[ Thanks to Brendan Scott for this
link. ]
“I have just had a look at a heap of videos from a
variety of conferences. They all make the (I think) mistake of
assuming video of the presenter is in the least bit relevant. If
you can’t see the slides (by which I mean actually read everything
on the slides – I’m looking at you absolutely tiny text shell
prompts and text editors showing code) then, assuming the presenter
is not using wild gesticulations which add meaning to their
presentation, the video might as well be plain audio. The value of
video is to sync what the speaker is saying with the slides. Seeing
the presenter shuffle from side to side looking up and down
occasionally (as charming as it might be) does not add anything to
the presentation. A 20MB download is better than a 200MB one if
they’re otherwise equivalent modulo video of a shuffling speaker.
Often they get the worst of both worlds, zooming out to get both
the speaker and the slides – and getting a poor
representation of both of them”