fix: vadMs sample lose on convert type

mybigday · Oct 19, 2023 · 59243ec · 59243ec
1 parent 244f215
commit 59243ec
Show file tree

Hide file tree

Showing 4 changed files with 9 additions and 9 deletions.
diff --git a/android/src/main/java/com/rnwhisper/WhisperContext.java b/android/src/main/java/com/rnwhisper/WhisperContext.java
@@ -82,9 +82,9 @@ private void rewind() {
   private boolean vad(ReadableMap options, short[] shortBuffer, int nSamples, int n) {
     boolean isSpeech = true;
     if (!isTranscribing && options.hasKey("useVad") && options.getBoolean("useVad")) {
-      int vadSec = options.hasKey("vadMs") ? options.getInt("vadMs") / 1000 : 2;
-      if (vadSec < 2) vadSec = 2;
-      int sampleSize = vadSec * SAMPLE_RATE;
+      int vadMs = options.hasKey("vadMs") ? options.getInt("vadMs") : 2000;
+      if (vadMs < 2000) vadMs = 2000;
+      int sampleSize = (int) (SAMPLE_RATE * vadMs / 1000);
       if (nSamples + n > sampleSize) {
         int start = nSamples + n - sampleSize;
         float[] audioData = new float[sampleSize];

diff --git a/example/ios/Podfile.lock b/example/ios/Podfile.lock
@@ -764,7 +764,7 @@ PODS:
     - SSZipArchive (~> 2.2)
   - SocketRocket (0.6.0)
   - SSZipArchive (2.4.3)
-  - whisper-rn (0.4.0-rc.1):
+  - whisper-rn (0.4.0-rc.2):
     - RCT-Folly
     - RCTRequired
     - RCTTypeSafety
@@ -1006,7 +1006,7 @@ SPEC CHECKSUMS:
   RNZipArchive: ef9451b849c45a29509bf44e65b788829ab07801
   SocketRocket: fccef3f9c5cedea1353a9ef6ada904fde10d6608
   SSZipArchive: fe6a26b2a54d5a0890f2567b5cc6de5caa600aef
-  whisper-rn: 6dc1b9a55eb0d9e9c0cb00999660c04c32d6732f
+  whisper-rn: 02c39fead176096e6d8420c4f7f4326f122b36e3
   Yoga: f7decafdc5e8c125e6fa0da38a687e35238420fa
   YogaKit: f782866e155069a2cca2517aafea43200b01fd5a
 

diff --git a/ios/RNWhisperContext.h b/ios/RNWhisperContext.h
@@ -30,7 +30,7 @@ typedef struct {
     NSString* audioOutputPath;
 
     bool useVad;
-    int vadSec;
+    int vadMs;
     float vadThold;
     float vadFreqThold;
 

diff --git a/ios/RNWhisperContext.mm b/ios/RNWhisperContext.mm
@@ -56,8 +56,8 @@ - (void)prepareRealtime:(NSDictionary *)options {
     self->recordState.audioOutputPath = options[@"audioOutputPath"];
 
     self->recordState.useVad = options[@"useVad"] != nil ? [options[@"useVad"] boolValue] : false;
-    self->recordState.vadSec = options[@"vadMs"] != nil ? [options[@"vadMs"] intValue] / 1000 : 2;
-    if (self->recordState.vadSec < 2) self->recordState.vadSec = 2;
+    self->recordState.vadMs = options[@"vadMs"] != nil ? [options[@"vadMs"] intValue] : 2000;
+    if (self->recordState.vadMs < 2000) self->recordState.vadMs = 2000;
 
     self->recordState.vadThold = options[@"vadThold"] != nil ? [options[@"vadThold"] floatValue] : 0.6f;
     self->recordState.vadFreqThold = options[@"vadFreqThold"] != nil ? [options[@"vadFreqThold"] floatValue] : 100.0f;
@@ -100,7 +100,7 @@ bool vad(RNWhisperContextRecordState *state, int16_t* audioBufferI16, int nSampl
 {
     bool isSpeech = true;
     if (!state->isTranscribing && state->useVad) {
-        int sampleSize = state->vadSec * WHISPER_SAMPLE_RATE;
+        int sampleSize = (int) (WHISPER_SAMPLE_RATE * state->vadMs / 1000);
         if (nSamples + n > sampleSize) {
             int start = nSamples + n - sampleSize;
             std::vector<float> audioBufferF32Vec(sampleSize);