mirrors
/
flac
mirror of https://github.com/xiph/flac
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Further improve calculation of when to use wide residual computation (#702)

This commit is contained in:
Martijn van Beurden 2024-05-16 16:00:19 +02:00 committed by GitHub
parent 1ab3c8e748
commit cfe3afca9b
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 16 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/libFLAC/include/private/lpc.h
View File

@ -181,6 +181,7 @@ void FLAC__lpc_compute_residual_from_qlp_coefficients_wide_intrin_avx2(const FLA
#endif /* !defined FLAC__INTEGER_ONLY_LIBRARY */
FLAC__uint64 FLAC__lpc_max_prediction_value_before_shift(uint32_t subframe_bps, const FLAC__int32 qlp_coeff[], uint32_t order);
uint32_t FLAC__lpc_max_prediction_before_shift_bps(uint32_t subframe_bps, const FLAC__int32 qlp_coeff[], uint32_t order);
uint32_t FLAC__lpc_max_residual_bps(uint32_t subframe_bps, const FLAC__int32 qlp_coeff[], uint32_t order, int lp_quantization);

25
src/libFLAC/lpc.c
View File

@ -939,27 +939,32 @@ FLAC__bool FLAC__lpc_compute_residual_from_qlp_coefficients_limit_residual_33bit
#endif /* !defined FLAC__INTEGER_ONLY_LIBRARY */
FLAC__uint64 FLAC__lpc_max_prediction_value_before_shift(uint32_t subframe_bps, const FLAC__int32 * flac_restrict qlp_coeff, uint32_t order)
{
FLAC__uint64 max_abs_sample_value = (FLAC__uint64)(1) << (subframe_bps - 1);
FLAC__uint32 abs_sum_of_qlp_coeff = 0;
uint32_t i;
for(i = 0; i < order; i++)
abs_sum_of_qlp_coeff += abs(qlp_coeff[i]);
return max_abs_sample_value * abs_sum_of_qlp_coeff;
}
uint32_t FLAC__lpc_max_prediction_before_shift_bps(uint32_t subframe_bps, const FLAC__int32 * flac_restrict qlp_coeff, uint32_t order)
{
/* This used to be subframe_bps + qlp_coeff_precision + FLAC__bitmath_ilog2(order)
* but that treats both the samples as well as the predictor as unknown. The
* predictor is known however, so taking the log2 of the sum of the absolute values
* of all coefficients is a more accurate representation of the predictor */
FLAC__uint32 abs_sum_of_qlp_coeff = 0;
uint32_t i;
for(i = 0; i < order; i++)
abs_sum_of_qlp_coeff += abs(qlp_coeff[i]);
return subframe_bps + FLAC__bitmath_extra_mulbits_unsigned(abs_sum_of_qlp_coeff);
return FLAC__bitmath_silog2(FLAC__lpc_max_prediction_value_before_shift(subframe_bps, qlp_coeff, order));
}
uint32_t FLAC__lpc_max_residual_bps(uint32_t subframe_bps, const FLAC__int32 * flac_restrict qlp_coeff, uint32_t order, int lp_quantization)
{
FLAC__int32 predictor_sum_bps = FLAC__lpc_max_prediction_before_shift_bps(subframe_bps, qlp_coeff, order) - lp_quantization;
if((int)subframe_bps > predictor_sum_bps)
return subframe_bps + 1;
else
return predictor_sum_bps + 1;
FLAC__uint64 max_abs_sample_value = (FLAC__uint64)(1) << (subframe_bps - 1);
FLAC__uint64 max_prediction_value_after_shift = -1 * ((-1 * (FLAC__int64)FLAC__lpc_max_prediction_value_before_shift(subframe_bps, qlp_coeff, order)) >> lp_quantization);
FLAC__uint64 max_residual_value = max_abs_sample_value + max_prediction_value_after_shift;
return FLAC__bitmath_silog2(max_residual_value);
}
#if defined(FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION) && !defined(FUZZING_BUILD_MODE_FLAC_SANITIZE_SIGNED_INTEGER_OVERFLOW)