diff --git a/faster_whisper/vad.py b/faster_whisper/vad.py
index 080795d..cf14d5c 100644
--- a/faster_whisper/vad.py
+++ b/faster_whisper/vad.py
@@ -20,7 +20,7 @@ def get_speech_timestamps(
     max_speech_duration_s: float = float("inf"),
     min_silence_duration_ms: int = 2000,
     window_size_samples: int = 1024,
-    speech_pad_ms: int = 200,
+    speech_pad_ms: int = 400,
 ) -> List[dict]:
     """This method is used for splitting long audios into speech chunks using silero VAD.
 
diff --git a/tests/test_transcribe.py b/tests/test_transcribe.py
index 5406535..8bebd2a 100644
--- a/tests/test_transcribe.py
+++ b/tests/test_transcribe.py
@@ -32,7 +32,7 @@ def test_vad(jfk_path):
     segments, _ = model.transcribe(
         jfk_path,
         vad_filter=True,
-        vad_parameters=dict(min_silence_duration_ms=500),
+        vad_parameters=dict(min_silence_duration_ms=500, speech_pad_ms=200),
     )
     segments = list(segments)