BUG: mic_vad_streaming partially recognizing voice

[burgil]

BUG: mic_vad_streaming partially recognizing voice

[burgil]

Forgot to say thanks for the helpers!! I will highly appreciate it!

[burgil]

[burgil]

Anaconda Environment:

[burgil]

In case this would help,
I recorded myself twice using the microphone quality setting set to 44100hz, I recorded one at 44.1khz and the other on 16khz since I recently found other speech services to work better when I do that, here are the recordings:
(Searched google for upload files forever and uploaded it as 7z since I had trouble putting it here): https://usaupload.com/5Kvd/test-voices.7z

I recorded my voice using the following script I had:

import pyaudio, wave, time, sys, os
from datetime import datetime

CHUNK = 8192
FORMAT = pyaudio.paInt16
CHANNELS = 1
RATE = 44100
RECORD_SECONDS = 4

#current_time = str(datetime.now())  #"Date/Time for File Name"
#current_time = "_".join(current_time.split()).replace(":","-")
#current_time = current_time[:-7]
#file_name = 'Audio_'+current_time+'.wav'
WAVE_OUTPUT_FILENAME = 'test.wav'
try:
  os.remove(WAVE_OUTPUT_FILENAME)
except:
  print("Nothing to clean")
p = pyaudio.PyAudio()

stream = p.open(format=FORMAT, channels = CHANNELS, rate = RATE, input = True, input_device_index = 0, frames_per_buffer = CHUNK)

print("* recording")

frames = []
for i in range(0, int(RATE / CHUNK * RECORD_SECONDS)):
    print(i)
    data = stream.read(CHUNK)
    frames.append(data)

print("* done recording")

stream.stop_stream()
stream.close()
p.terminate()

wf = wave.open(WAVE_OUTPUT_FILENAME, 'wb')
wf.setnchannels(CHANNELS)
wf.setsampwidth(p.get_sample_size(FORMAT))
wf.setframerate(RATE)
wf.writeframes(b''.join(frames))
wf.close()

record.cmd:

@echo off
call C:\Users\%username%\anaconda3\Scripts\activate.bat "C:\Users\%username%\anaconda3" & python record.py
pause

simply swapped RATE to 16000 and recorded again since these recordings always worked better than 44100 so I tested both just to be sure..

would be worth mentioning that I never actually tested any .wav files against DeepSpeech, just my microphone, which works as demonstrated above, what I meant is that I tested those wav files against "competitors" and it worked there, also my microphone worked there, just for reference.

Screenshots: