OK, thanks. Actually I already figured most of this out while I was waiting for a reply !

Now I would like to know how to get the audio values (rate, channels, sample size, signedness and endianness).

Also, does the add_ntsc_setup field do anything (I see that it is ignored in init, but can be set separately) ? I seem to be able to encode NTSC without it.

If anyone is interested, the code is here:
http://lives.cvs.sourceforge.net/lives/lives-plugins/plugins/decoders/dv_decoder.c

it could be a useful example of a simple decoder.


Regards,
Salsaman.
http://lives.sourceforge.net


On Mon, Sep 22, 2008 at 1:48 PM, Dan Dennedy <dan@dennedy.org> wrote:
On Sun, Sep 21, 2008 at 1:24 PM, salsaman <salsaman@gmail.com> wrote:
> Hi,
> sorry if this is posted twice, my original message seems to have not reached
> the list.
>
> could somebody please explain how yuv_mode of dv_decode_full_frame() works ?

It is by default YUY2. It requires a build-time define for YV12.
See here, start at line 278 for a quick example:
http://mlt.svn.sourceforge.net/viewvc/mlt/trunk/mlt/src/modules/dv/producer_libdv.c?v
iew=markup


> Also, do I need the dv_parse_header() and/or dv_parse_packs() ?

dv_parse_header just for decoding video, use dv_parse_packs if you
need certain metadata. It is safe to just call both.

--
+-DRD-+