Mailing List Archive

Re: [mythtv-commits] Ticket #10273: CEA-708 captions are delayed because of incorrect byte-pair handling
On Sat, Jan 21, 2012 at 6:51 PM, <> wrote:
> #10273: CEA-708 captions are delayed because of incorrect byte-pair handling
> ----------------------------------+-------------------------
>     Reporter:  argilo@…          |      Owner:  danielk
>         Type:  Patch - Bug Fix   |     Status:  new
>     Priority:  minor             |  Milestone:  unknown
>    Component:  MythTV - ATSC     |    Version:  Master Head
>     Severity:  medium            |   Keywords:  cea-708
> Ticket locked:  0                 |
> ----------------------------------+-------------------------
>  ATSC closed captions (CEA-708) are sometimes delayed due do a bug in the
>  way the DTVCC transport layer is packetized.  According to the CEA-708
>  specification:
>  "The end of the packet is indicated by either: (1) receipt of the header
>  (cc_valid = 1, cc_type = 11) of the next DTVCC packet, or (2) receipt of a
>  byte-pair where cc_valid = 0 and cc_type = 10 or cc_type = 11."
>  But only case (1) is handled currently.  As a result, a received packet
>  will not be processed until the next one arrives, which in some cases only
>  occurs after a long time has passed.  This usually manifests in roll-up
>  captions that are missing the last letter or two until more text appears,
>  on in pop-on captions that appear too late, not leaving enough time for
>  the viewer to read them before they disappear.
>  The attached patch adds handling for case (2).  I have tested it with the
>  PBS test material ( and
>  recorded programs from various local ATSC stations.  The delays are gone,
>  and I have not observed any new problems.

Thanks for this fix! Now, on my favorite test recording, according to
mythccextractor, EIA-708 caption durations and gaps between adjacent
captions differ by at most 3 frames from their EIA-608 equivalents.
As a result, I would say EIA-708 captions are now truly usable in

mythtv-dev mailing list