[mythtv-users] ATSC caption question

[Devin Heitmueller]

On Tue, Dec 22, 2009 at 1:10 PM, Yeechang Lee <ylee at pobox.com> wrote:
> I've been testing Taylor Ralph's patches to EIA-708 support (#7739,
> #7775, and #7776). As he says, they do indeed finally make ATSC
> captions usable. A few questions/observations:
>
> * In what situations would using ATSC captions over VBI (which I'd
>  used until these patches) be preferable, if both are available? I
>  know ATSC supports color foregrounds and backgrounds, and multiple
>  typefaces, but are such feature ever actually seen in the real
>  world?
>
>  On the other hand, during the past couple of days using Taylor's
>  work I've found a few situations in which VBI captions are
>  preferable to ATSC:
>
>  * A PBS program with almost-completely corrupt ATSC but intact VBI
>    captions.
>  * A movie in which, during dialogue, VBI captions' positions vary to
>    indicate who is saying what but ATSC captions don't. VBI captions
>    sync with dialogue more accurately, while ATSC captions tend to
>    appear/disappear much more quickly at the end of the lines they
>    depict.
>
> * I enable black backgrounds for captions for visibility. VBI
>  captions' backgrounds cover the letters and no more. ATSC captions
>  apparently reserve a fixed amount of screen space and so their
>  backgrounds cover more than just the words on the screen at a given
>  moment.
>
> * VBI captions depict musical notes properly while ATSC captions show
>  square boxes (an issue fixed with VBI captions some time ago). I use
>  the same fonts for both so it can't be a font issue.
>
> --
> Yeechang Lee <ylee at pobox.com> | San Francisco CA US

Hi Yeechang,

I don't doubt the points you are raising are visible - but I do wonder
whether they are really because of differences in the actual
broadcast, or whether they are bugs in MythTV's EIA-708
implementation.  I cannot speak for the developer who did the support,
but It might be worthwhile to make some sample captures available
which are suspect, and analyze whether it's really an inconsistency in
the actual content in the broadcast.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com