Traditional video mashup and summarization methods assume that all video clips have common audio, but with varying quality. Hence, selecting the best quality audio is sufficient. In this work we explore a new scenario ...