I'd like to combine a few data sources for telemetry overlays over a video. The MP4s have some metadata embedded from my GoPro but I have a few other data sources I'd like to include. I already have a way to generate a PNG of the data visualizations. What I am wondering if is ffmpeg includes a way to, every 16ms to take the current video frame, get the corresponding metadata values for that frame embedded within the mp4, send them off to the image creator, take that output and overlay over the video frame.
My assumption is that the mp4 metadata will correspond to the frame of the video. In that video frame's metadata the GPX values that I will need to generate the visualization.