diff --git a/discord/voice_client.py b/discord/voice_client.py
index 24a4081ae..ecbc4ee40 100644
--- a/discord/voice_client.py
+++ b/discord/voice_client.py
@@ -386,12 +386,14 @@ class VoiceClient:
         """
         command = 'ffmpeg' if not use_avconv else 'avconv'
         input_name = '-' if pipe else shlex.quote(filename)
-        cmd = command + ' -i {} -f s16le -ar {} -ac {} -loglevel warning pipe:1'
+        cmd = command + ' -i {} -f s16le -ar {} -ac {} -loglevel warning'
         cmd = cmd.format(input_name, self.encoder.sampling_rate, self.encoder.channels)
 
         if isinstance(options, str):
             cmd = cmd + ' ' + options
 
+        cmd += ' pipe:1'
+
         stdin = None if not pipe else filename
         args = shlex.split(cmd)
         try:
@@ -567,4 +569,3 @@ class VoiceClient:
         packet = self._get_voice_packet(encoded_data)
         sent = self.socket.sendto(packet, (self.endpoint_ip, self.voice_port))
         self.checked_add('timestamp', self.encoder.samples_per_frame, 4294967295)
-