W3C home > Mailing lists > Public > public-texttracks@w3.org > November 2011

Re: Roll-up captions in WebVTT

From: Shane Feldman <shane.feldman@nad.org>
Date: Wed, 30 Nov 2011 02:45:38 -0500
Message-ID: <CAAxSGs0C3YO1qSm4cP7+hRdtGLjRDmndnkcU_ghggKO_SU9qTw@mail.gmail.com>
To: John Foliot <jfoliot@stanford.edu>
Cc: Ian Hickson <ian@hixie.ch>, Christian Vogler <christian.vogler@gallaudet.edu>, Silvia Pfeiffer <silviapfeiffer1@gmail.com>, public-texttracks@w3.org
On Mon, Nov 28, 2011 at 8:00 PM, John Foliot <jfoliot@stanford.edu> wrote:

> Ian Hickson wrote:
> >
> > On Mon, 28 Nov 2011, Christian Vogler wrote:
> > >
> > > Take a look here: http://www.youtube.com/watch?v=F_q-RRXw-vY
> > >
> > > In that video, roll-up is actually very readable and leads the eye
> > very
> > > well with respect to focusing attention. No captioning or steno
> > errors
> > > in this video, but I hope this gets the point across.
> >
> > IMHO that's horrible compared to normal captions. It is always moving,
> > which means you can't read it as fast as normal captions, plus it's
> > continually distracting from the image.
>
> With all due respect, your opinion here should not be what is driving
> progress. If *you* don't like roll-up captions, don't use them. Others do
> and want this ability. Christian (for example) found it "...very
> readable...", so just because you don't like it is immaterial to this
> discussion.
>

While it is "readable" consumers generally prefer pop-up captions. If Ian
refers to pop-on captions as "normal" captions, we agree with him.
Consumers would prefer that once live captions are completed and time
tracks recorded, then they can be reconfigured and rebroadcast with pop-on
captions. We brought this up during the VPAAC WG1 discussion on live
captioning.

During subsequent discussions, we learned from Alex Varley of Media Access
Australia that a news station in Australia has technology where live
captions are presented as block captions (which looks like pop-on
captions). However, I have not seen a live demonstration of this
technology. I wonder if there is a delay and if the delay is acceptable. We
can learn more about this from Alex and if we like it, we can incorporate
support for it in the format.


>
>
> >
> > Good captions should be so low-overheard for the viewer that the viewer
> > can entirely forget that he's reading captions in the first place. You
> > simply can't get that effect with rollup captions.
>
> Again, this appears to be your opinion, but can you point us to a
> definitive source for this assertion?  We are not going for "effect" here,
> but rather functionality, and in some instances the need (or desire) to
> have rollup captions exist.  You asked for a use-case, and one was given.
> That you don't like it is no cause for it to be discarded.
>

There was a study on eye gaze and captions. You can find a summary at:
http://www.dcmp.org/caai/nadh133.pdf

In this study, consumers watched captions 84% of the time and watched the
video 14% of the time. While the study doesn't say (unless I missed it), I
assume that the consumers watched pop-on captions 100% of the time. If so,
based on experience, I think that the % of time watching captions would
increase and the % of time watching the video would decrease with paint on
or roll-up captions. It would be great if study was done to verify my
hypothesis.

Ian is right, captions should be low-overhead for the consumer. We want to
enjoy the video and not simply capture glimpses of the on-screen video.
However, as Andrew points out in the next email, the FCC will likely issue
a Report and Order in January mandating that captions be of equal or
greater quality than what was shown on TV. The FCC IP Captioning NPRM
proposes rules encouraging the industry to improve on captions that were
shown on TV.

So while the format needs to support archaic captions from TV, we would
like to see captions on the web (which are not from TV) of higher quality
and using technology that consumers prefer.


>
> JF
>
>
>
>
Received on Wednesday, 30 November 2011 07:46:28 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 8 May 2014 13:18:50 UTC