Why does the attacker need a second leak? Can they not just randomly clip each scene by a few milliseconds? This would defeat the identification, and not be noticeable to a viewer.