W3C home > Mailing lists > Public > public-media-capture@w3.org > June 2015

Re: Proposal to include channelCount constraint/capability

From: Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com>
Date: Sat, 20 Jun 2015 04:49:42 +0000
To: Joe Berkovitz <joe@noteflight.com>, "public-media-capture@w3.org" <public-media-capture@w3.org>
CC: Audio Working Group <public-audio@w3.org>
Message-ID: <1447FA0C20ED5147A1AA0EF02890A64B1D289697@ESESSMB209.ericsson.se>
Thanks for the proposal. One question in line.

On 19/06/15 00:22, Joe Berkovitz wrote:
> Hello Media Capture,
> I'm writing to you with the first of several proposals from the Audio WG
> that relate to our earlier Last Call comments. We've worked with the
> Media Capture co-chairs to revise our earlier ideas and make sure they
> fit in cleanly with the spec as it stands today, and with the history of
> previous decisions by the task force.
> Summary:
> The proposed change will associate a channel count with an audio
> MediaTrack, allowing both device discovery (via constraints) and device
> description (via capabilities and settings) to account for the number of
> audio channels in a track.
> Rationale:
> Channel counts are fundamental descriptors of audio tracks, as width or
> height are fundamental to video tracks (for input, at any rate). Knowing
> the potential and real channel counts of device streams are essential
> for audio-based apps to present or make meaningful choices between
> devices, and to adjust an application's behavior to best fit a given device.
> The Audio WG did consider using an enumerated type for channel layouts
> (e.g. "mono", "5.1", etc.) but has settled on channel counts because
> they are always knowable and discoverable in the UA's native
> environment, while knowledge of actual channel layouts are neither
> universally available nor standardized. Channel layout constraints could
> be added in the future if needed.
> Definitions:
> The channel count of an audio track is the number of independent
> channels of sound that it contains, i.e. the number of audio samples per
> sample frame.

And I guess that when using the channelCount constraint with 
getUserMedia it is a way to indicate to the UA what input audio device 
that would work best.

A question regarding multichannel audio input devices (I start thinking 
microphone arrays): are they usually capable of operating in different 
modes (e.g. an array of 5 microphones can deliver 5, 3 or 1 channel audio)?

> Proposed changes:
> - Add a *boolean channelCount* attribute to MediaTrackSupportedConstraints.
> - Add a *long channelCount* attribute to MediaTrackSettings
> - Add a *long/LongRange channelCount* attribute to MediaTrackCapabilities
> - Add a *ConstraintLong channelCount* attribute to MediaTrackConstraintSet
> Thank you very much for your consideration.
> Best,
> .            .       .    .  . ...Joe
> *Joe Berkovitz*
> W3C Audio WG Co-chair
> President, *Noteflight LLC*
> 49R Day Street / Somerville, MA 02144 / USA
> phone: +1 978 314 6271
> www.noteflight.com <http://www.noteflight.com>
> "Your music, everywhere"

Received on Saturday, 20 June 2015 04:50:13 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:26:33 UTC