Some questions for Niels ...

On Tue, Feb 24, 2015 at 12:07:51PM +0100, Diego Biurrun wrote:
> --- a/libavcodec/dca.h
> +++ b/libavcodec/dca.h
> @@ -69,6 +84,61 @@ enum DCAExtensionMask {
>  
> +typedef struct XllChSetSubHeader {
> +    int fs_interpolate;         ///< sampling frequency interpolation 
> multiplier

What is "fs" supposed to stand for?

> +    int hier_chset;

hier?

> @@ -164,12 +236,31 @@ typedef struct DCAContext {
> +    int xll_banddata_crcen;     ///< presence of CRC16 within each frequency 
> band

crcen?

> --- a/libavcodec/dca_exss.c
> +++ b/libavcodec/dca_exss.c
> @@ -242,21 +243,27 @@ static int dca_exss_parse_asset_header(DCAContext *s)
> @@ -309,13 +316,52 @@ void ff_dca_exss_parse_header(DCAContext *s)
> +            /* parse extensions that we know about */
> +            switch (mkr) {
> +            case 0x41a29547: /* XLL */
> +                if (s->xll_disable) {
> +                    av_log(s->avctx, AV_LOG_DEBUG,
> +                           "DTS-XLL: ignoring XLL extension\n");
> +                    break;
> +                }
> +                av_log(s->avctx, AV_LOG_DEBUG,
> +                       "DTS-XLL: decoding XLL extension\n");
> +                if (ff_dca_xll_decode_header(s)        == 0 &&
> +                    ff_dca_xll_decode_navi(s, end_pos) == 0)
> +                    s->exss_ext_mask |= DCA_EXT_EXSS_XLL;
> +                break;
> +            case 0x655e315e: /* XBR */
> +            case 0x47004a03: /* XXCH */
> +            default:
> +                av_log(s->avctx, AV_LOG_VERBOSE,
> +                       "DTS-ExSS: unknown marker = 0x%08"PRIx32"\n", mkr);
> +            }

The case labels should be enums or defines IMO.

Why do the XBR and XXCH cases fall through to unknown marker?

> --- /dev/null
> +++ b/libavcodec/dca_xll.c
> @@ -0,0 +1,749 @@
> +
> +    /* align to 8-bit and skip 16-bit CRC */

You mean "Align to 8 bits and skip 16-bit CRC."?

Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to