diff options
Diffstat (limited to 'libspeexdsp/preprocess.c')
| -rw-r--r-- | libspeexdsp/preprocess.c | 144 |
1 files changed, 72 insertions, 72 deletions
diff --git a/libspeexdsp/preprocess.c b/libspeexdsp/preprocess.c index b8e287a..3053eb5 100644 --- a/libspeexdsp/preprocess.c +++ b/libspeexdsp/preprocess.c @@ -1,6 +1,6 @@ /* Copyright (C) 2003 Epic Games (written by Jean-Marc Valin) - Copyright (C) 2004-2006 Epic Games - + Copyright (C) 2004-2006 Epic Games + File: preprocess.c Preprocessor with denoising based on the algorithm by Ephraim and Malah @@ -34,24 +34,24 @@ /* Recommended papers: - + Y. Ephraim and D. Malah, "Speech enhancement using minimum mean-square error - short-time spectral amplitude estimator". IEEE Transactions on Acoustics, + short-time spectral amplitude estimator". IEEE Transactions on Acoustics, Speech and Signal Processing, vol. ASSP-32, no. 6, pp. 1109-1121, 1984. - + Y. Ephraim and D. Malah, "Speech enhancement using minimum mean-square error - log-spectral amplitude estimator". IEEE Transactions on Acoustics, Speech and + log-spectral amplitude estimator". IEEE Transactions on Acoustics, Speech and Signal Processing, vol. ASSP-33, no. 2, pp. 443-445, 1985. - + I. Cohen and B. Berdugo, "Speech enhancement for non-stationary noise environments". Signal Processing, vol. 81, no. 2, pp. 2403-2418, 2001. - Stefan Gustafsson, Rainer Martin, Peter Jax, and Peter Vary. "A psychoacoustic - approach to combined acoustic echo cancellation and noise reduction". IEEE + Stefan Gustafsson, Rainer Martin, Peter Jax, and Peter Vary. "A psychoacoustic + approach to combined acoustic echo cancellation and noise reduction". IEEE Transactions on Speech and Audio Processing, 2002. - + J.-M. Valin, J. Rouat, and F. Michaud, "Microphone array post-filter for separation - of simultaneous non-stationary sources". In Proceedings IEEE International + of simultaneous non-stationary sources". In Proceedings IEEE International Conference on Acoustics, Speech, and Signal Processing, 2004. */ @@ -71,7 +71,7 @@ #define LOUDNESS_EXP 5.f #define AMP_SCALE .001f #define AMP_SCALE_1 1000.f - + #define NB_BANDS 24 #define SPEECH_PROB_START_DEFAULT QCONST16(0.35f,15) @@ -113,7 +113,7 @@ static inline spx_word16_t DIV32_16_Q8(spx_word32_t a, spx_word32_t b) a = SHL32(a,8); return PDIV32_16(a,b); } - + } static inline spx_word16_t DIV32_16_Q15(spx_word32_t a, spx_word32_t b) { @@ -181,7 +181,7 @@ struct SpeexPreprocessState_ { int sampling_rate; /**< Sampling rate of the input/output */ int nbands; FilterBank *bank; - + /* Parameters */ int denoise_enabled; int vad_enabled; @@ -194,7 +194,7 @@ struct SpeexPreprocessState_ { int echo_suppress; int echo_suppress_active; SpeexEchoState *echo_state; - + spx_word16_t speech_prob; /**< Probability last frame was speech */ /* DSP-related arrays */ @@ -256,7 +256,7 @@ static void conj_window(spx_word16_t *w, int len) spx_word16_t tmp; #ifdef FIXED_POINT spx_word16_t x = DIV32_16(MULT16_16(32767,i),len); -#else +#else spx_word16_t x = DIV32_16(MULT16_16(QCONST16(4.f,13),i),len); #endif int inv=0; @@ -281,10 +281,10 @@ static void conj_window(spx_word16_t *w, int len) } } - + #ifdef FIXED_POINT -/* This function approximates the gain function - y = gamma(1.25)^2 * M(-.25;1;-x) / sqrt(x) +/* This function approximates the gain function + y = gamma(1.25)^2 * M(-.25;1;-x) / sqrt(x) which multiplied by xi/(1+xi) is the optimal gain in the loudness domain ( sqrt[amplitude] ) Input in Q11 format, output in Q15 @@ -317,7 +317,7 @@ static inline spx_word16_t qcurve(spx_word16_t x) static void compute_gain_floor(int noise_suppress, int effective_echo_suppress, spx_word32_t *noise, spx_word32_t *echo, spx_word16_t *gain_floor, int len) { int i; - + if (noise_suppress > effective_echo_suppress) { spx_word16_t noise_gain, gain_ratio; @@ -343,8 +343,8 @@ static void compute_gain_floor(int noise_suppress, int effective_echo_suppress, } #else -/* This function approximates the gain function - y = gamma(1.25)^2 * M(-.25;1;-x) / sqrt(x) +/* This function approximates the gain function + y = gamma(1.25)^2 * M(-.25;1;-x) / sqrt(x) which multiplied by xi/(1+xi) is the optimal gain in the loudness domain ( sqrt[amplitude] ) */ @@ -410,8 +410,8 @@ EXPORT SpeexPreprocessState *speex_preprocess_state_init(int frame_size, int sam break; } } - - + + if (st->ps_size < 3*st->frame_size/4) st->ps_size = st->ps_size * 3 / 2; #else @@ -421,7 +421,7 @@ EXPORT SpeexPreprocessState *speex_preprocess_state_init(int frame_size, int sam N = st->ps_size; N3 = 2*N - st->frame_size; N4 = st->frame_size - N3; - + st->sampling_rate = sampling_rate; st->denoise_enabled = 1; st->vad_enabled = 0; @@ -436,15 +436,15 @@ EXPORT SpeexPreprocessState *speex_preprocess_state_init(int frame_size, int sam st->speech_prob_continue = SPEECH_PROB_CONTINUE_DEFAULT; st->echo_state = NULL; - + st->nbands = NB_BANDS; M = st->nbands; st->bank = filterbank_new(M, sampling_rate, N, 1); - + st->frame = (spx_word16_t*)speex_alloc(2*N*sizeof(spx_word16_t)); st->window = (spx_word16_t*)speex_alloc(2*N*sizeof(spx_word16_t)); st->ft = (spx_word16_t*)speex_alloc(2*N*sizeof(spx_word16_t)); - + st->ps = (spx_word32_t*)speex_alloc((N+M)*sizeof(spx_word32_t)); st->noise = (spx_word32_t*)speex_alloc((N+M)*sizeof(spx_word32_t)); st->echo_noise = (spx_word32_t*)speex_alloc((N+M)*sizeof(spx_word32_t)); @@ -457,19 +457,19 @@ EXPORT SpeexPreprocessState *speex_preprocess_state_init(int frame_size, int sam st->gain2 = (spx_word16_t*)speex_alloc((N+M)*sizeof(spx_word16_t)); st->gain_floor = (spx_word16_t*)speex_alloc((N+M)*sizeof(spx_word16_t)); st->zeta = (spx_word16_t*)speex_alloc((N+M)*sizeof(spx_word16_t)); - + st->S = (spx_word32_t*)speex_alloc(N*sizeof(spx_word32_t)); st->Smin = (spx_word32_t*)speex_alloc(N*sizeof(spx_word32_t)); st->Stmp = (spx_word32_t*)speex_alloc(N*sizeof(spx_word32_t)); st->update_prob = (int*)speex_alloc(N*sizeof(int)); - + st->inbuf = (spx_word16_t*)speex_alloc(N3*sizeof(spx_word16_t)); st->outbuf = (spx_word16_t*)speex_alloc(N3*sizeof(spx_word16_t)); conj_window(st->window, 2*N3); for (i=2*N3;i<2*st->ps_size;i++) st->window[i]=Q15_ONE; - + if (N4>0) { for (i=N3-1;i>=0;i--) @@ -569,7 +569,7 @@ static void speex_compute_agc(SpeexPreprocessState *st, spx_word16_t Pframe, spx float target_gain; float loudness=1.f; float rate; - + for (i=2;i<N;i++) { loudness += 2.f*N*st->ps[i]* st->loudness_weight[i]; @@ -587,7 +587,7 @@ static void speex_compute_agc(SpeexPreprocessState *st, spx_word16_t Pframe, spx st->init_max *= 1.f + .1f*Pframe*Pframe; } /*printf ("%f %f %f %f\n", Pframe, loudness, pow(st->loudness, 1.0f/LOUDNESS_EXP), st->loudness2);*/ - + target_gain = AMP_SCALE*st->agc_level*pow(st->loudness/(1e-4+st->loudness_accum), -1.0f/LOUDNESS_EXP); if ((Pframe>.5 && st->nb_adapt > 20) || target_gain < st->agc_gain) @@ -600,11 +600,11 @@ static void speex_compute_agc(SpeexPreprocessState *st, spx_word16_t Pframe, spx target_gain = st->max_gain; if (target_gain > st->init_max) target_gain = st->init_max; - + st->agc_gain = target_gain; } /*fprintf (stderr, "%f %f %f\n", loudness, (float)AMP_SCALE_1*pow(st->loudness, 1.0f/LOUDNESS_EXP), st->agc_gain);*/ - + for (i=0;i<2*N;i++) ft[i] *= st->agc_gain; st->prev_loudness = loudness; @@ -624,7 +624,7 @@ static void preprocess_analysis(SpeexPreprocessState *st, spx_int16_t *x) st->frame[i]=st->inbuf[i]; for (i=0;i<st->frame_size;i++) st->frame[N3+i]=x[i]; - + /* Update inbuf */ for (i=0;i<N3;i++) st->inbuf[i]=x[N4+i]; @@ -643,10 +643,10 @@ static void preprocess_analysis(SpeexPreprocessState *st, spx_int16_t *x) st->frame[i] = SHL16(st->frame[i], st->frame_shift); } #endif - + /* Perform FFT */ spx_fft(st->fft_lookup, st->frame, st->ft); - + /* Power spectrum */ ps[0]=MULT16_16(st->ft[0],st->ft[0]); for (i=1;i<N;i++) @@ -664,11 +664,11 @@ static void update_noise_prob(SpeexPreprocessState *st) int N = st->ps_size; for (i=1;i<N-1;i++) - st->S[i] = MULT16_32_Q15(QCONST16(.8f,15),st->S[i]) + MULT16_32_Q15(QCONST16(.05f,15),st->ps[i-1]) + st->S[i] = MULT16_32_Q15(QCONST16(.8f,15),st->S[i]) + MULT16_32_Q15(QCONST16(.05f,15),st->ps[i-1]) + MULT16_32_Q15(QCONST16(.1f,15),st->ps[i]) + MULT16_32_Q15(QCONST16(.05f,15),st->ps[i+1]); st->S[0] = MULT16_32_Q15(QCONST16(.8f,15),st->S[0]) + MULT16_32_Q15(QCONST16(.2f,15),st->ps[0]); st->S[N-1] = MULT16_32_Q15(QCONST16(.8f,15),st->S[N-1]) + MULT16_32_Q15(QCONST16(.2f,15),st->ps[N-1]); - + if (st->nb_adapt==1) { for (i=0;i<N;i++) @@ -695,7 +695,7 @@ static void update_noise_prob(SpeexPreprocessState *st) for (i=0;i<N;i++) { st->Smin[i] = MIN32(st->Smin[i], st->S[i]); - st->Stmp[i] = MIN32(st->Stmp[i], st->S[i]); + st->Stmp[i] = MIN32(st->Stmp[i], st->S[i]); } } for (i=0;i<N;i++) @@ -731,12 +731,12 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) spx_word16_t Pframe; spx_word16_t beta, beta_1; spx_word16_t effective_echo_suppress; - + st->nb_adapt++; if (st->nb_adapt>20000) st->nb_adapt = 20000; st->min_count++; - + beta = MAX16(QCONST16(.03,15),DIV32_16(Q15_ONE,st->nb_adapt)); beta_1 = Q15_ONE-beta; M = st->nbands; @@ -770,7 +770,7 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) st->update_prob[i] = 0; } */ - + /* Update the noise estimate for the frequencies where it can be */ for (i=0;i<N;i++) { @@ -788,17 +788,17 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) for (i=0;i<N+M;i++) { spx_word16_t gamma; - + /* Total noise estimate including residual echo and reverberation */ spx_word32_t tot_noise = ADD32(ADD32(ADD32(EXTEND32(1), PSHR32(st->noise[i],NOISE_SHIFT)) , st->echo_noise[i]) , st->reverb_estimate[i]); - + /* A posteriori SNR = ps/noise - 1*/ st->post[i] = SUB16(DIV32_16_Q8(ps[i],tot_noise), QCONST16(1.f,SNR_SHIFT)); st->post[i]=MIN16(st->post[i], QCONST16(100.f,SNR_SHIFT)); - + /* Computing update gamma = .1 + .9*(old/(old+noise))^2 */ gamma = QCONST16(.1f,15)+MULT16_16_Q15(QCONST16(.89f,15),SQR16_Q15(DIV32_16_Q15(st->old_ps[i],ADD32(st->old_ps[i],tot_noise)))); - + /* A priori SNR update = gamma*max(0,post) + (1-gamma)*old/noise */ st->prior[i] = EXTRACT16(PSHR32(ADD32(MULT16_16(gamma,MAX16(0,st->post[i])), MULT16_16(Q15_ONE-gamma,DIV32_16_Q8(st->old_ps[i],tot_noise))), 15)); st->prior[i]=MIN16(st->prior[i], QCONST16(100.f,SNR_SHIFT)); @@ -819,13 +819,13 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) for (i=N;i<N+M;i++) Zframe = ADD32(Zframe, EXTEND32(st->zeta[i])); Pframe = QCONST16(.1f,15)+MULT16_16_Q15(QCONST16(.899f,15),qcurve(DIV32_16(Zframe,st->nbands))); - + effective_echo_suppress = EXTRACT16(PSHR32(ADD32(MULT16_16(SUB16(Q15_ONE,Pframe), st->echo_suppress), MULT16_16(Pframe, st->echo_suppress_active)),15)); - + compute_gain_floor(st->noise_suppress, effective_echo_suppress, st->noise+N, st->echo_noise+N, st->gain_floor+N, M); - - /* Compute Ephraim & Malah gain speech probability of presence for each critical band (Bark scale) - Technically this is actually wrong because the EM gaim assumes a slightly different probability + + /* Compute Ephraim & Malah gain speech probability of presence for each critical band (Bark scale) + Technically this is actually wrong because the EM gaim assumes a slightly different probability distribution */ for (i=N;i<N+M;i++) { @@ -842,7 +842,7 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) #ifdef FIXED_POINT spx_word16_t tmp; #endif - + prior_ratio = PDIV32_16(SHL32(EXTEND32(st->prior[i]), 15), ADD16(st->prior[i], SHL32(1,SNR_SHIFT))); theta = MULT16_32_P15(prior_ratio, QCONST32(1.f,EXPIN_SHIFT)+SHL32(EXTEND32(st->post[i]),EXPIN_SHIFT-SNR_SHIFT)); @@ -867,12 +867,12 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) /* Convert the EM gains and speech prob to linear frequency */ filterbank_compute_psd16(st->bank,st->gain2+N, st->gain2); filterbank_compute_psd16(st->bank,st->gain+N, st->gain); - + /* Use 1 for linear gain resolution (best) or 0 for Bark gain resolution (faster) */ if (1) { filterbank_compute_psd16(st->bank,st->gain_floor+N, st->gain_floor); - + /* Compute gain according to the Ephraim-Malah algorithm -- linear frequency */ for (i=0;i<N;i++) { @@ -882,7 +882,7 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) spx_word16_t tmp; spx_word16_t p; spx_word16_t g; - + /* Wiener filter gain */ prior_ratio = PDIV32_16(SHL32(EXTEND32(st->prior[i]), 15), ADD16(st->prior[i], SHL32(1,SNR_SHIFT))); theta = MULT16_32_P15(prior_ratio, QCONST32(1.f,EXPIN_SHIFT)+SHL32(EXTEND32(st->post[i]),EXPIN_SHIFT-SNR_SHIFT)); @@ -893,22 +893,22 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) g = EXTRACT16(MIN32(Q15_ONE, MULT16_32_Q15(prior_ratio, MM))); /* Interpolated speech probability of presence */ p = st->gain2[i]; - + /* Constrain the gain to be close to the Bark scale gain */ if (MULT16_16_Q15(QCONST16(.333f,15),g) > st->gain[i]) g = MULT16_16(3,st->gain[i]); st->gain[i] = g; - + /* Save old power spectrum */ st->old_ps[i] = MULT16_32_P15(QCONST16(.2f,15),st->old_ps[i]) + MULT16_32_P15(MULT16_16_P15(QCONST16(.8f,15),SQR16_Q15(st->gain[i])),ps[i]); - + /* Apply gain floor */ if (st->gain[i] < st->gain_floor[i]) st->gain[i] = st->gain_floor[i]; /* Exponential decay model for reverberation (unused) */ /*st->reverb_estimate[i] = st->reverb_decay*st->reverb_estimate[i] + st->reverb_decay*st->reverb_level*st->gain[i]*st->gain[i]*st->ps[i];*/ - + /* Take into account speech probability of presence (loudness domain MMSE estimator) */ /* gain2 = [p*sqrt(gain)+(1-p)*sqrt(gain _floor) ]^2 */ tmp = MULT16_16_P15(p,spx_sqrt(SHL32(EXTEND32(st->gain[i]),15))) + MULT16_16_P15(SUB16(Q15_ONE,p),spx_sqrt(SHL32(EXTEND32(st->gain_floor[i]),15))); @@ -922,20 +922,20 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) { spx_word16_t tmp; spx_word16_t p = st->gain2[i]; - st->gain[i] = MAX16(st->gain[i], st->gain_floor[i]); + st->gain[i] = MAX16(st->gain[i], st->gain_floor[i]); tmp = MULT16_16_P15(p,spx_sqrt(SHL32(EXTEND32(st->gain[i]),15))) + MULT16_16_P15(SUB16(Q15_ONE,p),spx_sqrt(SHL32(EXTEND32(st->gain_floor[i]),15))); st->gain2[i]=SQR16_Q15(tmp); } filterbank_compute_psd16(st->bank,st->gain2+N, st->gain2); } - + /* If noise suppression is off, don't apply the gain (but then why call this in the first place!) */ if (!st->denoise_enabled) { for (i=0;i<N+M;i++) st->gain2[i]=Q15_ONE; } - + /* Apply computed gain */ for (i=1;i<N;i++) { @@ -944,7 +944,7 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) } st->ft[0] = MULT16_16_P15(st->gain2[0],st->ft[0]); st->ft[2*N-1] = MULT16_16_P15(st->gain2[N-1],st->ft[2*N-1]); - + /*FIXME: This *will* not work for fixed-point */ #ifndef FIXED_POINT if (st->agc_enabled) @@ -973,17 +973,17 @@ EXPORT int speex_preprocess_run(SpeexPreprocessState *st, spx_int16_t *x) } } #endif - + /* Synthesis window (for WOLA) */ for (i=0;i<2*N;i++) st->frame[i] = MULT16_16_Q15(st->frame[i], st->window[i]); /* Perform overlap and add */ for (i=0;i<N3;i++) - x[i] = st->outbuf[i] + st->frame[i]; + x[i] = WORD2INT(ADD32(EXTEND32(st->outbuf[i]), EXTEND32(st->frame[i]))); for (i=0;i<N4;i++) x[N3+i] = st->frame[N3+i]; - + /* Update outbuf */ for (i=0;i<N3;i++) st->outbuf[i] = st->frame[st->frame_size+i]; @@ -1016,11 +1016,11 @@ EXPORT void speex_preprocess_estimate_update(SpeexPreprocessState *st, spx_int16 M = st->nbands; st->min_count++; - + preprocess_analysis(st, x); update_noise_prob(st); - + for (i=1;i<N-1;i++) { if (!st->update_prob[i] || st->ps[i] < PSHR32(st->noise[i],NOISE_SHIFT)) @@ -1099,7 +1099,7 @@ EXPORT int speex_preprocess_ctl(SpeexPreprocessState *state, int request, void * case SPEEX_PREPROCESS_GET_VAD: (*(spx_int32_t*)ptr) = st->vad_enabled; break; - + case SPEEX_PREPROCESS_SET_DEREVERB: st->dereverb_enabled = (*(spx_int32_t*)ptr); for (i=0;i<st->ps_size;i++) @@ -1117,7 +1117,7 @@ EXPORT int speex_preprocess_ctl(SpeexPreprocessState *state, int request, void * /* FIXME: Re-enable when de-reverberation is actually enabled again */ /*(*(float*)ptr) = st->reverb_level;*/ break; - + case SPEEX_PREPROCESS_SET_DEREVERB_DECAY: /* FIXME: Re-enable when de-reverberation is actually enabled again */ /*st->reverb_decay = (*(float*)ptr);*/ |