diff --git a/disco/voice/__init__.py b/disco/voice/__init__.py
index 6a60306..b4a7f6c 100644
--- a/disco/voice/__init__.py
+++ b/disco/voice/__init__.py
@@ -1,2 +1,3 @@
 from disco.voice.client import *
 from disco.voice.player import *
+from disco.voice.playable import *
diff --git a/disco/voice/opus.py b/disco/voice/opus.py
index 1884d73..b40f6b5 100644
--- a/disco/voice/opus.py
+++ b/disco/voice/opus.py
@@ -156,125 +156,3 @@ class OpusEncoder(BaseOpus):
 
 class OpusDecoder(BaseOpus):
     pass
-
-
-class BufferedOpusEncoder(OpusEncoder):
-    def __init__(self, source, *args, **kwargs):
-        self.source = source
-        self.frames = Queue(kwargs.pop('queue_size', 4096))
-        super(BufferedOpusEncoder, self).__init__(*args, **kwargs)
-        gevent.spawn(self._encoder_loop)
-
-    def _encoder_loop(self):
-        while self.source:
-            raw = self.source.read(self.frame_size)
-            if len(raw) < self.frame_size:
-                break
-
-            self.frames.put(self.encode(raw, self.samples_per_frame))
-            gevent.idle()
-        self.source = None
-
-    def have_frame(self):
-        return self.source or not self.frames.empty()
-
-    def next_frame(self):
-        return self.frames.get()
-
-
-class GIPCBufferedOpusEncoder(OpusEncoder):
-    FIN = 1
-
-    def __init__(self, source, *args, **kwargs):
-        import gipc
-
-        self.source = source
-        self.parent_pipe, self.child_pipe = gipc.pipe(duplex=True)
-        self.frames = Queue(kwargs.pop('queue_size', 4096))
-        super(GIPCBufferedOpusEncoder, self).__init__(*args, **kwargs)
-
-        gipc.start_process(target=self._encoder_loop, args=(self.child_pipe, (args, kwargs)))
-
-        gevent.spawn(self._writer)
-        gevent.spawn(self._reader)
-
-    def _reader(self):
-        while True:
-            data = self.parent_pipe.get()
-            if data == self.FIN:
-                return
-
-            self.frames.put(data)
-        self.parent_pipe = None
-
-    def _writer(self):
-        while self.data:
-            raw = self.source.read(self.frame_size)
-            if len(raw) < self.frame_size:
-                break
-
-            self.parent_pipe.put(raw)
-            gevent.idle()
-
-        self.parent_pipe.put(self.FIN)
-
-    def have_frame(self):
-        return self.parent_pipe
-
-    def next_frame(self):
-        return self.frames.get()
-
-    @classmethod
-    def _encoder_loop(cls, pipe, (args, kwargs)):
-        encoder = OpusEncoder(*args, **kwargs)
-
-        while True:
-            data = pipe.get()
-            if data == cls.FIN:
-                pipe.put(cls.FIN)
-                return
-
-            pipe.put(encoder.encode(data, encoder.samples_per_frame))
-
-
-class DCADOpusEncoder(OpusEncoder):
-    def __init__(self, source, *args, **kwargs):
-        self.source = source
-        self.command = kwargs.pop('command', 'dcad')
-        super(DCADOpusEncoder, self).__init__(*args, **kwargs)
-        self._proc = None
-        self.header_size = struct.calcsize('<h')
-
-    @property
-    def proc(self):
-        if not self._proc:
-            self._proc = subprocess.Popen([
-                self.command,
-                '--channels', str(self.channels),
-                '--rate', str(self.sampling_rate),
-                '--size', str(self.samples_per_frame),
-                '--bitrate', '128',
-                '--fec',
-                '--packet-loss-percent', '30',
-                '--input', 'pipe:0',
-                '--output', 'pipe:1',
-            ], stdin=self.source.stdout, stdout=subprocess.PIPE)
-        return self._proc
-
-    def have_frame(self):
-        return bool(self.proc)
-
-    def next_frame(self):
-        header = self.proc.stdout.read(self.header_size)
-        if len(header) < self.header_size:
-            self._proc = None
-            return
-
-        size = struct.unpack('<h', header)[0]
-
-        data = self.proc.stdout.read(size)
-        if len(data) == 0:
-            self._proc = None
-            return
-
-        return data
diff --git a/disco/voice/playable.py b/disco/voice/playable.py
new file mode 100644
index 0000000..263d838
--- /dev/null
+++ b/disco/voice/playable.py
@@ -0,0 +1,290 @@
+import gevent
+import struct
+import subprocess
+
+
+from gevent.queue import Queue
+
+from disco.voice.opus import OpusEncoder
+
+
+try:
+    from cStringIO import cStringIO as StringIO
+except:
+    from StringIO import StringIO
+
+
+OPUS_HEADER_SIZE = struct.calcsize('<h')
+
+
+class BasePlayable(object):
+    pass
+
+
+
+class FFmpegPlayable(BasePlayable):
+    def __init__(self, source='-', command='avconv', sampling_rate=48000, channels=2, **kwargs):
+        self.source = source
+        self.command = command
+        self.sampling_rate = sampling_rate
+        self.channels = channels
+        self.kwargs = kwargs
+
+        self._buffer = None
+        self._proc = None
+        self._child = None
+
+    @classmethod
+    def create(_cls, *args, **kwargs):
+        cls = kwargs.pop('cls', BufferedOpusEncoder)
+        playable = _cls(*args, **kwargs)
+        playable.pipe(cls)
+        return playable
+
+    @classmethod
+    def create_youtube_dl(_cls, url, *args, **kwargs):
+        import youtube_dl
+
+        ydl = youtube_dl.YoutubeDL({'format': 'webm[abr>0]/bestaudio/best'})
+        info = ydl.extract_info(url, download=False)
+        entries = [info] if 'entries' not in info else info['entries']
+
+        for entry in entries:
+            playable = _cls.create(entry['url'], *args, **kwargs)
+            playable.info = entry
+            yield playable
+
+    @property
+    def stdout(self):
+        return self.proc.stdout
+
+    def read(self, sz):
+        if self.streaming:
+            return self.proc.stdout.read(sz)
+        else:
+            if not self._buffer:
+                data, _ = self.proc.communicate()
+                self._buffer = StringIO(data)
+            return self._buffer.read(sz)
+
+    def pipe(self, other, streaming=True):
+        if issubclass(other, OpusEncoder):
+            self._child = other(self, self.sampling_rate, self.channels, **self.kwargs)
+        else:
+            raise TypeError('Invalid pipe target')
+
+    @property
+    def samples_per_frame(self):
+        return self._child.samples_per_frame
+
+    @property
+    def proc(self):
+        if not self._proc:
+            args = [
+                self.command,
+                '-i', self.source,
+                '-f', 's16le',
+                '-ar', str(self.sampling_rate),
+                '-ac', str(self.channels),
+                '-loglevel', 'warning',
+                'pipe:1'
+            ]
+            self._proc = subprocess.Popen(args, stdin=None, stdout=subprocess.PIPE)
+        return self._proc
+
+    def have_frame(self):
+        return self._child and self._child.have_frame()
+
+    def next_frame(self):
+        return self._child.next_frame()
+
+
+class OpusFilePlayable(BasePlayable):
+    """
+    A Playable which supports reading from an on-disk opus-format file. This is
+    useful in combination with other playables and the OpusFileOutputDuplex.
+    """
+
+    def __init__(self, obj, sampling_rate=48000, frame_length=20, channels=2):
+        super(OpusFilePlayable, self).__init__()
+        self.obj = obj
+        self.sampling_rate = sampling_rate
+        self.frame_length = frame_length
+        self.channels = channels
+        self.samples_per_frame = int(self.sampling_rate / 1000 * self.frame_length)
+
+    def have_frame(self):
+        return self.obj
+
+    def next_frame(self):
+        header = self.obj.read(OPUS_HEADER_SIZE)
+        if len(header) < OPUS_HEADER_SIZE:
+            self.obj = None
+            return
+
+        size = struct.unpack('<h', header)[0]
+        raw = self.obj.read(size)
+        if len(raw) < size:
+            self.obj = None
+            return
+
+        return raw
+
+
+class BufferedOpusEncoder(BasePlayable, OpusEncoder):
+    def __init__(self, source, *args, **kwargs):
+        self.source = source
+        self.frames = Queue(kwargs.pop('queue_size', 4096))
+        super(BufferedOpusEncoder, self).__init__(*args, **kwargs)
+        gevent.spawn(self._encoder_loop)
+
+    def _encoder_loop(self):
+        while self.source:
+            raw = self.source.read(self.frame_size)
+            if len(raw) < self.frame_size:
+                break
+
+            self.frames.put(self.encode(raw, self.samples_per_frame))
+            gevent.idle()
+        self.source = None
+
+    def have_frame(self):
+        return self.source or not self.frames.empty()
+
+    def next_frame(self):
+        return self.frames.get()
+
+
+class GIPCBufferedOpusEncoder(BasePlayable, OpusEncoder):
+    FIN = 1
+
+    def __init__(self, source, *args, **kwargs):
+        import gipc
+
+        self.source = source
+        self.parent_pipe, self.child_pipe = gipc.pipe(duplex=True)
+        self.frames = Queue(kwargs.pop('queue_size', 4096))
+        super(GIPCBufferedOpusEncoder, self).__init__(*args, **kwargs)
+
+        gipc.start_process(target=self._encoder_loop, args=(self.child_pipe, (args, kwargs)))
+
+        gevent.spawn(self._writer)
+        gevent.spawn(self._reader)
+
+    def _reader(self):
+        while True:
+            data = self.parent_pipe.get()
+            if data == self.FIN:
+                return
+
+            self.frames.put(data)
+        self.parent_pipe = None
+
+    def _writer(self):
+        while self.data:
+            raw = self.source.read(self.frame_size)
+            if len(raw) < self.frame_size:
+                break
+
+            self.parent_pipe.put(raw)
+            gevent.idle()
+
+        self.parent_pipe.put(self.FIN)
+
+    def have_frame(self):
+        return self.parent_pipe
+
+    def next_frame(self):
+        return self.frames.get()
+
+    @classmethod
+    def _encoder_loop(cls, pipe, (args, kwargs)):
+        encoder = OpusEncoder(*args, **kwargs)
+
+        while True:
+            data = pipe.get()
+            if data == cls.FIN:
+                pipe.put(cls.FIN)
+                return
+
+            pipe.put(encoder.encode(data, encoder.samples_per_frame))
+
+
+class DCADOpusEncoder(BasePlayable, OpusEncoder):
+    def __init__(self, source, *args, **kwargs):
+        self.source = source
+        self.command = kwargs.pop('command', 'dcad')
+        super(DCADOpusEncoder, self).__init__(*args, **kwargs)
+        self._proc = None
+
+    @property
+    def proc(self):
+        if not self._proc:
+            self._proc = subprocess.Popen([
+                self.command,
+                '--channels', str(self.channels),
+                '--rate', str(self.sampling_rate),
+                '--size', str(self.samples_per_frame),
+                '--bitrate', '128',
+                '--fec',
+                '--packet-loss-percent', '30',
+                '--input', 'pipe:0',
+                '--output', 'pipe:1',
+            ], stdin=self.source.stdout, stdout=subprocess.PIPE)
+        return self._proc
+
+    def have_frame(self):
+        return bool(self.proc)
+
+    def next_frame(self):
+        header = self.proc.stdout.read(OPUS_HEADER_SIZE)
+        if len(header) < OPUS_HEADER_SIZE:
+            self._proc = None
+            return
+
+        size = struct.unpack('<h', header)[0]
+
+        data = self.proc.stdout.read(size)
+        if len(data) == 0:
+            self._proc = None
+            return
+
+        return data
+
+
+class OpusDuplexEncoder(BasePlayable):
+    def __init__(self, out, encoder, flush=False):
+        self.out = out
+        self.flush = flush
+        self.encoder = encoder
+        self.closed = False
+
+    def __getattr__(self, attr):
+        return getattr(self.encoder, attr)
+
+    def close(self):
+        if self.closed:
+            return
+
+        self.closed = True
+        self.out.flush()
+        self.out.close()
+
+    def have_frame(self):
+        have = self.encoder.have_frame()
+        if not have:
+            self.close()
+        return have
+
+    def next_frame(self):
+        frame = self.encoder.next_frame()
+
+        if frame:
+            self.out.write(struct.pack('<h', len(frame)))
+            self.out.write(frame)
+
+            if self.flush:
+                self.out.flush()
+        else:
+            self.close()
+        return frame
diff --git a/disco/voice/player.py b/disco/voice/player.py
index 3ca2bcc..0031146 100644
--- a/disco/voice/player.py
+++ b/disco/voice/player.py
@@ -1,129 +1,11 @@
 import time
 import gevent
-import struct
-import subprocess
 
 from six.moves import queue
 from holster.enum import Enum
 from holster.emitter import Emitter
 
 from disco.voice.client import VoiceState
-from disco.voice.opus import OpusEncoder, BufferedOpusEncoder
-
-try:
-    from cStringIO import cStringIO as StringIO
-except:
-    from StringIO import StringIO
-
-
-class FFmpegPlayable(object):
-    def __init__(self, source='-', command='avconv', sampling_rate=48000, channels=2, **kwargs):
-        self.source = source
-        self.command = command
-        self.sampling_rate = sampling_rate
-        self.channels = channels
-        self.kwargs = kwargs
-
-        self._buffer = None
-        self._proc = None
-        self._child = None
-
-    @property
-    def stdout(self):
-        return self.proc.stdout
-
-    def read(self, sz):
-        if self.streaming:
-            return self.proc.stdout.read(sz)
-        else:
-            if not self._buffer:
-                data, _ = self.proc.communicate()
-                self._buffer = StringIO(data)
-            return self._buffer.read(sz)
-
-    def pipe(self, other, streaming=True):
-        if issubclass(other, OpusEncoder):
-            self._child = other(self, self.sampling_rate, self.channels, **self.kwargs)
-        else:
-            raise TypeError('Invalid pipe target')
-
-    @property
-    def samples_per_frame(self):
-        return self._child.samples_per_frame
-
-    @property
-    def proc(self):
-        if not self._proc:
-            args = [
-                self.command,
-                '-i', self.source,
-                '-f', 's16le',
-                '-ar', str(self.sampling_rate),
-                '-ac', str(self.channels),
-                '-loglevel', 'warning',
-                'pipe:1'
-            ]
-            self._proc = subprocess.Popen(args, stdin=None, stdout=subprocess.PIPE)
-        return self._proc
-
-    def have_frame(self):
-        return self._child and self._child.have_frame()
-
-    def next_frame(self):
-        return self._child.next_frame()
-
-
-def create_ffmpeg_playable(*args, **kwargs):
-    cls = kwargs.pop('cls', BufferedOpusEncoder)
-    playable = FFmpegPlayable(*args, **kwargs)
-    playable.pipe(cls)
-    return playable
-
-
-def create_youtube_dl_playables(url, *args, **kwargs):
-    import youtube_dl
-
-    ydl = youtube_dl.YoutubeDL({'format': 'webm[abr>0]/bestaudio/best'})
-    info = ydl.extract_info(url, download=False)
-    entries = [info] if 'entries' not in info else info['entries']
-
-    for entry in entries:
-        playable = create_ffmpeg_playable(entry['url'], *args, **kwargs)
-        playable.info = entry
-        yield playable
-
-
-class OpusPlayable(object):
-    """
-    Represents a Playable item which is a cached set of Opus-encoded bytes.
-    """
-    def __init__(self, sampling_rate=48000, frame_length=20, channels=2):
-        self.frames = []
-        self.idx = 0
-        self.frame_length = 20
-        self.sampling_rate = sampling_rate
-        self.frame_length = frame_length
-        self.channels = channels
-        self.sample_size = int(self.sampling_rate / 1000 * self.frame_length)
-
-    @classmethod
-    def from_raw_file(cls, path):
-        inst = cls()
-        obj = open(path, 'r')
-
-        while True:
-            buff = obj.read(2)
-            if not buff:
-                return inst
-            size = struct.unpack('<h', buff)[0]
-            inst.frames.append(obj.read(size))
-
-    def have_frame(self):
-        return self.idx + 1 < len(self.frames)
-
-    def next_frame(self):
-        self.idx += 1
-        return self.frames[self.idx]
 
 
 class Player(object):