W3C home > Mailing lists > Public > public-web-and-tv@w3.org > December 2011

RE: [MPTF] ADR_Min_Control and minBandwidth

From: Jan Lindquist <jan.lindquist@ericsson.com>
Date: Fri, 16 Dec 2011 10:01:20 +0100
To: "Lewis, Jason" <Jason.Lewis@disney.com>
CC: "public-web-and-tv@w3.org" <public-web-and-tv@w3.org>
Message-ID: <82276AE38FD87A4C9CF6C820AC5276EA358FC3E44E@ESESSCMS0362.eemea.ericsson.se>
Hi Jason,

With my understanding of bandwidth or bit-rate I would question if this is the appropriate parameter. You hint at the association between the manifest file and that is what I understood is not the purpose of specifying the bandwidth. It is bandwidth agnostic. If we want to explicitly control the selection of representation in the manifest I suggest to use the maxLevel and minLevel as I propose in the other e-mail thread. I am very much in support of controlling the selected representation both min and max in the manifest but that is the other thread.

The reason I split the e-mails is to take the issues independently so they do not get mixed up and create incorrect expectations.

Hope I clarified the reasoning behind my proposal to remove minBandwidth.

Regards,
JanL

________________________________
From: Lewis, Jason [mailto:Jason.Lewis@disney.com]
Sent: den 16 december 2011 09:09
To: Jan Lindquist
Cc: public-web-and-tv@w3.org
Subject: Re: [MPTF] ADR_Min_Control and minBandwidth

I think minBandwidth should be kept.

It enables some valuable use cases.

It is useful to manage a single manifest file with bitrate representations across multiple devices (ie phone & tablet). The UA can be hinted to avoid the
lowest bitrates that may look acceptable on an phone, but are not on a tablet.

It could also provide the ability to set a target'starting' bitrate to optimize quality or startup time.
--Jason Lewis


On Dec 15, 2011, at 11:43 PM, "Jan Lindquist" <jan.lindquist@ericsson.com<mailto:jan.lindquist@ericsson.com>> wrote:

Hello,

To have a minimumBandwidth does not make sense from a control of the the stream perspective. If the stream is of type VBR there is going to be some very odd behaviour with increased quality levels at low bandwidth consuming portions of a stream. If comparing to what is described in Silverlight the max and min related to tracks which I interpret as the levels relating to the manifest. The support of maximumBandwidth is makes sense and should be kept but not minimumBandwidth.

Regards,
JanL
Received on Friday, 16 December 2011 09:01:56 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:44:06 UTC