[vlc-devel] [PATCH] cc.c : Added basic features of CIA-708 to cc.c

[Devin Heitmueller]

Hi Greeshma,


> I actually started working on this 6-7 days back. This is part of
> qualification task for Outreach program for women'15.

Ok, good to know the background.  Thanks.

>> - It looks like this patch just creates some basic data structures and
>> enums.  Do you have additional patches which demonstrate this
>> functionality?  I would be hesitant to see anything committed that
>> doesn't actually result in a functional change in the code.  I think
>> it would be better to see the development happen on some branch, and
>> then the entire patch series can be submitted for upstream inclusion
>> once you have something that works.  The alternative tends to result
>> in code that needs to be heavily refactored as you make it actually
>> work, dead code that isn't actually needed in the final solution, and
>> introducing new constraints on existing functionality.
>>
> As today is the deadline for application i have submitted all i could do
> till now. I am right now working on the proper functionalities.
>
>>
>> - I believe we're currently only feeding the EIA-608 bytes to the cc
>> decoder currently.  If we wanted to support EIA-708, the MPEG2 demux
>> would need to be modified to create additional decoder instances and
>> feed those bytes instead.
>
>
> Yes, MPEG2 demux has to be modified/extra features are to be added.As per my
> knowledge we now send captions included in picture user data.We need to
> change the decoder accordingly.

Captions can be present in either GOP or Picture User Data.  That
said, EIA-708 will typically want to use the Picture userdata due to
the increased bandwidth available.  I actually don't know if VLC
supports this today.

>> - We need to think about the GUI implications of supporting both
>> EIA-608 and 708.  Do we now show eight menu options for CC1-4 for
>> EIA-608 and CC1-4 for EIA-708?  Do we automatically pick EIA-708 over
>> EIA-608 (or the reverse) if both are present?  Do we have some global
>> configuration option which dictates which decoder gets used?
>
>
> EIA 708 has way lot more features than EIA 608.608 is kind of outdated
> format.There  is a way to convert 608 format to 708 but vice-versa is not
> possible.The transmission options decide which decoder to be used- eia608
> uses line 21 transmission whereas eia708 is embedded in MPEG-2 streams.

EIA-608 is also supported for embedding in MPEG2 streams.  In fact, in
the United States it is mandatory that both EIA-708 *and* EIA-608 be
embedded in terrestrial broadcasts.  As a result, most open source
implementations simply decode the EIA-608 stream rather than fully
supporting the complete feature set of EIA-708.

As a general trend, very few of the EIA-708 specific features are
being used.  Most broadcasters will generate an EIA-608 stream and
then convert it to EIA-708 so that both streams are present in the
resulting MPEG2 transport stream (thereby complying with the FCC
regulations).  As a result, the EIA-708 stream will generally only use
the features already found in EIA-608.  That's the key reason that
nobody has bothered implementing EIA-708 decoding.

On a related note, you should google for "WGBH EIA-708" if you want to
find some test streams that demonstrate the various EIA-708 features.
If you can't find them, email me and I'll point you to copies.

> Thank you for the guidance.I am trying to contribute to it through Outreach
> Program for Women -OPW'15.
> Even if its not possible i would definitely continue working on this and
> finish the decoder.

Good luck!

Devin

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