How to make Speechrecognition listen to system sounds (youtube, zoom, etc.) instead of the microphone?

Question

import pyaudio
import wave
import numpy as np
import pyautogui
import speech_recognition as sr


CHUNK = 1024
FORMAT = pyaudio.paInt16
RATE = 44100



p = pyaudio.PyAudio()


#stream usando o as_loopback para pegar som do SO
   
stream = p.open(
    format = FORMAT,
    channels = 2,
    rate = RATE,
    input=True,
    frames_per_buffer=CHUNK,
    input_device_index=16,)

#Função para ouvir e reconhecer a fala
def ouvir_microfone():
    #Habilita o microfone do usuário
    microfone = sr.Recognizer()
    
    
    
    #usando o microfone
    with sr.Microphone() as source:
        
        #Chama um algoritmo de reducao de ruidos no som
        microfone.adjust_for_ambient_noise(source)
        
        
        #Frase para o usuario dizer algo
        print("Diga alguma coisa: ")
        
        #Armazena o que foi dito numa variavel
        audio = microfone.listen(source)
        
    try:
        
        #Passa a variável para o algoritmo reconhecedor de padroes
        frase = microfone.recognize_google(audio,language='pt-BR')

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

        if "aula" in frase:
            pyautogui.PAUSE = 1
            pyautogui.keyDown('win')  # hold down the shift key
            pyautogui.press('1')     # press the left arrow key
            pyautogui.press('1')     # press the left arrow key
            pyautogui.keyUp('win')
            pyautogui.click(x=1311, y=988)
            pyautogui.write("Presente professor")
            pyautogui.press('enter')
            pyautogui.hotkey('alt', 'tab') 
            

            

        
        #Retorna a frase pronunciada
        print("Você disse: " + frase)
        
    #Se nao reconheceu o padrao de fala, exibe a mensagem
    except sr.UnkownValueError:
        print("Não entendi")
        
    return frase
    

ouvir_microfone()

Answer 1

Hello! Try changing the variable value input_device_index.

When you create the variable stream, you are passing several parameters to Pyaudio, this variable is responsible for selecting from which device the package will capture the audio:

stream = p.open(
    format = FORMAT,
    channels = 11,
    rate = RATE,
    input=True,
    frames_per_buffer=CHUNK,
    input_device_index=#ALTERAR AQUI,
    as_loopback=True)

PS: Maybe you have to search for these 'Devices' with:

for i in range(0, p.get_device_count()):
    print(i, p.get_device_info_by_index(i)['name'])

This way you can see which list of devices Pyaudio can capture, and choose which one you want. Let me know if you can.

If you still have questions, take a look in that reply, in this case it uses two variables stream, and simultaneously captures two devices (channels).