diff --git a/human/audio_handler.py b/human/audio_handler.py
index 6d5648b..94a7b41 100644
--- a/human/audio_handler.py
+++ b/human/audio_handler.py
@@ -3,6 +3,11 @@ from abc import ABC, abstractmethod
 
 
 class AudioHandler(ABC):
+    def __init__(self, context, handler):
+        self._context = context
+        self._handler = handler
+
     @abstractmethod
     def on_handle(self,  stream, index):
-        pass
+        if self._handler is not None:
+            self._handler.on_handle(stream, index)
diff --git a/human/audio_mal_handler.py b/human/audio_mal_handler.py
new file mode 100644
index 0000000..bb96c2f
--- /dev/null
+++ b/human/audio_mal_handler.py
@@ -0,0 +1,93 @@
+#encoding = utf8
+import logging
+import queue
+import time
+from queue import Queue
+from threading import Thread, Event
+
+import numpy as np
+
+from human import AudioHandler
+
+logger = logging.getLogger(__name__)
+
+
+class AudioMalHandler(AudioHandler):
+    def __init__(self, context, handler):
+        super().__init__(context, handler)
+
+        self._queue = Queue()
+        self._exit_event = Event()
+        self._thread = Thread(target=self._on_run)
+        self._exit_event.set()
+        self._thread.start()
+
+        self.frames = []
+        self.chunk = context.sample_rate() // context.fps()
+
+    def on_handle(self, stream, index):
+        if self._handler is not None:
+            self._handler.on_handle(stream, index)
+
+    def _on_run(self):
+        logging.info('chunk2mal run')
+        while self._exit_event.is_set():
+            self._run_step()
+            time.sleep(0.01)
+
+        logging.info('chunk2mal exit')
+
+    def _run_step(self):
+        for _ in range(self._context.batch_size() * 2):
+            frame, _type = self.get_audio_frame()
+            self.frames.append(frame)
+            # put to output
+            # self.output_queue.put((frame, _type))
+            self._human.push_out_put(frame, _type)
+        # context not enough, do not run network.
+        if len(self.frames) <= self.stride_left_size + self.stride_right_size:
+            return
+
+        inputs = np.concatenate(self.frames)  # [N * chunk]
+        mel = audio.melspectrogram(inputs)
+        # print(mel.shape[0],mel.shape,len(mel[0]),len(self.frames))
+        # cut off stride
+        left = max(0, self.stride_left_size * 80 / 50)
+        right = min(len(mel[0]), len(mel[0]) - self.stride_right_size * 80 / 50)
+        mel_idx_multiplier = 80. * 2 / self.fps
+        mel_step_size = 16
+        i = 0
+        mel_chunks = []
+        while i < (len(self.frames) - self.stride_left_size - self.stride_right_size) / 2:
+            start_idx = int(left + i * mel_idx_multiplier)
+            # print(start_idx)
+            if start_idx + mel_step_size > len(mel[0]):
+                mel_chunks.append(mel[:, len(mel[0]) - mel_step_size:])
+            else:
+                mel_chunks.append(mel[:, start_idx: start_idx + mel_step_size])
+            i += 1
+        # self.feat_queue.put(mel_chunks)
+        self._human.push_mel_chunks(mel_chunks)
+
+        # discard the old part to save memory
+        self.frames = self.frames[-(self.stride_left_size + self.stride_right_size):]
+
+    def get_audio_frame(self):
+        try:
+            frame = self._queue.get(block=True, timeout=0.01)
+            type_ = 0
+        except queue.Empty:
+            frame = np.zeros(self.chunk, dtype=np.float32)
+            type_ = 1
+
+        return frame, type_
+
+    def stop(self):
+        logging.info('stop')
+        if self._exit_event is None:
+            return
+
+        self._exit_event.clear()
+        if self._thread.is_alive():
+            self._thread.join()
+        logging.info('chunk2mal stop')
diff --git a/tts/Chunk2Mal.py b/tts/Chunk2Mal.py
index f284c38..7b0536a 100644
--- a/tts/Chunk2Mal.py
+++ b/tts/Chunk2Mal.py
@@ -51,8 +51,6 @@ class Chunk2Mal:
         for _ in range(self.batch_size * 2):
             frame, _type = self.get_audio_frame()
             self.frames.append(frame)
-            # put to output
-            # self.output_queue.put((frame, _type))
             self._human.push_out_put(frame, _type)
         # context not enough, do not run network.
         if len(self.frames) <= self.stride_left_size + self.stride_right_size: