[mythtv] Looking for Close Captioning info.

[Daniel Kristjansson]

On Thu, 2007-11-08 at 10:03 -0600, stuart wrote:

> Am looking for closed captioning information and how CC is embedded into 
> a mythtv video file.  I assume the TV tuner card captures the OTA NTSC 
> data off the VBI and presents it to ivtv which in turn presents it to 
> the mythtv back end.  Then what happens....?  How does the myth font end 
> decode it?

ivtv embeds the captions in an ivtv specific format into the mpeg file
it creates. In AvFormatDecoder::GetFrame() we look for the at codec_type
and codec_id and if we see CODEC_TYPE_DATA and CODEC_ID_MPEG2VBI we call
AvFormatDecoder::ProcessVBIDataPacket(). This further checks for either
"tv0" or "TV0" at the head of the data buffer. If either of these are
present we then decode the data using either the TeletextDecoder or
CC608Decoder depending on the buffer id of each line of data.

There are five types of data: TeleText, EIT-608, VPS, WSS and XDS.
We decode NTSC's TeleText, EIT-608, WSS, and XDS;
we decode PAL TeleText and WSS; and,
we decode SECAM TeleText.

VPS and XDS contain information relevant to recording programs such as
the program name now & next, duration, etc.

WSS tells you how to format the video, i.e. which areas are active vs
blank letterbox areas, and the true aspect ratio of the video.

You can see the VPS, XDS, and WSS data using -v vbi, but we currently
only use a limited amount of XDS data and no VPS or WSS data.

-- Daniel