Thu, 23 Feb 2006

Never Let It Be Said I Think Too Much...

So, here's the apparently most recent A Scanner Darkly trailer (direct link) (via Boing Boing). Looks kinda interesting but is also, I suspect, rather ripe for a disappointing execution.

Given the questions raised by the trailer, it seems an opportune time for a list of them...

Top Ten List (II)

  1. Have I become merely a pawn in GlobalMultiNationalBigMedia's attempts to own culture by propogating this link?
  2. Is Keanu forever (doomed to be?) type-cast in near-future techno-dystopia fantasy movies?
  3. Was it necessary to use the term techno-dystopia so gratuitously? (Easy answer: Yes.)
  4. Who will be the first person to mashup Johnny Mnemonic, The Matrix(ii), A Scanner Darkly and Point Break into a futuristic techno-surfer-dystopia?
  5. Are we already living in it?
  6. Will A Scanner Darkly be painfully disappointing?
  7. In what year will it be legal to create the mashup described in (4)?
  8. Is this a shameful attempt to justify viewing ASD as soon as it's released in New Zealand?
  9. Does it bother anyone else how many surveillence cameras there are in New Zealand and how little attention they seem to receive?
  10. Does anyone care at all about this list?
  11. Winona Ryder?? (Bonus.)

Auto Movie Mashup Generator

Speaking of the recent trend of movie mashups, a few years ago I wrote a Python toy that attempted to automatically split online comic strips such as Calvin and Hobbes into their component panels and then reassemble them randomly. I got partway through the process but obviously got distracted by something else before completion and release. (I've since seen actual implementations of similar functionality that someone took the time to complete and release... :-))

Anyway, I wonder when a practical implementation of an automatic movie mashup generator could be implemented. e.g. You specify the actors you want in the movie and it pulls segments together from different movies. Would it need some sort of meta-data annotation or could visual and audible recognition systems enable it to extract actors?

It could be a fun project. Although legal mis-adventures would seem inevitable. You could start by creating manually assisted mashups requiring lines of dialog to be user selected.

Posted at: 01:55 | category: / | Comments ()