[compsciphd]

and its easy enough to figure out who ripped the web-dl, because you can have each frame copied into 2 slightly different but functionally identical versions (or perhaps just a longer GOP, but still the same). Create 32 of these sets of 2 (i.e. bits, or even do it for every 32 sets in the video) and assign every user a unique set of 32 "bits" and you'll have the ability to uniquely identify every web-dl (as by definition they aren't modifying the streams) to the user who downloaded it and just a cost of 1x the storage vs not doing anything.