A method may include receiving video recorded with a camera and audio recorded with a microphone, wherein the video and audio were recorded simultaneously, and wherein the microphone is configured to move relative to the camera. The method may further include receiving a selection from a user of an object in the video, wherein the object is collocated with the microphone that recorded the audio. The method may further include tracking the object in the video and determining a location over time of the microphone relative to the camera based on the tracking of the object.