Solutions
TODO: Document in detail solutions videokollektiv members use.
Debconf/Pycon style:
Emphasis is on live editing and implementing a programmatic workflow of dealing with video capture/editing/transcoding/uploading. The biggest mistake people make with conference recording is to plan to do manual post-production in a non-linear video editor - it is completely unrealistic. Take Pycon US, for example; 5 days, 5 rooms, approx 8 hours a day of video = 200+ hours of video to edit. If mixing from multiple sources, probably 400+ hours of editing work. As much as possible needs to be automated, and a workflow/team built to suit individual tasks effectively - avoid trying to do anything post-conference. Using this workflow and suitable software, we're able to get videos out rapidly - typically within a few days of the conference ending, or if enough volunteers are assisting, during the conference (next-day, or same-day depending on infrastructure). We use several important things:
* live video/audio mixer saving to a digital file - i.e. DVSwitch - software DV mixer - http://dvswitch.alioth.debian.org/wiki/
* Conference video QA/editing/encoding/uploading software (encoding, publishing) - i.e. Veyepar (Pycon) - http://github.com/CarlFK/veyepar
* A defined workflow, training, senior team members with conference AV experience, etc - some docs/info @ http://www.uanywhere.com.au/pycon09_video/