Screenshots and movies
Making movies from the 3D brain is a semi-experimental feature in CVU. It will only work on linux systems with ffmpeg installed, though we may investigate implementations on other platforms in the future (additionally, it will only work on systems using the X11 display server rather than Mir or Wayland). Moreover, we have observed some differences in movie quality depending on the hardware. We have not tested this functionality on a wide range of hardware -- however, we suspect it has to do with the graphics card/driver.
Of course, instead of using the builtin feature, it is entirely possible for the user to use screen capturing software of any variety to capture the entire usage of CVU, rather than just the 3D brain.
The options available for making movies are described as follows:
######Framerate The number of frames per second of the resulting movie
######Bitrate The number of kilobytes per second used to encode the movie. Using higher values will cause the movie to be recorded at a higher quality and also to take up more disk space.
######Automatically Rotate When checked, a simple animation will play during the movie which will cause the camera to rotate around an axis. Further interactions are possible so that the user can move the camera around manually. This rotation is in addition to the animation.
######Rotation degrees When "Automatically Rotate" is checked, this parameter determines the number of degrees in a circle that the brain is rotated each movement. For a smoother animation and longer video, make this lower.
######Animation speed Adjust how quickly the animation moves from one state to another. This is independent of the framerate of the movie. That is, mayavi renders the scene and uses builtin animations to determine the rate at which the animation moves. ffmpeg captures this screen output and encodes a movie from this output at a certain framerate, but the sample rate of the movie has nothing to do with the sample rate of how often the animation is rendered.
######Debug ffmpeg ffmpeg has the annoying behavior of printing no text whatsoever to stdout and instead prints everything verbosely to stderr (a design choice I do not understand at all). This causes a lot of text to be printed in routine transcoding usage even when there is no error. When running cvu from a console this behavior is quite annoying since a whole lot of text is printed whenever making a movie, which precludes being able to read other error messages in the console. By default cvu suppresses this output and prints nothing. When this box is checked, cvu will print out ffmpeg's stderr stream to the console instead.