Re: New blog post from Philip Hodgetts. Worth the read.

[Jeremy Garchow] "Sorry, I'm beginning to lose you. It's my fault.

We have a transcription service.

The method I am proposing is theoretical. You are also mixing the two different projects I talked about today, one interview based, and one scene/take based."

Not your fault, I was combining ideas from a couple examples as a shortcut to get at what I see as the core issue that this thread has raised. Apologies for the confusion.

I think it boils down to the question of how much of the editorial process can be automated vs how much will always require human judgement. I agree that flexible metadata in an NLE is a good thing. What I'm highly skeptical about is the notion that it cuts the post-production timeframe in half or more, as claimed in the Hodgetts article. I'm with Walter in wanting to see the proof.

I'm also skeptical of metadata driven auto assembly. Sure, maybe as a first pass instead of an assistant's manual assembly. But to do anything good beyond that, there's never a substitute for knowing the material. Knowing the material takes time. Less time, less likelihood that the very best material will always be known and used. I don't see this dynamic changing anytime soon.

[Jeremy Garchow] "Anything else we should over analyze today?"

LOL, this thread is awfully long, isn't it?

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl