- From: Travis Leithead <travis.leithead@microsoft.com>
- Date: Thu, 15 Mar 2012 00:01:46 +0000
- To: Anant Narayanan <anant@mozilla.com>, "public-media-capture@w3.org" <public-media-capture@w3.org>
>-----Original Message----- >From: Anant Narayanan [mailto:anant@mozilla.com] > >On Mar 14, 2012, at 4:40 PM, Robert O'Callahan wrote: >> I think it would make sense to have VideoMediaStreamTrack (like >AudioMediaStreamTrack already), and APIs on video tracks to request setting >of various camera parameters (resolution, focus mode, etc) (async of >course), and an API to request a snapshot with/without flash and various >other parameters set momentarily (also async). The snapshot API should >probably trigger a callback passing a Blob which can be used as an <img> >source. It will be interesting to see how the "camera parameters" can be presented using our contraints/capabilities model being discussed. >This seems to be along the lines of of what Rich suggested earlier in the >thread as well. > >We've already kicked around the idea of a MediaStreamRecorder to capture >snippets of video from a stream, we might need a similar object to capture >images from a stream. I'll come up with a rough proposal for these and send >them to the list. +1 I've suspected that we need specific track types in the media stream to be able to segregate video and audio-specific APIs. Unless we expect a variety of different categories of track types, I think it would be safe to follow HTML5's HTMLMediaElement.audioTracks / .videoTracks concept and fork the tracks to make script processing easier. Earlier I drafted up a "MediaEncoder" concept [1] that might serve as inspiration for you (regarding drafting up a rough proposal). Ideally, I would like to see just one interface that can handle snapshots as well as video, but there was enough differences when I was sketching this out before that having two different interfaces seemed inevitable. Perhaps you'll have better luck :-) [1] http://lists.w3.org/Archives/Public/public-media-capture/2011Dec/0043.html
Received on Thursday, 15 March 2012 00:02:22 UTC