A high definition surveillance video storage optimization system executes purging directives according to policy and retention meta data. Segments of video are purged from storage when thresholds of retention metrics are applied and when policy permits. Retention metrics are determined from acoustic sound energy, radiation energy, access control events, and movement. Policies reflect calendars and elapsed time, percentage of available storage, statutory and judicial mandates, and utility of the stored segments. Extrinsic sensor data and operations are recorded into non-transitory retention meta data stores. Metrics are transformed into retention flags at either time of recordation or time of purgation. Security file storage includes metadata about the environment acquired by sensors include doors, audio noise, object identification and tracking and inputs which influence retention or purging. A circuit filters for motion and stores metadata per segment. Meta data enables content based purging of least consequential files.