Centre for Digital Music, Queen Mary, University of London.
This file copyright 2006-2008 Chris Cannam and QMUL.
- This program is free software; you can redistribute it and/or
- modify it under the terms of the GNU General Public License as
- published by the Free Software Foundation; either version 2 of the
- License, or (at your option) any later version. See the file
- COPYING included with this distribution for more information.
+ This file is part of vamp-aubio-plugins.
+
+ vamp-aubio is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ vamp-aubio is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with aubio. If not, see <http://www.gnu.org/licenses/>.
+
*/
#include <math.h>
using std::cerr;
using std::endl;
-Silence::Silence(float inputSampleRate, unsigned int apiVersion) :
+Silence::Silence(float inputSampleRate) :
Plugin(inputSampleRate),
- m_apiVersion(apiVersion),
m_ibuf(0),
m_pbuf(0),
- m_tmpptrs(0),
m_threshold(-80),
m_prevSilent(false),
m_first(true)
{
- if (m_apiVersion == 1) {
- cerr << "vamp-aubio: WARNING: using compatibility version 1 of the Vamp API for silence\n"
- << "detector plugin: upgrade your host to v2 for proper duration support" << endl;
- }
}
Silence::~Silence()
{
if (m_ibuf) del_fvec(m_ibuf);
if (m_pbuf) del_fvec(m_pbuf);
- if (m_tmpptrs) delete[] m_tmpptrs;
}
string
int
Silence::getPluginVersion() const
{
- if (m_apiVersion == 1) return 2;
- return 3;
+ return 4;
}
string
bool
Silence::initialise(size_t channels, size_t stepSize, size_t blockSize)
{
- m_channelCount = channels;
+ if (channels != 1) {
+ std::cerr << "Silence::initialise: channels must be 1" << std::endl;
+ return false;
+ }
+
m_stepSize = stepSize;
m_blockSize = blockSize;
- m_ibuf = new_fvec(stepSize, channels);
- m_pbuf = new_fvec(stepSize, channels);
- m_tmpptrs = new smpl_t *[channels];
+ m_ibuf = new_fvec(stepSize);
+ m_pbuf = new_fvec(stepSize);
return true;
}
desc = ParameterDescriptor();
desc.identifier = "silencethreshold";
desc.name = "Silence Threshold";
+ desc.description = "Threshold for silence detection";
desc.minValue = -120;
desc.maxValue = 0;
desc.defaultValue = -80;
OutputDescriptor d;
- if (m_apiVersion == 1) {
-
- d.identifier = "silencestart";
- d.name = "Beginnings of Silent Regions";
- d.description = "Return a single instant at the point where each silent region begins";
- d.hasFixedBinCount = true;
- d.binCount = 0;
- d.hasKnownExtents = false;
- d.sampleType = OutputDescriptor::VariableSampleRate;
- d.sampleRate = 0;
- list.push_back(d);
-
- d.identifier = "silenceend";
- d.name = "Ends of Silent Regions";
- d.description = "Return a single instant at the point where each silent region ends";
- d.hasFixedBinCount = true;
- d.binCount = 0;
- d.hasKnownExtents = false;
- d.sampleType = OutputDescriptor::VariableSampleRate;
- d.sampleRate = 0;
- list.push_back(d);
-
- } else {
+ d.identifier = "silent";
+ d.name = "Silent Regions";
+ d.description = "Return an interval covering each silent region";
+ d.hasFixedBinCount = true;
+ d.binCount = 0;
+ d.hasKnownExtents = false;
+ d.sampleType = OutputDescriptor::VariableSampleRate;
+ d.sampleRate = 0;
+ d.hasDuration = true;
+ list.push_back(d);
- d.identifier = "silent";
- d.name = "Silent Regions";
- d.description = "Return an interval covering each silent region";
- d.hasFixedBinCount = true;
- d.binCount = 0;
- d.hasKnownExtents = false;
- d.sampleType = OutputDescriptor::VariableSampleRate;
- d.sampleRate = 0;
- list.push_back(d);
-
- d.identifier = "noisy";
- d.name = "Non-Silent Regions";
- d.description = "Return an interval covering each non-silent region";
- d.hasFixedBinCount = true;
- d.binCount = 0;
- d.hasKnownExtents = false;
- d.sampleType = OutputDescriptor::VariableSampleRate;
- d.sampleRate = 0;
- list.push_back(d);
- }
+ d.identifier = "noisy";
+ d.name = "Non-Silent Regions";
+ d.description = "Return an interval covering each non-silent region";
+ d.hasFixedBinCount = true;
+ d.binCount = 0;
+ d.hasKnownExtents = false;
+ d.sampleType = OutputDescriptor::VariableSampleRate;
+ d.sampleRate = 0;
+ d.hasDuration = true;
+ list.push_back(d);
d.identifier = "silencelevel";
d.name = "Silence Test";
Vamp::RealTime timestamp)
{
for (size_t i = 0; i < m_stepSize; ++i) {
- for (size_t j = 0; j < m_channelCount; ++j) {
- fvec_write_sample(m_ibuf, inputBuffers[j][i], j, i);
- }
+ fvec_set_sample(m_ibuf, inputBuffers[0][i], i);
}
bool silent = aubio_silence_detection(m_ibuf, m_threshold);
fvec_t vec;
vec.length = incr * 4;
- vec.channels = m_channelCount;
- vec.data = m_tmpptrs;
for (size_t i = 0; i < m_stepSize - incr * 4; i += incr) {
- for (size_t j = 0; j < m_channelCount; ++j) {
- m_tmpptrs[j] = m_ibuf->data[j] + i;
- }
+ vec.data = m_ibuf->data + i;
bool subsilent = aubio_silence_detection(&vec, m_threshold);
if (silent == subsilent) {
off = i;
if (silent && (off == 0)) {
for (size_t i = 0; i < m_stepSize - incr; i += incr) {
- for (size_t j = 0; j < m_channelCount; ++j) {
- m_tmpptrs[j] = m_pbuf->data[j] + m_stepSize - i - incr;
- }
+ vec.data = m_pbuf->data + m_stepSize - i - incr;
bool subsilent = aubio_silence_detection(&vec, m_threshold);
if (!subsilent) {
off = -(long)i;
feature.values.clear();
- if (m_apiVersion == 1) {
+ if (!m_first) {
+ feature.timestamp = m_lastChange;
+ feature.hasDuration = true;
+ feature.duration = featureStamp - m_lastChange;
if (silent) {
- returnFeatures[0].push_back(feature);
- } else {
+ // non-silent regions feature
+ // (becoming silent, so this is a non-silent region)
returnFeatures[1].push_back(feature);
- }
- } else {
- if (!m_first) {
- feature.timestamp = m_lastChange;
- feature.hasDuration = true;
- feature.duration = featureStamp - m_lastChange;
- if (silent) {
- // becoming silent, so this is a non-silent region
- returnFeatures[1].push_back(feature);
- } else {
- // becoming non-silent, so this is a silent region
- returnFeatures[0].push_back(feature);
- }
- }
- m_lastChange = featureStamp;
+ } else {
+ // silent regions feature
+ // (becoming non-silent, so this is a silent region)
+ returnFeatures[0].push_back(feature);
+ }
}
+ m_lastChange = featureStamp;
m_prevSilent = silent;
m_first = false;
// swap ibuf and pbuf data pointers, so that this block's data is
// available in pbuf when processing the next block, without
// having to allocate new storage for it
- smpl_t **tmpdata = m_ibuf->data;
+ smpl_t *tmpdata = m_ibuf->data;
m_ibuf->data = m_pbuf->data;
m_pbuf->data = tmpdata;
{
FeatureSet returnFeatures;
- if (m_prevSilent) {
- if (m_lastTimestamp > m_lastChange) {
- Feature feature;
- feature.hasTimestamp = true;
- feature.timestamp = m_lastChange;
- feature.hasDuration = true;
- feature.duration = m_lastTimestamp - m_lastChange;
- if (m_prevSilent) {
- returnFeatures[0].push_back(feature);
- } else {
- returnFeatures[1].push_back(feature);
- }
+// std::cerr << "Silence::getRemainingFeatures: m_lastTimestamp = " << m_lastTimestamp << ", m_lastChange = " << m_lastChange << ", m_apiVersion = " << m_apiVersion << ", m_prevSilent = " << m_prevSilent << std::endl;
+
+ if (m_lastTimestamp > m_lastChange) {
+
+ Feature feature;
+ feature.hasTimestamp = true;
+
+ feature.timestamp = m_lastChange;
+ feature.hasDuration = true;
+ feature.duration = m_lastTimestamp - m_lastChange;
+ if (m_prevSilent) {
+ // silent regions feature
+ returnFeatures[0].push_back(feature);
+ } else {
+ // non-silent regions feature
+ returnFeatures[1].push_back(feature);
+ }
+
+ if (!m_prevSilent) {
+ Feature silenceTestFeature;
+ silenceTestFeature.hasTimestamp = true;
+ silenceTestFeature.timestamp = m_lastTimestamp;
+ silenceTestFeature.values.push_back(0);
+ returnFeatures[2].push_back(silenceTestFeature);
}
}
- return FeatureSet();
+ return returnFeatures;
}