1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-23 12:43:46 +02:00

use dsputil for float to signed 16-bit sample conversion

Originally committed as revision 9914 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Justin Ruggles 2007-08-04 20:59:22 +00:00
parent 97be28d8d7
commit 4e09232070

View File

@ -150,8 +150,11 @@ typedef struct {
MDCTContext imdct_512; //for 512 sample imdct transform
MDCTContext imdct_256; //for 256 sample imdct transform
DSPContext dsp; //for optimization
float add_bias; ///< offset for float_to_int16 conversion
float mul_bias; ///< scaling for float_to_int16 conversion
DECLARE_ALIGNED_16(float, output[AC3_MAX_CHANNELS-1][256]); //output after imdct transform and windowing
DECLARE_ALIGNED_16(short, int_output[AC3_MAX_CHANNELS-1][256]); ///< final 16-bit integer output
DECLARE_ALIGNED_16(float, delay[AC3_MAX_CHANNELS-1][256]); //delay - added to the next block
DECLARE_ALIGNED_16(float, tmp_imdct[256]); //temporary storage for imdct transform
DECLARE_ALIGNED_16(float, tmp_output[512]); //temporary storage for output before windowing
@ -262,6 +265,14 @@ static int ac3_decode_init(AVCodecContext *avctx)
dsputil_init(&ctx->dsp, avctx);
av_init_random(0, &ctx->dith_state);
if(ctx->dsp.float_to_int16 == ff_float_to_int16_c) {
ctx->add_bias = 385.0f;
ctx->mul_bias = 1.0f;
} else {
ctx->add_bias = 0.0f;
ctx->mul_bias = 32767.0f;
}
return 0;
}
/*********** END INIT FUNCTIONS ***********/
@ -651,7 +662,7 @@ static inline void do_imdct(AC3DecodeContext *ctx)
ctx->tmp_imdct);
}
ctx->dsp.vector_fmul_add_add(ctx->output[ch-1], ctx->tmp_output,
ctx->window, ctx->delay[ch-1], 384, 256, 1);
ctx->window, ctx->delay[ch-1], ctx->add_bias, 256, 1);
ctx->dsp.vector_fmul_reverse(ctx->delay[ch-1], ctx->tmp_output+256,
ctx->window, 256);
}
@ -921,7 +932,7 @@ static int ac3_parse_audio_block(AC3DecodeContext *ctx, int blk)
/* apply scaling to coefficients (headroom, dynrng) */
for(ch=1; ch<=ctx->nchans; ch++) {
float gain = 2.0f;
float gain = 2.0f * ctx->mul_bias;
if(ctx->acmod == AC3_ACMOD_DUALMONO && ch == 2) {
gain *= ctx->dynrng2;
} else {
@ -934,17 +945,12 @@ static int ac3_parse_audio_block(AC3DecodeContext *ctx, int blk)
do_imdct(ctx);
return 0;
}
/* convert float to 16-bit integer */
for(ch=0; ch<ctx->out_channels; ch++) {
ctx->dsp.float_to_int16(ctx->int_output[ch], ctx->output[ch], 256);
}
static inline int16_t convert(int32_t i)
{
if (i > 0x43c07fff)
return 32767;
else if (i <= 0x43bf8000)
return -32768;
else
return (i - 0x43c00000);
return 0;
}
/* Decode ac3 frame.
@ -960,10 +966,6 @@ static int ac3_decode_frame(AVCodecContext * avctx, void *data, int *data_size,
AC3DecodeContext *ctx = (AC3DecodeContext *)avctx->priv_data;
int16_t *out_samples = (int16_t *)data;
int i, blk, ch;
int32_t *int_ptr[6];
for (ch = 0; ch < 6; ch++)
int_ptr[ch] = (int32_t *)(&ctx->output[ch]);
//Initialize the GetBitContext with the start of valid AC3 Frame.
init_get_bits(&ctx->gb, buf, buf_size * 8);
@ -999,7 +1001,7 @@ static int ac3_decode_frame(AVCodecContext * avctx, void *data, int *data_size,
}
for (i = 0; i < 256; i++)
for (ch = 0; ch < ctx->out_channels; ch++)
*(out_samples++) = convert(int_ptr[ch][i]);
*(out_samples++) = ctx->int_output[ch][i];
}
*data_size = NB_BLOCKS * 256 * avctx->channels * sizeof (int16_t);
return ctx->frame_size;