[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ProgSoc] Film and PCs (was: physical to electronic)
> On a technical note - someone mentioned that at 4000x2000*? came out
> to something like a CD's worth of data per second of film - which
> implied it was infeasible to work with it (in the context of non-linear
> editing), but now it's being stated that film actually *is* manipulated
You can work on it, sure, but not in real-time. That means adding digital
effects is OK but you can't use a CPU as part of the pipeline in your
actual distribution and projection system (until the CPUs get faster which
they obviously will). Also it means that digital storage is not suitable
for most normal packaging requirements (you don't want to be driving a semi-
trailor full of CDs into the cinema). Again digital storage density is
going up up up, so just a little time should fix the problem.
> Does this mean they work on only a few seconds of film at
> a time, and/or just how big are the machines that they do this kind of
> thing on? (Terminator II, Jurassic Park style films, I mean.)
Well the classic example would have to be Titanic which used a room
full of alpha boxes running linux kernels and wired into something similar
to a Beowolf cluster (though the software that did the actual calculation
was completely private so there is no way to tell how similar it was to
a Beowolf machine in function). From memory it took about three months
but someone may have better memory than mine. Exactly how much of the
film was computer generated, how much was models and how much was stage
furniture is anyone's guess.
> always assumed that kind of work would be intense, tedious,
> intensely-tedious ..
Exactly the sort of work you would give a computer to do.
You are subscribed to the progsoc mailing list. To unsubscribe, send a
message containing "unsubscribe" to email@example.com.
If you are having trouble, ask firstname.lastname@example.org for help.
This list is archived at <http://www.progsoc.uts.edu.au/lists/progsoc/>