MajenkoProjects/AudiobookRecorder
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Decimate by 4

Browse Source
This commit is contained in:
Matt Jenkins
2018-10-02 22:37:31 +01:00
parent b566b9e89e
commit 1244bd60ed
2 changed files with 12 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/uk/co/majenko/audiobookrecorder/AudiobookRecorder.java
View File

@@ -679,7 +679,7 @@ public class AudiobookRecorder extends JFrame {
sphinxConfig.setDictionaryPath("resource:/edu/cmu/sphinx/models/en-us/cmudict-en-us.dict");
sphinxConfig.setLanguageModelPath("resource:/edu/cmu/sphinx/models/en-us/en-us.lm.bin");
sphinxConfig.setSampleRate((int)(book.getAudioFormat().getSampleRate() / 3f));
sphinxConfig.setSampleRate((int)(book.getAudioFormat().getSampleRate() / 4f));
StreamSpeechRecognizer recognizer;

29
src/uk/co/majenko/audiobookrecorder/Sentence.java
View File

@@ -579,30 +579,23 @@ public class Sentence extends DefaultMutableTreeNode implements Cacheable {
setText("[recognising...]");
AudiobookRecorder.window.bookTreeModel.reload(this);
AudioInputStream s = AudioSystem.getAudioInputStream(getFile());
AudioFormat format = getAudioFormat();
int frameSize = format.getFrameSize();
int length = (int)s.getFrameLength();
byte[] data = new byte[length * frameSize];
byte[] inData = getRawAudioData();
int inLength = inData.length;
int bps = format.getFrameSize();
int inSamples = inLength / bps;
s.read(data);
int outSamples = inSamples / 4;
byte[] outData = new byte[outSamples * bps];
int channels = format.getChannels();
int newLen = (length / 3);
byte[] decimated = new byte[newLen * 2];
for (int i = 0; i < newLen; i++) {
if (channels == 1) {
decimated[i * 2] = data[i * 6];
decimated[i * 2 + 1] = data[i * 6 + 1];
} else {
decimated[i * 2] = data[i * 12];
decimated[i * 2 + 1] = data[i * 12 + 1];
for (int i = 0; i < outSamples; i++) {
for (int j = 0; j < bps; j++) {
outData[i * bps + j] = inData[(i * 4) * bps + j];
}
}
ByteArrayInputStream bas = new ByteArrayInputStream(decimated);
ByteArrayInputStream bas = new ByteArrayInputStream(outData);
recognizer.startRecognition(bas);
SpeechResult result;
String res = "";
@@ -633,7 +626,7 @@ public class Sentence extends DefaultMutableTreeNode implements Cacheable {
AudioInputStream s = AudioSystem.getAudioInputStream(getFile());
AudioFormat format = getAudioFormat();
sphinxConfig.setSampleRate((int)(format.getSampleRate() / 3f));
sphinxConfig.setSampleRate((int)(format.getSampleRate() / 4f));
StreamSpeechRecognizer recognizer;