Might be simpler to do on the recording side but should be relatively easy to do as a processing step when a lecture is uploaded