RE: [MSE] Homework re: issues in the context of upcoming FPWD

> I think issue #19673 [1] would benefit the most from being included in the FPWD, and exposed to a broader audience.

Why would this bug "benefit the most"?  I am becoming quite concerned that we will never get a FPWD if we insist on getting everyone's "most important bug" solved in the FPWD.

In my view it is time to move to a "date driven schedule" for both the MSE and EME FPWD's.  We should pick a date and agree that we will all work to get as many bugs resolved by that date.  Anything not done by that date will simply wait for a subsequent WD which could be as soon as we want after the FPWD.

/paulc

Paul Cotton, Microsoft Canada
17 Eleanor Drive, Ottawa, Ontario K2E 6A3
Tel: (425) 705-9596 Fax: (425) 936-7329


-----Original Message-----
From: Pierre-Anthony Lemieux [mailto:pal@sandflow.com] 
Sent: Tuesday, December 11, 2012 1:47 AM
To: public-html-media@w3.org
Subject: [MSE] Homework re: issues in the context of upcoming FPWD

Hi all,

Per our last telecon, I looked at the bugs I filed in the context of the upcoming MSE FPWD.

I think issue #19673 [1] would benefit the most from being included in the FPWD, and exposed to a broader audience. Audio splicing is not trivial and has been visited many times before. In my mind, MSE should provide guidance to implementers based on existing practices and match the capabilities of existing consumer audio-video formats.

Specifically, while Sections 2.1.2 and 2.1.3 already discuss behavior at splice points, I think the specification would benefit from additional guidance in order to prevent audio artifacts and enable seamless splices. For instance, the statement "implementations must support dropping the old audio frame [...] and insert silence for the small gap that is created" seems to exclude non-frame-based audio and suggest inserting silence without a fade.

One approach would be to collect all splice point recommendations in an (informative?) Annex. I think the document referenced in issue
#19673 could be a starting point -- but I am biased of course.

Looking forward to the discussion. Happy to contribute.

Best,

-- Pierre

[1] https://www.w3.org/Bugs/Public/show_bug.cgi?id=19673

Received on Wednesday, 12 December 2012 09:23:38 UTC