Re: First draft of "MediaStream Capture Scenarios" now available

On 12/06/2011 02:45 AM, Travis Leithead wrote:
> As per my Action 461 [1], and as requested by this task force, I've
> put together the first draft of the scenarios for media capture [2].
Thanks, this is very valuable.

I have some comments, more or less in random order:

1. while there are examples for stream recording, taking pictures, there 
is no scenario for streaming in real time to another browser (i.e. the 
webrtc scenario). I think that should be added for completeness

2. 3.2, 3.3: Regarding what should happen if only a fraction of the 
requested capabilities (audio+video) is granted, or if a part of a 
MediaStream stops (e.g. because the camera is unplugged): I think we're 
moving much of the functionality of the MediaStream to the 
MediaStreamTrack level, this should make it more straightforward to 
handle these cases

3. 3.3: An additional sink is a PeerConnection.

I don't think audio preview is a problem. It would be a very badly 
written app that does not mute the audio if a video element is used for 
a self-view.

I don't think having a 1:1 mapping between source and sink; a simple 
example is that you would like a video element (for self view) _and_ a 
PeerConnection both be sinks for the same media stream; or for a multi 
party session several PeerConnection objects would be sinks of the same 
media stream

4. 3.5, 3.6 (Pre- and post processing): Is it necessary to split up? 
Given that a media stream can have several sinks (motivated above), 
can't the same tools as can be used for post processing also be used for 
pre processing? E.g. You can use canvas to detect motion, and start 
recording/streaming to a peer when there is motion.

And, one toolbox that should be added is the Web Audio API 
(<https://dvcs.w3.org/hg/audio/raw-file/tip/webaudio/specification.html>). 
I even think there are some implementations.

I would also like to know more about the "audio canvas"!

Stefan

>
> This draft is a combination of scenarios mingled with some
> commentary, issues, and points of discussion.
>
> Writing this document has given me a wonderful opportunity to
> mentally explore a variety of possible API designs and patterns. It
> also started pushing me in a particular direction which you might be
> able to tease out from the document. With the Chair's permission, I'd
> like to write down my ideas in the form of a media capture API member
> submission and send them to this group for further discussion. I
> would expect that it would take me about a week to get it all written
> down.
>
> Thanks, and feedback welcome.
>
> [1] https://www.w3.org/2009/dap/track/actions/461 [2]
> https://dvcs.w3.org/hg/dap/raw-file/tip/media-stream-capture/scenarios.html
>
>  -Travis
>
>
>
>

Received on Tuesday, 13 December 2011 14:40:00 UTC