Re: Action item. definition and use of Clean audio in European television

Amazing how much easier it is to find information when you can search on
the name everyone else has been using! Thanks, Sean, for ferreting out
this critical datum.

Yes, I believe we should adjust our docs and requests. No reason to go
against the flow.

With respect to independently adjusting volume of channels (or their pan
positions, etc), may I suggest we've asked for that in our User
Requirements?

     * (DV-5) Allow the author to independently adjust the volumes of
     * the audio description and original soundtracks.
          * (DV-6) Allow the user to independently adjust the volumes of
   * the audio description and original soundtracks, with
          the user's settings overriding the author's.
             * (DV-8) Allow the author to provide * fade and pan controls to be
      * accurately synchronised with the * original soundtrack.
     * (DV-11) Support the simultaneous playback of both the described * and non-described audio tracks so that one may be
            directed at separate outputs (e.g., a speaker and headphones).

Janina

Silvia Pfeiffer writes:
> On Sat, May 14, 2011 at 1:33 PM, Silvia Pfeiffer
> <silviapfeiffer1@gmail.com> wrote:
> > Ha, that's funny: have we been using the wrong term all the time?
> >
> > In out requirements it's called "clear audio" (and that's also what we
> > have requested in the bug, see
> > http://www.w3.org/Bugs/Public/show_bug.cgi?id=12544 ). In these
> > documents it's called "clean audio".
> >
> > I think we will need to change our documents/request for @kind label...
> >
> > I've now (using the proper name) also found this document:
> > http://www.etsi.org/deliver/etsi_ts/101100_101199/101154/01.09.01_60/ts_101154v010901p.pdf
> > which is "Digital Video Broadcasting (DVB); Specification for the use
> > of Video and Audio Coding in Broadcasting Applications based on the
> > MPEG-2 Transport Stream" . Annex E talks about Clean Audio as a
> > supplementary audio service in the form of receiver-mixed data. This
> > is in contrast to broadcast-mixed delivery, which would be as a
> > separate alternative audio stream, which DVB does not support.
> >
> > So, it seems that while we are asking for a @kind="clean audio" label
> > on audio tracks that are delivered separately to the main audio track
> > and provide a better understandable audio mix, this isn't actually
> > what DVB supports (and I assume DVB is the only way in which clean
> > audio is currently realized?).
> >
> > So, the question is: what do we do in HTML5 to support the kind of
> > clean audio tracks that DVB expects.
> >
> > IIUC, the receiver-mixed means of delivery of clean audio (and audio
> > description for that matter) delivers it through multi-channel audio
> > but within the main audio track. This means, on a 3, 5.1 or more audio
> > channel file, some of these channels are reserved for non-main audio.
> 
> Actually, that's incorrect for audio description. A "supplementary
> audio" channel is an additional audio channel, so it would take up the
> number of audio file with stereo, they would deliver 3 channels, for
> one with 5.1 it would have 7 channels etc. I think though for clean
> audio it is still true that they use center channel for speech only in
> a 5.1 file.
> 
> Now the big question is: how common is all this. And would we
> encourage people to rather provide multitrack than multichannel for
> accessibility needs?
> 
> Cheers,
> Silvia.
> 
> 
> > There is a marker of "visual impaired commentary" for AD and "hearing
> > impaired" for clean audio used on the individual channels that have a
> > separate use.
> >
> > The images shown in the spec show the use of clean audio only in the
> > case of 5.1 channels with the center channel used for clean audio,
> > while when using audio description it shows the use of 3 channels:
> > left, right, and a mono audio description.The user has the possibility
> > to separately increase the volume of the audio description and of the
> > clean audio.
> >
> > If such a file with multiple audio channels was to be used in a HTML5
> > page, we wouldn't get the same level of control that DVB has defined,
> > because there is no control that allows to change the volume mix of
> > multi-channel audio. It is important to understand that there is a
> > difference between multi-channel audio and multi-track audio:
> > multi-channel audio is encoded within one audio track just by
> > interleaving the channels within that track. In HTML5 we don't have a
> > means to change the volume of the individual channels, but only to
> > change the volume of a track.
> >
> > Cheers,
> > Silvia.
> >
> >
> > On Fri, May 13, 2011 at 7:59 PM, Sean Hayes <Sean.Hayes@microsoft.com> wrote:
> >> Here is a summary of the European digital television definition for clean
> >> audio [adapted from a forthcoming edition of TS 101 154 from the DVB]. The
> >> system relies on client side mixing, under control of the programme-maker.
> >>
> >>
> >>
> >> Clean Audio refers to audio providing improved intelligibility. It is
> >> targeted for viewers with hearing impairments, but can as well serve as
> >> improvement for listening in noisy environments.… an important requirement
> >> is to adjust, on a passage by passage basis, the relative level of programme
> >> sound in the mix which the secondary audio user hears. The programme maker
> >> is best able to determine the level under controlled conditions when
> >> authoring the secondary audio information to modulate the level of programme
> >> sound in the secondary audio capable receiver so suitable secondary audio
> >> information is thus transmitted within the secondary audio stream.
> >>
> >>
> >>
> >> Individual secondary audio users will have different aural acuity, and there
> >> are, in practice, differences in audio signal level for different home
> >> receivers. An essential requirement is for the user to be able to adjust the
> >> volume to suit his/her condition.
> >>
> >>
> >>
> >> A descriptor is sent in a service to indicate a clean audio service. The
> >> level by which the main audio service should be attenuated is sent in the
> >> metadata (on a per channel basis). The client should accentuate the levels
> >> of the channel containing the dialogue and attenuate the other channels,
> >> accordingly. The level of the centre channel added should additionally be
> >> under user control to allow individual tailoring of the sound for
> >> audibility.
> >>
> >>
> >>
> >> Background:
> >>  http://www.rnid.org.uk/VirtualContent/84923/background_noise_PDF.pdf
> >>
> >> European HD TV requirements: http://tech.ebu.ch/docs/tech/tech3333.pdf [p15]
> >

-- 

Janina Sajka, Phone: +1.443.300.2200
  sip:janina@asterisk.rednote.net

Chair, Open Accessibility janina@a11y.org 
Linux Foundation  http://a11y.org

Chair, Protocols & Formats
Web Accessibility Initiative http://www.w3.org/wai/pf
World Wide Web Consortium (W3C)

Received on Sunday, 15 May 2011 01:25:55 UTC