, I had the opportunity to paint Mars from direct eyepiece
observations at Lick observatory as Mars came closest to earth. I used
the Amiga to create reference material to aid my observation and painting.
Scanning a USGS Mars map derived from Viking missions, I used DPaint to
create a 24 frame animation of the spinning planet. The eyepiece of the
telescope inverted the image of Mars, but the Amiga made it simple to invert
the map and spin the planet in the proper direction relative to our viewing
instrument.
I printed out the 24 frames on a single page with hourly notations about
the central meridian. This provided a visual reference each night to interpret
the albedo (light and dark) features I was seeing. For Mars, 24 frames
was serendipitous as Mars' rate of rotation is similar to our own, a little
over 24 hours. For other planets, the number of frames would need to be
adjusted to match its rotation period in order to use the sequence as a
clock reference.
For instant color drawings, I scanned my pencil sketches then used a paint program to change the palette of the image from grey to a gradient of color matching what I perceived through the eyepiece. I also made detailed paintings directly in the computer. Later I produced large acrylic paintings for exhibition based on these renderings. I used the Amiga monitor as a reference while I painted.
The work with historical Mars maps got me thinking about perceptions
of Mars over the years. I decided to analyze several maps for consistencies
of perceived albedo features. The first task was to distort a group of
maps to the same dimensions and orientation. ADPro has an excellent interface
for cropping and scaling, but you can get workable results from any paint
program. I had to crop all maps to 60 degrees, the greatest common north/south
dimension among the group. I then used DPaint to cut and reassemble each
map, aligning them to the zero/zero coordinate near Sinus Meridiani. Photoshop
has excellent tools for this process.
My first comparison of map features was simply to load each one to a different
frame of an animation, then play them at high speed. Areas of consistency
were apparent in the rapid display. My next experiment was to merge two
maps at 50% transparency. I used ADPro's Composite options, but the same
experiment can be performed in any paint program that allows image merging
with controlled transparency. In the resulting image, areas that are either
light in both images or dark in both images remain light and dark. However,
areas that are light on one map but dark on the other average each other
out and become a neutral tone in the new image. This makes it easy to see
where the perceptions of different observers were consistent and where
they differed. It may also offer some insight into the changes in albedo
features over time.
Going a step further in map comparisons, you could use Image FX, ADPro
or ImageMaster (or Photoshop) to process each map image before combining
them. One observer might use much darker tones in his drawings than would
another. Image processing software can adjust the Dynamic Range, Contrast,
Gamma and Histograms of images to bring them subjectively closer to a consistent
base of values. In the same programs, images can then be combined with
new rules such as addition or subtraction of color values. This can reveal
new information. You can approximate these results in paint programs by
using Lighten or Darken or Shade modes to adjust overall value levels.
Creative use of color stenciling during a merge can also produce interesting
results.
I wrote a quick utility to compare two maps pixel by pixel. The program
weighed the difference in value of each pixel pair then created a new image.
If the difference was slight, it averaged the value for the new pixel.
If the difference was great, it substituted a very bright value. The illustration
shows the result of that process. Image #1 is Schiaparelli's 1877 map.
Image #2 is from a USGS map. Image #3 is the comparative map produced by
my analytic routine. I used Director2 which is no longer available, but
Basic or any good authoring software would work.
Continuing the computer Mars exploration, you can generate landscapes and
animations using VistaPro or Scenery Animator. This is an exciting option,
seeing regions of the planet from vantages otherwise inaccessible. On the
scientific side, an Amiga explorer can adjust the vertical exaggeration
of the data to see the computer landscapes without distortion. Every computer
generated Mars flyover you have seen on TV has been stretched vertically
for effect. You can take the law into your own hands and see a different
Mars reality if you so choose.
I have worked with a lot of educators eager to make use of computer graphics
in the classroom but at a loss for good projects. Perhaps this column will
suggest a line of thinking in which generalized software such as paint
or image manipulation programs can be used as tools of investigation and
exploration.
end