livemic.py

"""
Okay so the plan is to record from the mic
and feed that into whisper
in 5 second chunks since processing time is under a second
larger chunks would have better quality and less missed words
but you sacrifice the 'real-time' feel
"""

# import required libraries
import sounddevice as sd
import wavio as wv
import whisper
import multiprocessing



def record(conn):
    # Sampling frequency
    freq = 44100
    # Recording duration
    duration = 5
    print('Recording')
    
    while True:
        newmsg = False
        conn.send(newmsg)
        # Start recorder with the given values
        # of duration and sample frequency
        recording = sd.rec(int(duration * freq),
                        samplerate=freq, channels=2)

        # Record audio for the given number of seconds
        sd.wait()
        # i wonder if i can bounce between two files so i don't lose words

        # Convert the NumPy array to audio file
        wv.write("recording0.wav", recording, freq, sampwidth=2)
        newmsg = True
        conn.send(newmsg)

def transcribe(conn, model):

    while True:
        while True:
            newmsg = conn.recv()
            if newmsg:
                break
        
        audio = whisper.load_audio("recording0.wav")
        audio = whisper.pad_or_trim(audio)
        
        mel = whisper.log_mel_spectrogram(audio).to(model.device)
        options = whisper.DecodingOptions(language= 'en', fp16=False)
        result = whisper.decode(model, mel, options)

        print(result.text)


if __name__=="__main__":
    model = whisper.load_model("small.en")

    to_mic, to_whisper = multiprocessing.Pipe()
    mic = multiprocessing.Process(target=record, args = (to_whisper,))
    write = multiprocessing.Process(target=transcribe, args = (to_mic, model))
    mic.start()
    write.start()
    mic.join()
    write.join()