Lynne:
> Jan 9, 2021, 22:01 by andreas.rheinha...@gmail.com:
>
>> Lynne:
>>
>>> @@ -165,7 +164,11 @@ typedef struct AC3EncodeContext {
>>> AVCodecContext *avctx; ///< parent AVCodecContext
>>> PutBitContext pb; ///< bitstream writer context
>>> AudioDSPContext adsp;
>>> +#if AC3ENC_FLOAT
>>> AVFloatDSPContext *fdsp;
>>> +#else
>>> + AVFixedDSPContext *fdsp;
>>> +#endif
>>> MECmpContext mecc;
>>> AC3DSPContext ac3dsp; ///< AC-3 optimized functions
>>> FFTContext mdct; ///< FFT context for MDCT
>>> calculation
>>>
>> [...]
>>
>>> @@ -118,9 +89,10 @@ static CoefType calc_cpl_coord(CoefSumType energy_ch,
>>> CoefSumType energy_cpl)
>>> static av_cold void ac3_fixed_mdct_end(AC3EncodeContext *s)
>>> {
>>> ff_mdct_end(&s->mdct);
>>> + av_freep(&s->fdsp);
>>> + av_freep(&s->mdct_window);
>>> }
>>>
>>
>> ff_ac3_encode_close already unconditionally frees fdsp, so freeing it
>> above is either unnecessary or ac3_float_mdct_end should also free its
>> fdsp (and ff_ac3_encode_close shouldn't). Freeing mdct_window can also
>> be moved to ff_ac3_encode_close (which already frees several buffers
>> whose pointed-to-type depends upon the encoding mode).
>> Notice that ac3enc.c uses the fixed-point mode, but the layout of
>> AC3EncodeContext does not depend upon this (apart from pointed-to-types,
>> of course). Actually, ff_mdct_end does the same for both fixed- and
>> floating-point mode, so one could even incorporate
>> ac3_fixed/float_mdct_end into ff_ac3_encode_close.
>>
> Done. Left ac3_fixed/float_mdct_end as-is for now.
> New patch attached.
> >
> @@ -129,8 +99,31 @@ static av_cold void ac3_fixed_mdct_end(AC3EncodeContext
> *s)
> */
> static av_cold int ac3_fixed_mdct_init(AC3EncodeContext *s)
> {
> + int32_t *iwin;
> + float fwin[AC3_BLOCK_SIZE];
> +
> int ret = ff_mdct_init(&s->mdct, 9, 0, -1.0);
> - s->mdct_window = ff_ac3_window;
You forgot to remove this table.
> + if (ret < 0)
> + return ret;
> +
> + iwin = av_malloc_array(AC3_WINDOW_SIZE, sizeof(*iwin));
> + if (!iwin)
> + return AVERROR(ENOMEM);
> +
> + ff_kbd_window_init(fwin, 5.0, AC3_WINDOW_SIZE/2);
> +
> + for (int i = 0; i < AC3_WINDOW_SIZE/2; i++)
> + iwin[i] = lrintf(fwin[i] * (1 << 22));
> +
Does this lead to a different result than using ff_kbd_window_init_fixed
directly?
> + for (int i = 0; i < AC3_WINDOW_SIZE/2; i++)
> + iwin[AC3_WINDOW_SIZE-1-i] = iwin[i];
> +
> + s->mdct_window = iwin;
> +
> + s->fdsp = avpriv_alloc_fixed_dsp(s->avctx->flags &
> AV_CODEC_FLAG_BITEXACT);
> + if (!s->fdsp)
> + return AVERROR(ENOMEM);
> +
> return ret;
> }
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
