On Wed, 2005-04-20 at 10:02 +0200, Alexander Schmehl wrote: > Hi! > > * Gunnar Wolf <gwolf@gwolf.org> [050419 21:09]: > > > A _very_ important point. Also, if audience < 100, many people prefer > > not to use a mike, even if specifically asked to do so, so they have > > both hands free. An unobtrusive, wireless microphone will get us more > > recordings - Of course, if possible we should add to it a decent > > enviromental (omnidirectional?) mike, as it'd be worth it to record > > the attendee's interventions as well. > > And please remind the speakers before their talk to use the mice and to > repeat questions for the mice. I once did a talk, which was recorded, > but was quite useless, since they just told me about it the day before, > and I forgot to put on mice and repeat the question in the time > passed... > This is a good point to remember. In my local LUG I've been trying to get all the speakers to restate questions and comments from the audience for the benefit of the recording. If the main presentation room is big, restating is a good idea regardless of whether it's being recorded or not. Odds are that someone in the audience couldn't hear it. We really need to focus in on what we're striving for though. There are three different ways we can go about this: (1) Live streaming + recording of the streams. (2) Very simple recording (1 camera per presentation with no real editing) and a rapid turnaround time for putting the presentations online. (3) More complex recording and editing with a longer turnaround time. We've already discounted (1) as a primary goal because no one has experience with it. I take it that (3) makes some people skittish because timeliness is an issue and there are worries that complex editing will never get done; (3) is what I've been aiming for at my local LUG. So if we're shooting for (2), rapid turnaround as top priority, then we'll end up with 1 camera panning between the speaker and the screen. You film the presentation in one take, dump it to a PC using kino, run it through mencoder to create a web-suitable clip, put it online somewhere and save the original Quicktime DV files in case there's any call for a higher quality version. If a single camera shoots 8 hours of video in a day, the same camera will need to do 8 hours worth of playback that night. The odds of any given consumer DV camera playing a tape from another consumer DV camera are not perfect. It will probably take another 8 hours or more to do some basic clipping in kino and run the video through mencoder, though this could be done on a different machine. If we don't rush to complete this during Debconf I have a friend who works in the A/V department of a local company and is willing to help dump tapes to PC and edit the footage. He has access to a professional playback deck that will eliminate concerns about compatibility but doesn't speed up the dumping process any. Professional mini-DV decks are still 1x speed from what he tells me. For wireless vs wired mics, we use wireless at HLUG so that the speaker is free to move around without worrying about the mic. There is occasionally RF interference on the tapes, but it has never ruined a presentation. Pack 100 laptops and cellphones in a room and it might be a different story. When will the schedule be announced? It would help to know exactly how many hours worth of footage we'll be dealing with to decide what level of complexity is realistic. There are a wide variety of ways we could make better recordings (multiple cameras, multiple mics, titles/credits, denoising the audio track, two-pass encoding, you name it.) It's really just a matter of how much time will be available. John
Attachment:
signature.asc
Description: This is a digitally signed message part