Re: [libdv-dev] dv_decode_full_frame

[salsaman <sal...@gm...>]

I should have mentioned cdata->width=720


Salsaman.
http://lives.sourceforge.net


On Sat, Sep 20, 2008 at 1:52 PM, salsaman <sal...@gm...> wrote:

> Hi,
> could somebody please explain how dv_decode_full_frame() works ?
>
> I have the following code for example (for PAL), but it is crashing:
>
>     uint8_t fbuffer[144000];
>
>     int frame=1;
>     size_t bytes=144000;
>     int rowstrides[3];
>
>     if (cdata.palette==YUV420) {
>       rowstrides[0]=cdata.width;
>       rowstrides[1]=cdata.width>>1;
>       rowstrides[2]=cdata.width>>1;
>     }
>     else {
>       rowstrides[0]=cdata.width*6; // YUV411 - 6 bytes per pixel
>     }
>
>     lseek(fd,bytes,SEEK_SET);
>
>     if (read (fd, fbuffer, 144000) < 144000) return FALSE;
>
>     dv_parse_header(dv_dec, fbuffer);
>     //dv_parse_packs(dv_dec, fbuffer);
>
>     dv_decode_full_frame(dv_dec,fbuffer,e_dv_color_yuv,(uint8_t
> **)pixel_data,rowstrides);
>
>
>
>
> Here pixel_data is an array of pointers to the Y,U, and V planes for
> YUV420.
>
> Also, do I need the dv_parse_header() and/or dv_parse_packs() ?
>
>
> Regards,
> Salsaman.
> http://lives.sourceforge.net
>
>
>
>