08:33:16 <RRSAgent> RRSAgent has joined #inbandtracks
08:33:16 <RRSAgent> logging to http://www.w3.org/2013/11/13-inbandtracks-irc
08:33:17 <aizu> Present+ aizu
08:33:30 <xiaoqian_> chairs: cyril
08:33:33 <Noriya> Noriya sakamoto Toshiba
08:34:17 <olivier> olivier has joined #inbandtracks
08:34:24 <olivier> rrsagent, make logs world
08:34:34 <olivier> Present+ Olivier Thereaux (BBC)
08:34:46 <cyril> http://www.w3.org/community/inbandtracks/
08:34:53 <israelh> israelh has joined #inbandtracks
08:34:53 <cyril> http://www.w3.org/community/inbandtracks/wiki/Main_Page
08:35:09 <olivier> scribe: olivier
08:35:31 <olivier> cyril: the idea of this community group is to improve integration of media containers with HTML5 media elements
08:35:42 <Takahiro> Takahiro has joined #inbandtracks
08:35:53 <olivier> ... there is a section in HTML5 on how to create text tracks from tracks that are in a media file like mp4, webm, ogg and so on
08:35:59 <olivier> ... it is however not complete
08:36:01 <JatinderMann> JatinderMann has joined #inbandtracks
08:36:08 <JatinderMann> present+ JatinderMann
08:36:20 <olivier> ... for example in the case of mp4 it only talks about metadata tracks, but not subtitle trakc
08:36:54 <olivier> ... secondly, it is not easy to automatically map the video track to the video element. is it the main video, additional, etc
08:37:03 <olivier> ... so there is a need for better mapping
08:37:09 <olivier> ... onto the kind attribute
08:37:29 <olivier> ... third problem is that the tracks are not exposed if the format is not recognised
08:38:03 <olivier> ... e.g in the case of an audio codec, you could decode it in js, but you'd need the tracks to be exposed to js
08:38:14 <olivier> israelh: how would you guarantee syncing
08:38:22 <olivier> giuseppe: that's a use case on its own
08:38:31 <olivier> use cases
08:38:41 <olivier> 1. mapping of descriptors to kind
08:38:58 <olivier> 2. mapping of text tracks not yet covered by html5 spec
08:39:03 <olivier> 3. mapping of other tracks
08:39:18 <olivier> cyril: any thought? are any of the use cases more relevant?
08:39:35 <olivier> giuseppe: at least 2 other groups were trying to work on this mapping
08:39:43 <olivier> ... so doing it at w3c should avoid redundancy
08:40:00 <olivier> cyril: we started from a spec by cablelabs
08:40:05 <olivier> ... referenced by DLNA
08:40:12 <olivier> ... liaison with OIPF
08:40:22 <olivier> ... and there's the HTML5 spec itself
08:41:11 <olivier> israelh: does live broadcast fit in this?
08:41:24 <olivier> giuseppe: [scribe missed]
08:41:43 <olivier> MarkV: the idea is that multiple browsers get the same mediastreams
08:41:59 <olivier> ... thanks of mapping they'd give the same messages, same order, same data
08:42:15 <nigel> nigel has joined #inbandtracks
08:42:17 <olivier> giuseppe: there are hints in the spec, but not normative
08:42:21 <olivier> cyril: and not complete
08:42:57 <xiaoqian_> chair: cyril
08:43:03 <olivier> Igarashi: sufficient to support live?
08:43:07 <olivier> MarkV: we believe so
08:43:21 <olivier> ... will test by completing the mapping
08:43:56 <olivier> cyril: everyone understands and agrees with use cases?
08:44:21 <nigel> rrsagent, make minutes
08:44:21 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html nigel
08:44:24 <olivier> igarashi: switching between tracks is also in scope?
08:44:40 <olivier> MarkV: we believe there are mechanisms to switch between tracks
08:44:46 <michael> michael has joined #inbandtracks
08:44:57 <olivier> ... again, by completing the mapping will we be able to test that assumption
08:45:06 <olivier> cyril: there's also MSE which can be used for that
08:45:28 <olivier> q+
08:45:41 <olivier> cyril: don't think the switching should be in scope of this CG
08:46:06 <cyril> olivier: how do we envision keeping this up-to-date when new formats appear?
08:46:13 <cyril> ... registry or updates ?
08:46:41 <Youngsun_Ryu> Youngsun_Ryu has joined #inbandtracks
08:46:54 <olivier> MarkV: wonder if we could keep it in same way as those auxiliary page kept by the whatwg
08:46:57 <Zakim> Zakim has joined #inbandtracks
08:47:14 <olivier> cyril: we'd need to coordinate closely with mpeg, dvb and providers of track content
08:47:28 <olivier> ... at the same time I think we will need basic mechanism to expose the track
08:47:34 <nigel> q+
08:47:39 <olivier> ... and if a specific mechanism is needed we can refine
08:48:00 <olivier> igarashi: html5 media tf are currently thinking about a registry for MSE
08:48:15 <olivier> cyril: that is a registry to define what is a segment
08:48:22 <olivier> igarashi: we could use something similar
08:48:43 <olivier> nigel: thinking of the split between payload of tracks and wrapper
08:48:59 <olivier> [scribe missed question, sorry]
08:49:17 <olivier> cyril: what you need is timing info
08:49:38 <olivier> ... browsers have parser, you would'nt want to do the mp2 TS parsing
08:49:50 <olivier> ... in js
08:50:05 <olivier> MarkV: suggest looking at CableLabs spec
08:52:00 <olivier> cyril: looking at html5 spec, each track has id, kind, language...
08:52:19 <olivier> ... how do you set those attributes in an interoperable way?
08:52:31 <olivier> rrsagent, draft minutes
08:52:31 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html olivier
08:52:49 <Rossen_away> Rossen_away has joined #inbandtracks
08:53:21 <olivier> cyril: looking at proposal for language descripor from MPEG2
08:54:20 <Youngsun_Ryu> Youngsun_Ryu has joined #inbandtracks
08:54:27 <nigel> http://www.w3.org/community/inbandtracks/wiki/Main_Page
08:55:06 <olivier> cyril: been working on intermediate solution
08:55:16 <olivier> ... you could preprocess your file, extract and expose using WebVTT
08:55:27 <cyril> http://concolato.wp.mines-telecom.fr/2013/10/24/using-webvtt-to-carry-media-streams/
08:55:29 <a12u> a12u has joined #inbandtracks
08:56:21 <olivier> cyril: strawman proposal
08:56:25 <xiaoqian_> present+ igarashi , israelh , kinjim , michael , nigel , Noriya , Rossen, Takahiro , Youngsun_Ryu , xiaoqian_
08:57:38 <olivier> ... could put the payload as base64, or just an offset into VTT
08:57:47 <olivier> ... as a way to expose any kind of track to HTML5
08:57:56 <olivier> ... would like to replace that with an interop standard
08:58:14 <olivier> ... ideally would generate not a text track cue but a data cue
09:00:10 <olivier> cyril: one problem is size
09:00:27 <olivier> ... you'd end up with whole stream in memory
09:00:45 <olivier> ... also, events. you'd receive an event for each cue
09:01:08 <olivier> ... ok for subtitles, might be too many at a time if using with HFR video
09:01:27 <olivier> igarashi: is it just for text data?
09:01:41 <olivier> cyril: any kind. text, xml, json, audio, video...
09:02:07 <olivier> ... you could use to support depth of video even if not natively supported
09:02:16 <olivier> ... so I would like to expose any kind of track
09:02:44 <olivier> ... and let the browser decide
09:03:00 <olivier> ... e.g. Eric Carlson says that they don't have access to all the tracks on some devices
09:03:33 <olivier> israelh: so wouldn't be limited to certain types of tracks
09:03:44 <giuseppep> giuseppep has joined #inbandtracks
09:04:20 <olivier> MarkV: our scope is dealing with standard media streams
09:04:40 <olivier> ... it's already possible to add data to a track
09:04:56 <olivier> ... also if browser implementation gives you access to data tracks
09:05:12 <olivier> ... generally it's the audio and video that are not
09:05:38 <olivier> cyril: generally not the case. all goes directly to the hardware
09:05:54 <olivier> sheau: another issue is that timing information is convoluted
09:06:13 <olivier> ... without solving sync, don't you end up having independent track which end up diverging
09:06:46 <olivier> giuseppep: sync is a different concern
09:07:16 <olivier> sheau: @@
09:07:39 <Kiyoshi> Kiyoshi has joined #inbandtracks
09:07:52 <olivier> cyril: you could get as a track all the data with all the timing
09:07:56 <olivier> ... and do sync
09:08:13 <olivier> igarashi: video is decoded by the UA
09:08:37 <olivier> ... so you'd have a delay between that and the timing info in your track
09:09:26 <olivier> ... because you don't have access to the timing of the player
09:09:32 <olivier> ... and you can't control it
09:09:43 <olivier> cyril: you can query currenttime
09:09:49 <olivier> igarashi: not accurate
09:10:31 <olivier> ??: not accurate, 100ms off
09:10:36 <olivier> sheau: and it would drift
09:10:42 <nigel_> nigel_ has joined #inbandtracks
09:11:12 <olivier> cyril: agree sync is a concern, but there is a need to do all the work on accessing the tracks first
09:11:19 <olivier> s/??/Sylvia
09:11:23 <Noriya> Noriya has joined #inbandtracks
09:11:39 <olivier> cyril: CG has just started
09:11:52 <olivier> ... start with work in Wiki
09:12:07 <olivier> ... decide whether to extend HTML5 spec or include
09:12:40 <olivier> nigel_: does this help with situation where video is delivered in mpeg2 and data is provided out of band
09:12:49 <olivier> cyril: if your timestamps match, yes
09:12:59 <olivier> sylvia: and if there's no drift
09:13:17 <olivier> nigel: question is whether we're exposing the right data
09:13:32 <olivier> ... fundamental question is how do we sync
09:13:39 <olivier> sylvia: the UA does that for you
09:14:03 <olivier> ... you could do it yourself but we're not talking about that
09:14:12 <olivier> cyril: we are talking about using the sync engine in the browser
09:14:28 <olivier> MarkV: you could imagine writing an editor in browser
09:14:49 <olivier> ... but if you just have 2 tracks, use what is provided by the browser
09:15:10 <olivier> cyril: question about scope
09:15:19 <yo> yo has joined #inbandtracks
09:15:21 <olivier> ... MP4, MPEG2 TS, Matroska and OGG
09:15:24 <olivier> ... also RTP?
09:15:33 <olivier> ... DASH? HLS?
09:15:34 <silvia> silvia has joined #inbandtracks
09:15:49 <olivier> ... think we may want to start with simple formats?
09:16:12 <olivier> MarkV: do we need to treat Matroska and WebM separately?
09:16:35 <olivier> DavidD: Matroska/WebM. it's just a subset
09:16:40 <olivier> ... webm is a subset
09:16:46 <HJLEE> HJLEE has joined #inbandtracks
09:16:56 <olivier> israelh: request for clarification
09:17:16 <olivier> ... is it about getting tracks, or preprocessing?
09:17:27 <olivier> MarkV: the goal is just to define standard mapping
09:17:49 <olivier> ... to HTML5 standard APIs exposing those tracks
09:18:13 <olivier> ... our work is not needed if you're just playing
09:18:42 <olivier> nigel: if you are thinking about DASH, a DASH player would be able to do all this already
09:18:58 <olivier> MarkV: you could extract some data and do additional things with it
09:19:18 <olivier> nigel_: managing sync of media tracks playing back is really hard
09:19:49 <olivier> ... question is whether w3c should be defining something that is already defined elsewhere
09:19:55 <olivier> giuseppep: it's a mapping exercise
09:20:01 <olivier> cyril: the first step is mapping
09:20:06 <olivier> ... second step is sync
09:20:14 <olivier> ... is it preprocessing, post processing?
09:20:35 <olivier> israelh: that implies it's at playtime?
09:20:50 <olivier> MarkV: typically, yes, but it would be whenever you get the data
09:21:07 <olivier> cyril: if you can preload you could grab and prcess
09:21:12 <olivier> ... would still be best effort sync
09:22:53 <olivier> cyril: admin point - where do we host the work?
09:23:09 <olivier> silvia: wiki is a good start
09:23:58 <cyril> ilivier: form my experience, if you choose early what you want to use, you have to show DoC
09:24:09 <cyril> some groups have been happy using GitHub and issue tracking
09:24:19 <cyril> if you want to use GitHub, start early
09:25:01 <olivier> silvia: PLH has mentioned need to have bugs in w3c space, but there are talks of api
09:25:25 <Rossen__> Rossen__ has joined #inbandtracks
09:25:26 <olivier> nigel: question of WG?
09:25:35 <olivier> cyril: no, the CG does the work, targets the HTML Media TF
09:25:52 <olivier> cyril: there is also a mailing-list associated with the CG
09:26:03 <olivier> ... we've been discussing mpping for the id attribute
09:26:14 <nigel_> http://www.w3.org/community/inbandtracks/
09:26:36 <olivier> giuseppep: editorial comment - would be more readable if attributes were in rows, and formats in cols
09:27:15 <olivier> MarkV: (in response to Igarashi) - the goal is for this work to replace the cablelabs spec
09:27:35 <olivier> cyril: every track interface has an id attribute
09:27:49 <olivier> ... how do we map between container id and the id of the text track
09:28:11 <olivier> ... so far we looked at media fragment identifiers
09:29:12 <olivier> giuseppep: @@
09:29:20 <xiaoqian_> RRSAgent, make munites
09:29:20 <RRSAgent> I'm logging. I don't understand 'make munites', xiaoqian_.  Try /msg RRSAgent help
09:29:31 <xiaoqian_> RRSAgent, make minutes
09:29:31 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html xiaoqian_
09:29:32 <olivier> rrsagent, make minutes
09:29:32 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html olivier
09:29:57 <olivier> cyril: I will report a good discussion, ~25 people. agreement on use cases.
09:30:14 <olivier> ... discussion on second priority use case for sync
09:30:21 <olivier> ... agreement on scope
09:30:30 <olivier> [adjourn]
09:30:36 <olivier> rrsagent, make minutes
09:30:36 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html olivier
09:31:09 <silvia> Silvia Pfeiffer, NICTA
09:32:04 <israelh> israelh has joined #inbandtracks
09:32:07 <olivier> (quick round of intro, including latecomers)
09:32:14 <olivier> rrsagent, make minutes
09:32:14 <RRSAgent> I have made the request to generate http://www.w3.org/2013/11/13-inbandtracks-minutes.html olivier
09:32:24 <israelh> Participant: israelh, Rossen, Jatinder (Microsoft)
09:32:34 <nigel_> nigel_ has left #inbandtracks
09:40:08 <Takahiro> Takahiro has joined #inbandtracks
09:45:13 <silvia> silvia has joined #inbandtracks
09:56:30 <Takahiro_> Takahiro_ has joined #inbandtracks
10:09:19 <olivier> olivier has left #inbandtracks
13:28:12 <Zakim> Zakim has left #inbandtracks
15:18:44 <silvia> silvia has joined #inbandtracks