Skip to content
Snippets Groups Projects
J-M Valin (rewritten by T.B. Terriberry)'s avatar
Jean-Marc Valin authored
LPC_inverse_pred_gain_QA() was failing to detect unstable filters
because of insufficient numerical precision. Precision was increased
to Q24 for LPC coefs and the reciprocal now uses all the precision
available (variable shift right from the start).
b23a6ca5
History
Name Last commit Last update
..
fixed
float
A2NLSF.c
API.h
CNG.c
HP_variable_cutoff.c
Inlines.h
LPC_analysis_filter.c
LPC_inv_pred_gain.c
LP_variable_cutoff.c
MacroCount.h
MacroDebug.h
NLSF2A.c
NLSF_VQ.c
NLSF_VQ_weights_laroia.c
NLSF_decode.c
NLSF_del_dec_quant.c
NLSF_encode.c
NLSF_stabilize.c
NLSF_unpack.c
NSQ.c
NSQ_del_dec.c
PLC.c
PLC.h
SigProc_FIX.h
VAD.c
VQ_WMat_EC.c
ana_filt_bank_1.c
apply_sine_window.c
array_maxabs.c
autocorr.c
biquad_alt.c
burg_modified.c
bwexpander.c
bwexpander_32.c
check_control_input.c
code_signs.c
control.h
control_SNR.c
control_audio_bandwidth.c
control_codec.c
create_init_destroy.c
debug.c
debug.h
dec_API.c
decode_core.c
decode_frame.c
decode_indices.c
decode_parameters.c
decode_pitch.c
decode_pulses.c
decoder_set_fs.c
define.h
enc_API.c
encode_indices.c
encode_pulses.c
errors.h
gain_quant.c
init_encoder.c
inner_prod_aligned.c
interpolate.c
k2a.c
k2a_Q16.c
lin2log.c
log2lin.c
macros.h
main.h
pitch_analysis_core.c
pitch_est_defines.h
pitch_est_tables.c
process_NLSFs.c
quant_LTP_gains.c
resampler.c
resampler_down2.c
resampler_down2_3.c
resampler_private.h
resampler_private_AR2.c
resampler_private_IIR_FIR.c
resampler_private_down_FIR.c
resampler_private_up2_HQ.c
resampler_rom.c
resampler_rom.h
resampler_structs.h
scale_copy_vector16.c
scale_vector.c
schur.c
schur64.c
shell_coder.c
sigm_Q15.c
silk_common.vcxproj
silk_common.vcxproj.filters
sort.c
stereo_LR_to_MS.c
stereo_MS_to_LR.c
stereo_decode_pred.c
stereo_encode_pred.c
stereo_find_predictor.c
stereo_quant_pred.c
structs.h
sum_sqr_shift.c