One of the thing that bothers me (and that has been somewhat posted before) is that almost all movies shared on the net are technically Data Container Formats (see »en.wikipedia.org/wiki/Media_container
for a crude explanation).
Typically (but not always - rare occurrence) a Data Container Format has the following components (oversimplified, of course):
1 - Header
2 - Data Streams (interleaved)
3 - Index
Typically (but not always - a rare occurrence) the data in the Data Streams cannot be understood without the Header and the Index.
Said Header and Index are physically located at the beginning and end of said Data Container, this is, at the beginning and end of a movie file.
The bittorrent process segments files into hundredths to thousands of "chunks". This means, that for a given movie, most of these chunks will be simply Data Streams.
Now, in a movie, a Data Stream is meaningless without the Header and the Index which indicates how to "play" the movie (i.e. how to execute the interleaved data streams by indicating how to de-compress them and their location in the Data Stream).
This is, if I share a chunk containing a piece of a Data Stream, that data or information *cannot* be construed as a part of a copyrighted work. Why? Simple: it is indistinguishable from garbage.
How do we recognize a "part" or "portion" of a copyrighted work? By looking at its similarities with the work. By getting a portion of the information available in the work. Does this happen with Data Stream chunks? NO.
Given "chunks" can one view a portion of the movie? NO.
Given "chunks" can one listen to a portion of the sound track? NO
Given "chunks" can one extract snapshots or pictures of the movie? NO
In other words, I could potentially download all of the movie but without the Header and the Index and STILL have garbage in my HDD.
So, if a chunk is totally and utterly unrecognizable from any part of the copyrighted work without the Header and the Index, what exactly are people "sharing"??? Basically garbage, NOT the copyrighted work.
So, a copyright holder would have to prove that you are uploading *the entire movie*, particularly the Header and the Index for it to be in a state of copyright infringement. Since without the Header and the Index, the file cannot be de-compressed dynamically and converted into a watchable movie.
This is not what Voltage claimed in their papers. Granted, they are *very* technically confusing indeed, but the most likely interpretation is that they got a few chunks from some people, checked their hashes, assembled 1 movie and viewed the movie against an original.
I cannot read their papers as they downloading 2300 full movies from all involved IPs and then comparing each of the 2300 movies against an original sample.
So, technically, they are claiming copyright infringement based on people uploading garbage!