Soft video capture: A gratis method for capturing video on all Linux operating systems

The need for the visual presentation of material is increasing. While static graphs, images and charts can convey information in a thorough manner, video is of greater demand. Whether it be a tutorial for a software package, an online course for training, or a presentation of research material, nothing surpasses video as a tool for delivering information to your audience. Traditionally video capture on personal computers has been done in hardware. Normally an add-on board called a "video capture card" had to be installed. Software solutions are also available at cost, which almost always interfere with the rest of the processes running on a system. At Northrop Grumman the software models we develop are normally highly complex and resource - intensive. Nearly everything runs on Linux, with some of the servers having as much as twenty four processors and terabytes of volatile memory. Explaining to the customer what the results of my research are can be challenging at times. Scientific data visualization is a must, however it is often presented as static pie charts, bar graphs and line charts. With a tightening defense budget, reducing cost is an important goal. A process has been developed for capturing video on a Linux based system for no cost. All of the tools are freely available (some conform to the General Public License) and the impact on the operating system is little to none at all. The catalyst is the notion that Linux systems use X, whose entire windowing information is available at any given time. This information is simply copied to an alternate location in real-time. The postprocessing of the window information involves conversion of geometric data to image data. The individual images are then "glued" together and can be encoded with a variety of freely available codecs (including DivX, Mpeg1 and WMV). The resulting videos have been successful in demonstrating proof of concept. Also, it is of great use as a tool for communicating with end - users. Becaus- of its minimal resource utilization on a system, it can run as a background process while the end-user operates the foreground application. The resulting video of the end-user\´s actions are played back, and the human-error is quickly realized and corrected. This is exceptionally useful in dangerous, high stress combat environments. The end-results of this process have been successfully demonstrated to a captive audience and received much interest. This paper serves to convey the research involved and share it with the IEEE community. This process allows the end-user to capture real-time video on Linux for entertainment, informative and instructional use.