Mars in 3-D is a stereographic film of imagery taken by the NASA Viking 1 and 2 spacecraft from both Mars orbit and on the surface, from both lander locations, between 1976 and 1979. The original film was produced at Stanford University in 1979 using 16mm film. It is a valuable and unique historical presentation of the results of one of the United States' most important space achievements of the decade following the Apollo moon program. After being shown around the world at scientific conferences and concerts, the film reels and related materials were donated to the NASA archives at Ames Research Center in Mountain View, CA.
The twin Viking spacecraft arrived at Mars in 1976 and operated for several years. In 1979, one member of the Viking Imaging Team, Dr. Elliott Levinthal, was by then working at the Stanford Medical School developing medical imaging technology. He received support from NASA and Stanford University to produce a scientific documentary based on the 3D imagery from the spacecraft.
Each spacecraft had both an orbiter and lander component. Each lander had two cameras separated by about 0.8 meters, which when used together could produce left/right stereo images. The original purpose of this was to determine precise distances to nearby features for programming the soil scoop arm, but it also proved useful in understanding the overall geology of the surrounding landscape. It also really brings home the tremendous human and technological achievement of seeing the surface of another planet up close for the first time in human history. And although the tests for life on Mars were then deemed inconclusive, the results are still debated by scientists today, with some arguing for a positive interpretation.
For the orbital views, two images taken by the moving spacecraft from slightly different locations were used to produce the stereo images. The exaggerated 3D that this provides dramatically reveals the topography of the large-scale surface features. The film also includes more conventional stereo images of the Viking test lander taken at the Jet Propulsion Laboratory in Pasadena, and some non-stereo scenes of the narrator on-screen.
In 1979, CCRMA was the leading center in the world for developing the use of computers for musical synthesis and composition (and arguably remains so). Dr. Levinthal approached Professor John Chowning, then the directory of CCRMA, for help in producing a stereo soundtrack for the film. Prof. Chowning turned to two of his graduate students, myself and Mr. Schottstaedt, and we each produced about 15 minutes of music for the 32 minute film. This was an inspiring opportunity for me, having been an amateur astronomer since childhood and an avid supporter of the NASA manned and robotic exploration missions. In fact, astronomy was my original career goal.
In 2009, CCRMA proposed a concert to honor its founder and former director, Professor John Chowning, who proposed to present the concert in a movie theater and include the Mars in 3-D film. However, using 16mm projection with the original reels was out of the question due to the degraded quality of the film and soundtrack, and difficulty using outdated 16mm stereo technology. So, with the cooperation of NASA/Ames we began an effort to find, restore and convert the film to modern Digital Cinema format and 5.1 surround audio, for presentation using current 3D cinema projection technology. Several copies of the 16mm left and right film reels, as well as the original narration audio tapes, were located at NASA Ames. These were then scanned to HD video.
The original quadraphonic surround music has been re-synthesized using a software emulator, built by Mr. Schottstaedt, that exactly reproduces the functionality of the hardware synthesizer we used in 1979. This, along with digitized copies of the original narration recordings, were used to create a new 5.1 surround audio soundtrack.For the restoration, each reel (left and right eyes), was first processed according to the following steps:
After this, several major problems still remained. First, the SmoothCam filter, while successfully removing most of the jitter, had adversely affected the left-right eye convergence both vertically and horizontally, and differently so for each scene. Second, the original film had some inherent 3-D "fusion" problems arising from the wide separation of the lander cameras and the exaggerated 3D of the orbiter images. These scenes need to be "pulled in". Third, some of the 3D scenes involved left and right images that were shot at different times of day, leading to problematic differences in shadow orientation and color between the two. These latter problems made it difficult for some people to fuse the 3D in some Mars scenes even in the original movie, and really needed to be improved.
Trying to correct all these 3D problems using normal FCP tools quickly proved highly impractical. Instead I used the Stereo3D Toolbox plugin from Dashwood Cinema Solutions. I primarily used the Geometry and Image Balance plugins for convergence and final color and contrast matching of the left and right eye sequences. The plugin also allowed me to do this in one pass and automate various parameters, since I was basically operating on one long sequence with pre-existing scenes and cross-faded transitions, but the parameters still vary from scene to scene, and sometimes within one scene.