Files
voice-cat/third_party/rnnoise/src/pitch.h

148 lines
4.5 KiB
C
Raw Normal View History

feat(audio): real noise suppression via vendored RNNoise (send + receive) The two-sided NR plumbing (RemoteStream::recv_ns + the per-listener vc_set_remote_stream noise_reduction toggle) was wired but inert: ApmProcessor::create() returned a no-op passthrough, because the originally-planned webrtc-audio-processing has no working Windows/macOS build. Drop in RNNoise as the real backend behind the same ApmProcessor interface, lighting up both NR paths. - Vendor RNNoise (BSD-3 + CC0) at third_party/rnnoise/ — the vcpkg port is !windows !arm, so it can't cover our primary targets. Shrunk int8 model (78MB -> 11.7MB via upstream scripts/shrink_model.sh), built as a standalone C static lib with no RTCD (portable scalar path on x86, auto-NEON on arm64) under -DDISABLE_DEBUG_FLOAT. Model is baked in (rnnoise_create(NULL)); no runtime file. - New RnnoiseProcessor (core/src/audio/apm_processor.cpp) selected by ApmProcessor::create() when VOICECAT_HAS_NS. Mono/48kHz/480-sample; our clock is fixed 48kHz and Opus frame sizes are multiples of 480, so no resampling. RT-safe: allocates at construction, lock-free in the capture/playback callbacks. - Receive-side: lit up via the factory; gated to mono streams (a stereo stream is a screen-audio share, not voice). - Send-side (new): vc_set_input_noise_reduction(client, enable) ABI + vc_client::mic_ns_, run before input gain/VAD in on_capture_frame. A stereo mic is downmixed to mono ONLY when NR is on — with NR off a stereo mic keeps full stereo (never collapse mic quality unasked). - Enable C as a project language for the vendored lib. - New noise_suppression test: white noise through ApmProcessor::create() drops ~99.9% RMS. ctest --preset dev green, 28/28. windows-client DLL builds clean with vc_set_input_noise_reduction exported, system-only deps. - Docs synced: voice.md §10, tech-stack.md §1/§5, third_party/README.md, vcpkg.json note, PROGRESS.md, CLAUDE.md. Client on/off UI toggles (Windows/macOS/iOS) are the remaining follow-up. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-23 13:30:54 +02:00
/* Copyright (c) 2007-2008 CSIRO
Copyright (c) 2007-2009 Xiph.Org Foundation
Written by Jean-Marc Valin */
/**
@file pitch.h
@brief Pitch analysis
*/
/*
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
- Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef PITCH_H
#define PITCH_H
#include "arch.h"
void rnn_pitch_downsample(celt_sig *x[], opus_val16 *x_lp,
int len, int C);
void rnn_pitch_search(const opus_val16 *x_lp, opus_val16 *y,
int len, int max_pitch, int *pitch);
opus_val16 rnn_remove_doubling(opus_val16 *x, int maxperiod, int minperiod,
int N, int *T0, int prev_period, opus_val16 prev_gain);
/* OPT: This is the kernel you really want to optimize. It gets used a lot
by the prefilter and by the PLC. */
static OPUS_INLINE void xcorr_kernel(const opus_val16 * x, const opus_val16 * y, opus_val32 sum[4], int len)
{
int j;
opus_val16 y_0, y_1, y_2, y_3;
celt_assert(len>=3);
y_3=0; /* gcc doesn't realize that y_3 can't be used uninitialized */
y_0=*y++;
y_1=*y++;
y_2=*y++;
for (j=0;j<len-3;j+=4)
{
opus_val16 tmp;
tmp = *x++;
y_3=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_0);
sum[1] = MAC16_16(sum[1],tmp,y_1);
sum[2] = MAC16_16(sum[2],tmp,y_2);
sum[3] = MAC16_16(sum[3],tmp,y_3);
tmp=*x++;
y_0=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_1);
sum[1] = MAC16_16(sum[1],tmp,y_2);
sum[2] = MAC16_16(sum[2],tmp,y_3);
sum[3] = MAC16_16(sum[3],tmp,y_0);
tmp=*x++;
y_1=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_2);
sum[1] = MAC16_16(sum[1],tmp,y_3);
sum[2] = MAC16_16(sum[2],tmp,y_0);
sum[3] = MAC16_16(sum[3],tmp,y_1);
tmp=*x++;
y_2=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_3);
sum[1] = MAC16_16(sum[1],tmp,y_0);
sum[2] = MAC16_16(sum[2],tmp,y_1);
sum[3] = MAC16_16(sum[3],tmp,y_2);
}
if (j++<len)
{
opus_val16 tmp = *x++;
y_3=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_0);
sum[1] = MAC16_16(sum[1],tmp,y_1);
sum[2] = MAC16_16(sum[2],tmp,y_2);
sum[3] = MAC16_16(sum[3],tmp,y_3);
}
if (j++<len)
{
opus_val16 tmp=*x++;
y_0=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_1);
sum[1] = MAC16_16(sum[1],tmp,y_2);
sum[2] = MAC16_16(sum[2],tmp,y_3);
sum[3] = MAC16_16(sum[3],tmp,y_0);
}
if (j<len)
{
opus_val16 tmp=*x++;
y_1=*y++;
sum[0] = MAC16_16(sum[0],tmp,y_2);
sum[1] = MAC16_16(sum[1],tmp,y_3);
sum[2] = MAC16_16(sum[2],tmp,y_0);
sum[3] = MAC16_16(sum[3],tmp,y_1);
}
}
static OPUS_INLINE void dual_inner_prod(const opus_val16 *x, const opus_val16 *y01, const opus_val16 *y02,
int N, opus_val32 *xy1, opus_val32 *xy2)
{
int i;
opus_val32 xy01=0;
opus_val32 xy02=0;
for (i=0;i<N;i++)
{
xy01 = MAC16_16(xy01, x[i], y01[i]);
xy02 = MAC16_16(xy02, x[i], y02[i]);
}
*xy1 = xy01;
*xy2 = xy02;
}
/*We make sure a C version is always available for cases where the overhead of
vectorization and passing around an arch flag aren't worth it.*/
static OPUS_INLINE opus_val32 celt_inner_prod(const opus_val16 *x,
const opus_val16 *y, int N)
{
int i;
opus_val32 xy=0;
for (i=0;i<N;i++)
xy = MAC16_16(xy, x[i], y[i]);
return xy;
}
void rnn_pitch_xcorr(const opus_val16 *_x, const opus_val16 *_y,
opus_val32 *xcorr, int len, int max_pitch);
#endif