During 1986 – 1987, my partner, Douglas McGuire, and I undertook the task of developing a suite of commercial software programs to facilitate the generation of accurate three-dimensional (3D) data from video camera measurement systems. The result of this effort was AMASS (ADTech Motion Analysis Software System) which included components for camera linearization, system calibration, automatic marker tracking at the 3D level, 3D marker identification, and a graphics program (ADG) to display the final results which were in the format of C3D files. I must thank the Biomechanics Laboratory at the National Institutes of Health (Bethesda, Maryland), and in particular, Dr. Lynn Gerber and Dr. Steven Stanhope, for providing encouragement and support through of laboratory facilities that enabled the project to be completed.
Shortly after its completion AMASS was licensed to Oxford Metrics Ltd. (Oxford, England), and sold independently to a number of biomechanics laboratories. The subsequent introduction and success of the VAX/VMS based Vicon-VX system by Oxford Metrics resulted in the widespread use of AMASS and C3D files within the biomechanics community
In the past, several factors have contributed to prevent a still wider acceptance of the C3D file format. The first was the lack of thorough and complete documentation of the file structure and parameter contents by the AMASS manuals. The second, partially resulting from the first, was an insufficient understanding by programmers of the capabilities and flexibility of the file structure. This lack of understanding resulted in some attempts to put “round pegs into square holes”, and generated a legacy of C3D files and applications that digressed from the original format and intention. Many of these files and their applications are still around today and cause considerable problems for programmers who wish to handle every C3D file. A third factor was that a formal standard for the format was never established or universally agreed upon, resulting in uncertainties for programmers trying to implement it. In my estimation, this manual should go a long way towards belatedly overcoming all of these shortcomings.
While I no longer have any commercial interests in the C3D file format, I will continue to be available to provide assistance and support for its promotion as a tool for all those who wish to use its capabilities.
May 2002
Andrew Dainis, Ph.D.