forked from scheb/sound-to-light-osc
-
Notifications
You must be signed in to change notification settings - Fork 0
/
recorder.py
74 lines (63 loc) · 2.79 KB
/
recorder.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
import matplotlib
matplotlib.use('TkAgg') # <-- THIS MAKES IT FAST!
import numpy
import pyaudio
import threading
class InputRecorder:
"""Simple, cross-platform class to record from the default input device."""
def __init__(self):
self.RATE = 44100
self.BUFFERSIZE = 2**10
self.secToRecord = .1
self.kill_threads = False
self.has_new_audio = False
self.setup()
def setup(self):
self.buffers_to_record = int(self.RATE * self.secToRecord / self.BUFFERSIZE)
if self.buffers_to_record == 0:
self.buffers_to_record = 1
self.samples_to_record = int(self.BUFFERSIZE * self.buffers_to_record)
self.chunks_to_record = int(self.samples_to_record / self.BUFFERSIZE)
self.sec_per_point = 1. / self.RATE
self.p = pyaudio.PyAudio()
# make sure the default input device is broadcasting the speaker output
# there are a few ways to do this
# e.g., stereo mix, VB audio cable for windows, soundflower for mac
print("Using default input device: {:s}".format(self.p.get_default_input_device_info()['name']))
self.in_stream = self.p.open(format=pyaudio.paInt16,
channels=1,
rate=self.RATE,
input=True,
frames_per_buffer=self.BUFFERSIZE)
self.audio = numpy.empty((self.chunks_to_record * self.BUFFERSIZE), dtype=numpy.int16)
def close(self):
self.kill_threads = True
self.p.close(self.in_stream)
### RECORDING AUDIO ###
def get_audio(self):
"""get a single buffer size worth of audio."""
audio_string = self.in_stream.read(self.BUFFERSIZE)
return numpy.fromstring(audio_string, dtype=numpy.int16)
def record(self):
while not self.kill_threads:
for i in range(self.chunks_to_record):
self.audio[i*self.BUFFERSIZE:(i+1)*self.BUFFERSIZE] = self.get_audio()
self.has_new_audio = True
def start(self):
self.t = threading.Thread(target=self.record)
self.t.start()
def fft(self, data=None, trim_by=2, log_scale=False, div_by=100):
if not data:
data = self.audio.flatten()
left, right = numpy.split(numpy.abs(numpy.fft.fft(data)), 2)
ys = numpy.add(left, right[::-1])
if log_scale:
ys = numpy.multiply(20, numpy.log10(ys))
xs = numpy.arange(self.BUFFERSIZE/2, dtype=float)
if trim_by:
i = int((self.BUFFERSIZE/2) / trim_by)
ys = ys[:i]
xs = xs[:i] * self.RATE / self.BUFFERSIZE
if div_by:
ys = ys / float(div_by)
return xs, ys