See also: IRC log
<glennd> Roll
<glennd> Last minutes 4-22-15 https://lists.w3.org/Archives/Public/public-web-and-tv/2015Apr/0006.html
<glennd> Use case review: https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Content_Capture#Capture_UC-4_3rd_Party_Composite_Asset_From_Multiple_Sources
The roll was called by Bill Rose
Glenn called for changes to the agenda and the minutes of the last meeting - no changes to either.
Glenn noted that there were a number of companies at the Streaming Media East Conference in NYC last week showing streaming content solutions, products and services. He asked that members reach out to these companies when they find them to join GGIE.
Bill Rose and Glenn Deen described a demo at the conference where a company was combing video and audio from multiple HDMI inputs on dedicated hardware to compose a TV experience.
The company was Skreens Entertainment Tech, out of Boston.
Glenn noted that the application would benefit greatly from the use of metadata associated with each stream which would allow the client/app to do far more than a simple HDMI connection can. It was also noted that accurate timing information is necessary.
Andrew noted that Cisco has been working along these areas for several years.
<kaz> ACTION: glennd to Look at the identification problems with respect to the ID of the content as well as an offset for the content for use in second screen applications. [recorded in http://www.w3.org/2015/05/20-webtv-minutes.html#action01]
<trackbot> Created ACTION-211 - look at the identification problems with respect to the id of the content as well as an offset for the content for use in second screen applications. [on Glenn Deen - due 2015-05-27].
<glennd> Bill led us through : https://www.w3.org/2011/webtv/wiki/GGIE_TF/UseCases/Content_Capture#Capture_UC-4_3rd_Party_Composite_Asset_From_Multiple_Sources
<inserted> scribenick: glennd
bill & andrew - a compositor could assemble the composite stream by reference not by directly sourcing the stream
andrew - would this be realtime or offline/timeshifted
bill - could be all three - realtime, next week
andrew - metadata could be included in video somehow - Question: does time of day of capture metadata get recorded in video
glennd - no not consistently, some cameras will add it in extended metadata, but professional composed/edited content does not do that.
andrew - metadata includes time of day, angle, location in stadium - all could be used to composite and present
glennd - in still photos GPS data is part of exif - don't know if video containers/codecs capture location etc metadata
glennd - potential GAP - need camera location metadata to reassemble 3D spacial positioning
bill rose- We need time too - so while cameras may capture it, it may get stripped - so also need way to associate it with the content ID so it can be retrieved
andrew - exif data in photos gets easily corrupted/lost during editing/transcoding - we need a way to find either the metadata or the identifier itself - doesn't need complicated image analysis.
glennd - potential gap - how to deal with a piece of content/stream that isn't marked at all by it's creator or distributor.
glennd - This maybe an implementors issue to deal with unmarked connected.
<inserted> scribenick: BillRose
Kaz - we should have an authentication mechanism to accept information from the 3rd party compositor in Capture UC4 to determine if the information provided is correct.
<azamlerc> I'm going to drop off because I need to run to another meeting. I'll summarize some of the use cases that we previously explored at Cisco in an email to the group. Thanks guys!
Glenn - you mean the camera may be mis-representing itself or the compositor provides misinformation.
Glenn - 2 gaps. No way to validate the information being provided. e.g. signing the metadata
Second gap - if the metadata is being referred to by the ID (metadata is stored externally for retrival) how can that metadata be validated?
Glenn - 3rd gap: If we are doing these references, lookup, etc., do we have a privacy issue that needs to be addressed? Clearly there is a privacy issue.
Glenn: New applications have made it possible to determine camera angles, height, etc., and use it to correct a picture but also to re-capture privacy information.
Glenn: Discussion on next call
and next steps. Noted that W3C is switching to WebEx in the
future. The next call is currently scheduled for June 3rd which
will still be on the current system. However he will be on a
plane so he will need to reschedule it. Glenn will send
notification as to potential new daters.
... Would like to shift work in mid-July to prioritizing the
Use Cases we will report on in October. First rank the Use
Cases and select the ones for the report. Then do a further
analysis of the selected ones and dig more deeply into the
gaps.
Then start to discuss the gaps, how they can be closed, and where the standardization work might best be done (which organizations).
Glenn: Current scheduled calls are for June 3rd, June 17, July 1st, and July 15th
<ldaigle> I will be there (IETF Prague). Bought plane tix yesterday.
<ldaigle> July 19 - 24
Glenn will be in the IETF meeting in Prague in July. Will others be there at the meeting and if so, let Glenn know and he will try to create an opportunity to meet there.
<alex_deacon> thanks glenn!
Glen closed the meeting at 12:10 PM ET.
<kaz> [ adjourned ]