I'm streaming microphone input from my laptop computer using Python. I'm currently using PyAudio and .wav to create a 2 second batches (code below) and then read out the frame representations of the newly created .wav file in a loop.
However I really just want the np.ndarray
represented by "signal" in the code that is the Int16 representation of the .wav file. Is there a way to bypass writing to .wav entirely and make my application appear to be "real-time" instead of micro-batch?
import pyaudio
import wave
#AUDIO INPUT
FORMAT = pyaudio.paInt16
CHANNELS = 1
RATE = 44100
CHUNK = 1024
RECORD_SECONDS = 2
WAVE_OUTPUT_FILENAME = "output.wav"
audio = pyaudio.PyAudio()
# start Recording
stream = audio.open(format=FORMAT, channels=CHANNELS,
rate=RATE, input=True,
frames_per_buffer=CHUNK)
while(1):
print "recording"
frames = []
for i in range(0, int(RATE / CHUNK * RECORD_SECONDS)):
data = stream.read(CHUNK)
frames.append(data)
waveFile = wave.open(WAVE_OUTPUT_FILENAME, 'wb')
waveFile.setnchannels(CHANNELS)
waveFile.setsampwidth(audio.get_sample_size(FORMAT))
waveFile.setframerate(RATE)
waveFile.writeframes(b''.join(frames))
waveFile.close()
spf = wave.open(WAVE_OUTPUT_FILENAME,'r')
#Extract Raw Audio from Wav File
signal = spf.readframes(-1)
signal = np.fromstring(signal, 'Int16')
copy= signal.copy()
# stop Recording stream.stop_stream() stream.close() audio.terminate()