usual annual question: will there be audio/video recordings at an Internet Research conference? usual annual response: 'no' usually phrased as 'the conference group/exec will discuss it', which usually means 'how much will it cost?', which resolves into, 'well, yes it can be done, but only if you donate a large sum of money, and or do it yourself', which usually resolves into 'no, it will not happen.'
folks who're interested in videoing IR should spend a couple of hours reading the blogs of the folks who have been doing the video of the Debconf series of conferences. [starting point - http://layer-acht.org/slides/20070122_debian-meetings-archive_SLUG.pdf ] they've invested SERIOUS time, hardware, and resources into making it work. it is a *hard* problem - completely nontrivial. I like video as much as the next person - probably more - but this would be a huge committment.
Personally, I'm against recording any session other than keynotes. I think photographs, blogging, and a back-channel are great additions, but public recording of session stifles the communal atmosphere where people can be frank and collegial.
Pretty much on-board with what jeremy just said. The backchannel is particularly useful, IMHO.... --e