Update

sv.py

@@ -194,7 +194,7 @@ def format_time(seconds, use_short_format=True, always_use_seconds=False):
 
     if always_use_seconds or (use_short_format and hours == 0 and minutes == 0):
         return f"{seconds:06.3f}s"
-    elif use_short_format or hours == 0:
+    elif use_short_format and hours == 0:
         return f"{minutes:02d}:{seconds:06.3f}"
     else:
         return f"{hours:02d}:{minutes:02d}:{seconds:06.3f}"
@@ -275,7 +275,7 @@ def process_audio(audio_path, language="yue", fs=16000):
 
     # Determine if the audio is less than one minute
     total_duration = sum(duration for _, _, duration, _ in diarization_segments)
-    use_short_format = total_duration < 60
+    use_long_format = total_duration >= 60
 
     # Process the audio in chunks based on diarization segments
     results = []
@@ -321,9 +321,19 @@ def process_audio(audio_path, language="yue", fs=16000):
     # Format the results
     formatted_text = ""
     for speaker, start, end, duration, text in results:
-        start_str = format_time(start, always_use_seconds=True)
-        end_str = format_time(end, always_use_seconds=True)
-        duration_str = format_time(duration, always_use_seconds=True)
+        start_str = (
+            format_time(start, use_short_format=False)
+            if use_long_format
+            else format_time(start, use_short_format=True)
+        )
+        end_str = (
+            format_time(end, use_short_format=False)
+            if use_long_format
+            else format_time(end, use_short_format=True)
+        )
+        duration_str = format_time(
+            duration, use_short_format=True
+        )  # Always use short format for duration
         speaker_num = "1" if speaker == "SPEAKER_00" else "2"
         line = f"{start_str} - {end_str} ({duration_str}) Speaker {speaker_num}: {text}"
         formatted_text += line + "\n"