Techniques for transcoding a video with frame-synchronous metadata are described. A portion of a digital video transmission is obtained, the portion including a first sequence of video frames, a plurality of metadata values, and a first set of timing data relating the plurality of metadata values to the first sequence of video frames. A rendition of the first sequence of video frames is encoded, the rendition including a second sequence of video frames based on the first sequence of video frames. A plurality of segment files for the rendition is generated, each segment file including a group of frames from the second sequence of video frames and a group of metadata values, each metadata value in the group of metadata values having a correspondence to a frame in the group of frames based on the first set of timing data.