Re: Roll-up captions in WebVTT

Just to sum up my opinion right now, since I think the conversation may
have muddied it: I'm not convinced that a significant enough number of
viewers actually prefer roll-up captions to pop-on captions to support it
for that reason, given that almost all movie subtitles are pop-on.  I think
roll-on captions *may* be required for live captioning, but Gal claims
otherwise and I ask for more information below.  I'm not strongly for or
against the rendering mode as a whole, but I'm strongly against certain
methods of implementing them (especially the "repeat the cue over and over
manually" one).


On Mon, Dec 19, 2011 at 3:10 AM, Gal Klein <gal@plymedia.com> wrote:

> As I was posting in previous emails, we have been doing LIVE caption for
> online video for a while now.
> We never use roll-up captions as they make it very difficult to follow
> with the video (there have been many articles stating the roll-up caption
> obscure the video viewage).
> Pop-up caption can be transmitted in Real Time with just some basic logics
> that can be inserted to an online caption tool.
> Will be happy to explore this with the team and also show you examples
> (like the WSJ front page 4 times a day), improved synchronization is also
> possible with some additional effort, and we are working on implementing
> this as well.
>

Can you go into detail about this?  We should learn exactly how this works
in practice (eg. does it increase caption delays?), what it looks like to
users, and video demos would be ideal.  You might want to start another
thread about this (this one is getting cluttered).


On Mon, Dec 19, 2011 at 6:18 AM, Silvia Pfeiffer <silviapfeiffer1@gmail.com>
wrote:
> Not necessarily. Ian invented a <redoline> (redo-line) tag in this bug
> https://www.w3.org/Bugs/Public/show_bug.cgi?id=14104 . That could be
> used for rollup as well as popup captions and it is editing done
> through markup.

Which he labelled as a variant, and he said that WebVTT was designed for
static resources.  WebVTT isn't a streaming format, and trying to bend it
into one will likely make a mess of the whole thing.  I'm not saying live
editing shouldn't be supported in some form, but it sounds like something
for a layer on top of WebVTT.  (I wouldn't say it's *impossible* to wedge
this into WebVTT cleanly, but it seems at least improbable and that it
would complicate WebVTT for everyone.)

Let's not tangent too far into the question of how to transmit live
captions, so we don't get too far off track.  It's orthogonal to the
rendering mode question.

> How would the markup look such that you could render it to either
> rollup or popup? Do you have an example markup that could be rendered
> either way with using just UA settings?

Any WebVTT text.  Again, there'd be work to do here and questions to
answer, but it's solvable.  For example, render all text as roll-up, with
cues with a cue line position of auto or of >= 50 being grouped at the
bottom (rolling up) and other cues being grouped at the top rolling down.
 (I'm not going to try to lay out a complete rendering algorithm for
roll-up; I'm not at all familiar enough with that part of the spec to even
try.  This is the part that would require a fair bit of work.)

> My suggestion to solve this problem was to have a "class" on cues that
> would group cues together such that they can be rendered as rollup
> (see my first email in this thread). It is minimal additional markup
> and would indeed allow to create rollup or popup captions from the
> same content. It could be changed through preferences.

Not if you really want to be able to see roll-ups in general, since you'd
only be able to see roll-ups where the author has specified it.  Almost no
authors will do that--you can disagree with that premise if you want, but I
think I'm right--so you'd end up seeing pop-ons most of the time.

>> Bitmap subtitles on DVDs (the
>> form used by most movies) didn't even support roll-up (from what I
>> recall when I implemented a decoder many years back).
>
> DVD is never used for live and captions for live recordings on DVD
> would always have been reformatted. So, I understand why that never
> existed.

We're talking about the use of roll-up in general, including prerecorded
captions.  You said that roll-up captions are more natural to US readers,
and that pop-on captions are more common with anime than other genres.
 That simply doesn't seem to be true; roll-up captions seem exceptionally
rare outside of live captions and not even supported by many media.  I
showed samples from several media formats and different countries to
support this.


On Mon, Dec 19, 2011 at 11:30 AM, David Singer <singer@apple.com> wrote:
> That would, indeed, but I am not sure it's been suggested?

I was responding to "I think you have to provide two different files with
different makrup for people to choose from if you want to support both
means."

> It's only evil if you can't tell it's a duplicate when you need to know
(e.g. when using TTS), and what I am suggesting is tagging to say that, for
those that need to know.

It's evil and ugly regardless.  For credits, you would need hundreds or
even thousands of copies of each cue to scroll all the way up the screen.
I'm actually a bit taken aback that it's being put forward seriously.

-- 
Glenn Maynard