(feat) Make audio playback cross platform

Author: dfgHiatus

BabbleApp/babble_processor.py

@@ -11,7 +11,7 @@
 import cv2
 from enum import Enum
 from one_euro_filter import OneEuroFilter
-from utils.misc_utils import playSound, SND_FILENAME, SND_ASYNC, onnx_providers
+from utils.misc_utils import playSound, onnx_providers
 import importlib
 from osc import Tab
 from osc_calibrate_filter import *
@@ -37,7 +37,7 @@ def wrapper(*args, **kwargs):
 async def delayed_setting_change(setting, value):
     await asyncio.sleep(5)
     setting = value
-    playSound("Audio/completed.wav", SND_FILENAME | SND_ASYNC)
+    playSound(os.path.join("Audio", "completed.wav"))
 
 
 class BabbleProcessor:

BabbleApp/camera_widget.py

@@ -3,14 +3,13 @@
 from threading import Event, Thread
 import FreeSimpleGUI as sg
 import cv2
+import os
 from babble_processor import BabbleProcessor, CamInfoOrigin
 from camera import Camera, CameraState, MAX_RESOLUTION
 from config import BabbleConfig
 from osc import Tab
 from utils.misc_utils import (
     playSound,
-    SND_FILENAME,
-    SND_ASYNC,
     list_camera_names,
     get_camera_index_by_name,
     bg_color_highlight,
@@ -19,7 +18,6 @@
 )
 from lang_manager import LocaleStringManager as lang
 
-
 class CameraWidget:
     def __init__(self, widget_id: Tab, main_config: BabbleConfig, osc_queue: Queue):
         self.gui_camera_addr = f"-CAMERAADDR{widget_id}-"
@@ -441,7 +439,7 @@ def render(self, window, event, values):
                 values[self.use_calibration] == True
             ):  # Don't start recording if the calibration filter is disabled.
                 self.babble_cnn.calibration_frame_counter = 1500
-                playSound("Audio/start.wav", SND_FILENAME | SND_ASYNC)
+                playSound(os.path.join("Audio", "start.wav"))
 
         if event == self.gui_stop_calibration:
             if (

BabbleApp/landmark_processor.py

@@ -11,7 +11,7 @@
 import cv2
 from enum import Enum
 from one_euro_filter import OneEuroFilter
-from utils.misc_utils import playSound, SND_FILENAME, SND_ASYNC, onnx_providers
+from utils.misc_utils import playSound, onnx_providers
 from osc_calibrate_filter import *
 from tab import CamInfo, CamInfoOrigin
 from landmark_model_loader import *
@@ -28,13 +28,10 @@ def wrapper(*args, **kwargs):
     wrapper.has_run = False
     return wrapper
 
-
 async def delayed_setting_change(setting, value):
     await asyncio.sleep(5)
     setting = value
-    playSound('Audio/completed.wav', SND_FILENAME | SND_ASYNC)
-
-
+    playSound(os.path.join("Audio", "completed.wav"))
 
 class LandmarkProcessor:
     def __init__(

BabbleApp/osc.py

@@ -1,25 +1,23 @@
 from pythonosc import udp_client, osc_server, dispatcher
-from utils.misc_utils import playSound, SND_FILENAME, SND_ASYNC
+from utils.misc_utils import playSound
 import queue
 import threading
 from enum import IntEnum
 import time
 from config import BabbleConfig
 import traceback
 import math
+import os
 from lang_manager import LocaleStringManager as lang
 
-
 class Tab(IntEnum):
     CAM = 0
     SETTINGS = 1
     ALGOSETTINGS = 2
     CALIBRATION = 3
 
-
 import numpy as np
 
-
 def delay_output_osc(array, delay_seconds, self):
     time.sleep(delay_seconds)
     output_osc(array, self)
@@ -152,7 +150,7 @@ def recalibrate_mouth(self, address, osc_value):
         if osc_value:
             for cam in self.cams:
                 cam.babble_cnn.calibration_frame_counter = 300
-                playSound("Audio/start.wav", SND_FILENAME | SND_ASYNC)
+                playSound(os.path.join("Audio", "start.wav"))
 
     def run(self):
 

BabbleApp/osc_calibrate_filter.py

@@ -1,17 +1,15 @@
 import numpy as np
 from math import log
 from enum import IntEnum
-from utils.misc_utils import playSound, SND_FILENAME, SND_ASYNC
+from utils.misc_utils import playSound
+import os
 from lang_manager import LocaleStringManager as lang
 
-
 class CamId(IntEnum):
     CAM = 0
     SETTINGS = 1
 
-
 class cal:
-
     def __init__(self):
         self.calibration_frame_counter = None
         self.calibrated_array = None
@@ -57,7 +55,7 @@ def cal_osc(self, array):
                     f'[{lang._instance.get_string("log.info")}] Calibration completed.'
                 )
 
-                playSound("Audio/completed.wav", SND_FILENAME | SND_ASYNC)
+                playSound(os.path.join("Audio", "completed.wav"))
 
             if self.settings.calibration_mode == "Neutral":
                 self.min_max_array = np.fromstring(

BabbleApp/requirements.txt

@@ -15,4 +15,6 @@ comtypes==1.4.8;
 pygrabber==0.2;
 psutil==7.0.0;
 requests==2.32.3;
-v4l2py==3.0.0;
+v4l2py==3.0.0;
+sounddevice=0.5.1;
+soundfile-0.13.1;

BabbleApp/utils/misc_utils.py

@@ -7,6 +7,8 @@
 import cv2
 import re
 import subprocess
+import sounddevice as sd
+import soundfile as sf
 
 bg_color_highlight = "#424042"
 bg_color_clear = "#242224"
@@ -169,23 +171,15 @@ def get_camera_index_by_name(name):
 
     return None
 
-
-# Placeholder for sound functions on Windows
-def playSound(*args, **kwargs):
-    pass
-
+# Set environment variable before importing sounddevice. Value is not important.
+os.environ["SD_ENABLE_ASIO"] = "1"
+def playSound(file):
+    data, fs = sf.read(file)
+    sd.play(data, fs)
+    sd.wait()
 
 # Handle debugging virtual envs.
 def ensurePath():
     if os.path.exists(os.path.join(os.getcwd(), "BabbleApp")):
         os.chdir(os.path.join(os.getcwd(), "BabbleApp"))
 
-
-SND_FILENAME = SND_ASYNC = 1
-
-if os_type == 'Windows':
-    import winsound
-
-    playSound = winsound.PlaySound
-    SND_FILENAME = winsound.SND_FILENAME
-    SND_ASYNC = winsound.SND_ASYNC