So, On Thu, Aug 10, 2017 at 06:18:55PM -0400, Wouter Verhelst wrote: > As an aside to that, I think it also makes sense to do some internal > evaluation. Did things work as expected? Do we need to fix things > somewhere? What worked, what didn't work?
This. As for my bit, I think sreview performed as I had expected it. There is still room for improvement (especially in that ffmpeg defaults to pretty horrid resolution settings for VP8, hadn't noticed that), but other than that it worked well. Unless people disagree, I intend to use it for future events, rather than veyepar. However, I'm not planning to disable the veyepar instance on vittoria, so it can still be used if needs be. I think the cameras worked well, we chose good. Their zoom is okay, their low-light performance is great, nothing more you'd want than that. As I've said a few times, we should move away from MPEG2 encoding and onto something that is HTML5-compatible. MPEG2 means sreview can't use the original encodings for the previews, which makes it take a lot of time to generate a preview. Additionally, MPEG2 takes a lot of diskspace, a lot more than the DV recordings did (and those were horribly inefficient). FOSDEM has had great results with H.264; Mark van den Borre can tell you more about that. We should investigate this at the sprint in Cambridge. Any other thoughts? -- Could you people please use IRC like normal people?!? -- Amaya Rodrigo Sastre, trying to quiet down the buzz in the DebConf 2008 Hacklab _______________________________________________ Debconf-video mailing list Debconf-video@lists.debconf.org http://lists.debconf.org/mailman/listinfo/debconf-video