Fix my error. The headphones should work. Do not suggest using VB-Audio. Modify the…
- User
Fix my error. The headphones should work. Do not suggest using VB-Audio. Modify the code.
--- error below
✅ Format supported: Microphone (Razer Seiren Mini) (2 ch, 44100 Hz) 🎙️ Recording from: Microphone (Razer Seiren Mini) (Index: 1, Channels: 2, Sample Rate: 44100 Hz) ⚠️ Note: Bluetooth devices like WH-1000XM4 may not support loopback. Consider using VB-Audio Virtual Cable (https://vb-audio.com/Cable/). ⚠️ Format not supported: Headphones (2- WH-1000XM4) (2 ch, 44100 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (1 ch, 44100 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (2 ch, 48000 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (1 ch, 48000 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (2 ch, 44100 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (1 ch, 44100 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (2 ch, 32000 Hz): ('Invalid number of channels', -9998) ⚠️ Format not supported: Headphones (2- WH-1000XM4) (1 ch, 32000 Hz): ('Invalid number of channels', -9998) ❌ Speaker format not supported: Headphones (2- WH-1000XM4) (0 ch, 44100 Hz) 🔈 Disabling speaker recording 🔄 Starting capture loop... 🔄 Loop iteration 1, recording=True
--- code begins
import threading import time import os import wave import subprocess import sys import shutil
import numpy as np import pyaudiowpatch as pyaudio from pystray import Icon, MenuItem, Menu from PIL import Image, ImageDraw
--- Globals ---
recording = False mic_chunks = [] sr_mic = None chunk_mic = 8192 # Increased for Bluetooth stability tray_icon = None start_time = None selected_mic = None # Selected microphone device info selected_spk = None # Selected speaker (loopback) device info p = None # PyAudio instance for global access
--- Get available devices ---
def get_available_devices(): global p, selected_mic, selected_spk if p is None: p = pyaudio.PyAudio() input_devices = [] output_devices = [] for i in range(p.get_device_count()): dev = p.get_device_info_by_index(i) dev_info = { 'name': dev['name'], 'index': dev['index'], 'input_channels': min(dev['maxInputChannels'], 2) if dev['maxInputChannels'] > 0 else 0, 'output_channels': min(dev['maxOutputChannels'], 2) if dev['maxOutputChannels'] > 0 else 0, 'sample_rate': int(dev['defaultSampleRate']) } if dev['maxInputChannels'] > 0: input_devices.append(dev_info) if 'Razer Seiren Mini' in dev['name'] and not selected_mic: selected_mic = dev_info # Default to Razer Seiren Mini if dev['maxOutputChannels'] > 0: output_devices.append(dev_info) # Validate selected_spk against current devices if selected_spk: if not any(dev['index'] == selected_spk['index'] for dev in output_devices): print(f"⚠️ Previously selected output {selected_spk['name']} no longer available. Resetting to None.") selected_spk = None return input_devices, output_devices
--- Check if device is selected ---
def is_input_selected(device): return selected_mic and selected_mic['index'] == device['index']
def is_output_selected(device): return selected_spk and selected_spk['index'] == device['index'] if device else selected_spk is None
--- Select device ---
def select_input_device(device): def inner(icon, item): global selected_mic selected_mic = device print(f"🎙️ Selected input: {selected_mic['name']} (Index: {selected_mic['index']})") icon.title = f"Tray Recorder: {selected_mic['name']}" icon.menu = build_menu() # Rebuild menu with persisted selections return inner
def select_output_device(device): def inner(icon, item): global selected_spk old_spk = selected_spk['name'] if selected_spk else "None" selected_spk = device new_spk = selected_spk['name'] if selected_spk else "None" print(f"🔈 Output changed from {old_spk} to {new_spk} (Index: {selected_spk['index'] if selected_spk else 'N/A'})") icon.title = f"Tray Recorder: {selected_mic['name'] if selected_mic else 'Idle'}" icon.menu = build_menu() # Rebuild menu with persisted selections return inner
--- Build dynamic menu ---
def build_menu(): input_devices, output_devices = get_available_devices() input_menu = Menu( *[MenuItem(f"{dev['name']} ({dev['input_channels']} ch, {dev['sample_rate']} Hz)", select_input_device(dev), checked=lambda item, dev=dev: is_input_selected(dev)) for dev in input_devices] ) output_menu = Menu( MenuItem("None", select_output_device(None), checked=lambda item: is_output_selected(None)), *[MenuItem(f"{dev['name']} ({dev['output_channels']} ch, {dev['sample_rate']} Hz)", select_output_device(dev), checked=lambda item, dev=dev: is_output_selected(dev)) for dev in output_devices] ) return Menu( MenuItem("Select Input Device", input_menu), MenuItem("Select Output Device", output_menu), MenuItem("Start Recording", start_record), MenuItem("Stop Recording", stop_record), MenuItem("Quit", quit_app) )
--- Test device format ---
def test_device_format(device, sample_rate, channels, is_input=True): sample_rates = [sample_rate, 48000, 44100, 32000] # Try default, then common rates for sr in sample_rates: for ch in [channels, 1]: # Try specified channels, then 1 try: p.is_format_supported( sr, input_device=device['index'] if is_input else None, output_device=device['index'] if not is_input else None, input_channels=ch if is_input else None, output_channels=ch if not is_input else None, input_format=pyaudio.paInt16 if is_input else None, output_format=pyaudio.paInt16 if not is_input else None ) print(f"✅ Format supported: {device['name']} ({ch} ch, {sr} Hz)") return True, ch, sr except ValueError as e: print(f"⚠️ Format not supported: {device['name']} ({ch} ch, {sr} Hz): {e}") return False, 0, sample_rate
--- Audio capture worker ---
def record_worker(): global recording, mic_chunks, sr_mic, chunk_mic, selected_mic, selected_spk, p
if p is None: p = pyaudio.PyAudio() # Log all devices print("🔍 Available devices:") for i in range(p.get_device_count()): dev = p.get_device_info_by_index(i) print(f"Device {i}: {dev['name']}, in_ch={dev['maxInputChannels']}, " f"out_ch={dev['maxOutputChannels']}, sr={dev['defaultSampleRate']}") # Validate selected devices if not selected_mic: print("❌ No microphone selected") recording = False p.terminate() return # Microphone setup sr_mic = selected_mic['sample_rate'] mic_idx = selected_mic['index'] mic_ch = selected_mic['input_channels'] mic_supported, mic_ch, sr_mic = test_device_format(selected_mic, sr_mic, mic_ch, is_input=True) if not mic_supported: print(f"❌ Microphone format not supported: {selected_mic['name']} ({mic_ch} ch, {sr_mic} Hz)") recording = False p.terminate() return print(f"🎙️ Recording from: {selected_mic['name']} " f"(Index: {mic_idx}, Channels: {mic_ch}, Sample Rate: {sr_mic} Hz)") mic_stream = None try: mic_stream = p.open( format=pyaudio.paInt16, channels=mic_ch, rate=sr_mic, input=True, input_device_index=mic_idx, frames_per_buffer=chunk_mic ) except Exception as e: print(f"❌ Cannot open microphone {selected_mic['name']}: {e}") recording = False p.terminate() return # Speaker loopback setup (if selected) spk_stream = None spk_chunks = [] sr_spk = None spk_ch = None if selected_spk: print("⚠️ Note: Bluetooth devices like WH-1000XM4 may not support loopback. Consider using VB-Audio Virtual Cable (https://vb-audio.com/Cable/).") sr_spk = selected_spk['sample_rate'] spk_idx = selected_spk['index'] spk_ch = selected_spk['output_channels'] spk_supported, spk_ch, sr_spk = test_device_format(selected_spk, sr_spk, spk_ch, is_input=True) if not spk_supported: print(f"❌ Speaker format not supported: {selected_spk['name']} ({spk_ch} ch, {sr_spk} Hz)") print("🔈 Disabling speaker recording") selected_spk = None else: print(f"🔈 Recording from: {selected_spk['name']} " f"(Index: {spk_idx}, Channels: {spk_ch}, Sample Rate: {sr_spk} Hz)") for attempt in range(5): # Retry up to 5 times try: spk_stream = p.open( format=pyaudio.paInt16, channels=spk_ch, rate=sr_spk, input=True, input_device_index=spk_idx, frames_per_buffer=chunk_mic ) print(f"✅ Speaker stream opened on attempt {attempt + 1}") break except Exception as e: print(f"⚠️ Attempt {attempt + 1} failed to open speaker {selected_spk['name']}: {e}") time.sleep(1.0) # Longer delay for Bluetooth if attempt == 4: print("❌ Cannot open speaker after 5 attempts. Disabling speaker recording.") selected_spk = None break mic_chunks.clear() spk_chunks.clear() # Capture loop loop_count = 0 timeout = time.time() + 30 print("🔄 Starting capture loop...") while recording and time.time() < timeout: loop_count += 1 print(f"🔄 Loop iteration {loop_count}, recording={recording}") try: # Check microphone device try: dev_info = p.get_device_info_by_index(mic_idx) if dev_info['maxInputChannels'] <= 0: print(f"⚠️ Microphone {selected_mic['name']} unavailable") recording = False break except Exception as e: print(f"⚠️ Error checking microphone: {e}") recording = False break # Microphone capture if mic_stream and mic_stream.is_active(): for _ in range(5): try: start_read = time.time() data_mic = mic_stream.read(chunk_mic, exception_on_overflow=False) read_time = time.time() - start_read print(f"🎙️ Read {len(data_mic)} bytes from mic in {read_time:.3f}s") if data_mic: arr_mic = np.frombuffer(data_mic, dtype=np.int16).reshape(-1, mic_ch) mono_mic = arr_mic.mean(axis=1).astype(np.int16).tobytes() mic_chunks.append(mono_mic) mic_max = np.max(np.abs(arr_mic)) if arr_mic.size > 0 else 0 print(f"🎙️ Mic chunk {len(mic_chunks)}, max amplitude: {mic_max}") break else: print(f"⚠️ Empty data from {selected_mic['name']}") time.sleep(0.2) except Exception as e: print(f"⚠️ Mic read error: {e}") if "Stream closed" in str(e): break time.sleep(0.2) else: print(f"⚠️ Mic read failed after retries") try: mic_stream.close() mic_stream = p.open( format=pyaudio.paInt16, channels=mic_ch, rate=sr_mic, input=True, input_device_index=mic_idx, frames_per_buffer=chunk_mic ) print(f"🎙️ Mic stream reinitialized") except Exception as e: print(f"❌ Failed to reinitialize mic stream: {e}") recording = False break # Speaker capture (if selected) if selected_spk and spk_stream and spk_stream.is_active(): for _ in range(5): try: start_read = time.time() data_spk = spk_stream.read(chunk_mic, exception_on_overflow=False) read_time = time.time() - start_read print(f"🔈 Read {len(data_spk)} bytes from speaker in {read_time:.3f}s") if data_spk: arr_spk = np.frombuffer(data_spk, dtype=np.int16).reshape(-1, spk_ch) mono_spk = arr_spk.mean(axis=1).astype(np.int16).tobytes() spk_chunks.append(mono_spk) spk_max = np.max(np.abs(arr_spk)) if arr_spk.size > 0 else 0 print(f"🔈 Speaker chunk {len(spk_chunks)}, max amplitude: {spk_max}") break else: print(f"⚠️ Empty data from {selected_spk['name']}") time.sleep(0.2) except Exception as e: print(f"⚠️ Speaker read error: {e}") if "Stream closed" in str(e): break time.sleep(0.2) else: print(f"⚠️ Speaker read failed after retries") try: spk_stream.close() spk_stream = p.open( format=pyaudio.paInt16, channels=spk_ch, rate=sr_spk, input=True, input_device_index=selected_spk['index'], frames_per_buffer=chunk_mic ) print(f"🔈 Speaker stream reinitialized") except Exception as e: print(f"❌ Failed to reinitialize speaker stream: {e}") selected_spk = None break except Exception as e: print(f"⚠️ Error in capture loop: {e}") recording = False break time.sleep(0.01) print(f"🛑 Capture loop exited (reason: {'recording=False' if not recording else 'timeout'}, " f"mic_chunks: {len(mic_chunks)}, spk_chunks: {len(spk_chunks) if selected_spk else 0}, loops: {loop_count})") # Cleanup if mic_stream: try: if mic_stream.is_active(): mic_stream.stop_stream() mic_stream.close() except Exception as e: print(f"⚠️ Error closing mic stream: {e}") if spk_stream: try: if spk_stream.is_active(): spk_stream.stop_stream() spk_stream.close() except Exception as e: print(f"⚠️ Error closing speaker stream: {e}") p.terminate() p = None--- Save & mix via ffmpeg ---
def save_recording(): global mic_chunks, sr_mic, selected_mic, selected_spk
print(f"💾 Saving: {len(mic_chunks)} mic chunks from {selected_mic['name']}") spk_chunks = [] if selected_spk: print(f"💾 Saving: {len(spk_chunks)} speaker chunks from {selected_spk['name']}") if not mic_chunks and not spk_chunks: print("❌ No audio recorded—skipping save.") return mic_silent = True if mic_chunks: mic_data = np.frombuffer(b"".join(mic_chunks), dtype=np.int16) mic_max = np.max(np.abs(mic_data)) if mic_data.size > 0 else 0 mic_silent = mic_max < 10 print(f"🎙️ Mic audio max amplitude: {mic_max}") spk_silent = True if selected_spk and spk_chunks: spk_data = np.frombuffer(b"".join(spk_chunks), dtype=np.int16) spk_max = np.max(np.abs(spk_data)) if spk_data.size > 0 else 0 spk_silent = spk_max < 10 print(f"🔈 Speaker audio max amplitude: {spk_max}") if mic_silent and spk_silent: print("❌ Audio is silent—skipping save.") return ts = int(time.time()) wav_files = [] # Save microphone audio if mic_chunks and not mic_silent: wav_mic = f"recording_mic_{time.strftime('%Y%m%d_%H%M%S')}.wav" with wave.open(wav_mic, "wb") as wf: wf.setnchannels(1) wf.setsampwidth(2) wf.setframerate(sr_mic) wf.writeframes(b"".join(mic_chunks)) wav_files.append(wav_mic) # Save speaker audio if selected_spk and spk_chunks and not spk_silent: wav_spk = f"recording_spk_{time.strftime('%Y%m%d_%H%M%S')}.wav" with wave.open(wav_spk, "wb") as wf: wf.setnchannels(1) wf.setsampwidth(2) wf.setframerate(sr_spk if sr_spk else sr_mic) wf.writeframes(b"".join(spk_chunks)) wav_files.append(wav_spk) if not wav_files: print("❌ No valid audio to save.") return # Check for ffmpeg ffmpeg_path = shutil.which("ffmpeg") if ffmpeg_path is None: # Try local ffmpeg.exe in script directory script_dir = os.path.dirname(os.path.abspath(__file__)) local_ffmpeg = os.path.join(script_dir, "ffmpeg.exe") if os.path.exists(local_ffmpeg): ffmpeg_path = local_ffmpeg else: print("❌ FFmpeg not found. Please install FFmpeg and add it to your system PATH, or place ffmpeg.exe in the script directory (C:\\Users\\Radek\\Desktop\\audio).") print(f"✅ Audio saved as WAV: {wav_files[0]}") return out_mp3 = time.strftime("recording_%Y%m%d_%H%M%S.mp3") # Mix or save single file try: if len(wav_files) == 1: cmd = [ffmpeg_path, "-y", "-i", wav_files[0], "-codec:a", "libmp3lame", "-qscale:a", "2", out_mp3] else: cmd = [ffmpeg_path, "-y"] for wav in wav_files: cmd.extend(["-i", wav]) cmd.extend(["-filter_complex", "amix=inputs=2:duration=longest", "-codec:a", "libmp3lame", "-qscale:a", "2", out_mp3]) subprocess.run(cmd, check=True) print(f"✅ Saved: {out_mp3}") # Clean up WAV files after successful MP3 conversion for wf in wav_files: try: os.remove(wf) except OSError: print(f"⚠️ Could not delete temporary WAV file: {wf}") except (subprocess.CalledProcessError, FileNotFoundError) as e: print(f"❌ FFmpeg failed: {e}") print(f"✅ Audio saved as WAV: {wav_files[0]}")--- Tray & Ctrl+C handling ---
def make_icon_image(is_rec=False): img = Image.new("RGB", (64, 64), "black") d = ImageDraw.Draw(img) d.ellipse((16, 16, 48, 48), fill="red" if is_rec else "grey") return img
def start_record(icon, item): global recording, start_time if recording: return if not selected_mic: print("❌ Please select a microphone first") return recording = True start_time = time.time() print("▶ Recording started.") threading.Thread(target=record_worker, daemon=True).start() icon.icon = make_icon_image(True) icon.title = f"Recording: {selected_mic['name']}" icon.menu = build_menu() # Update menu with persisted selections
def stop_record(icon=None, item=None): global recording, start_time if not recording: return elapsed = time.time() - start_time if start_time else 0 if elapsed < 5: print("⚠️ Stop ignored: recording too short (< 5s)") return print(f"🛑 Stop record called (elapsed: {elapsed:.1f}s)") recording = False if icon: icon.icon = make_icon_image(False) icon.title = "Saving..." time.sleep(1.0) save_recording() if icon: icon.title = f"Tray Recorder: {selected_mic['name'] if selected_mic else 'Idle'}" icon.menu = build_menu() # Update menu with persisted selections
def quit_app(icon, item): stop_record(icon, item) icon.stop()
--- Main ---
if name == "main": # Initialize tray icon tray_icon = Icon( "TrayRecorder", icon=make_icon_image(False), title="Tray Recorder: Idle" )
# Set initial menu tray_icon.menu = build_menu() tray_icon.run_detached() print("Tray recorder running. Use tray menu to select devices and Start/Stop, or Ctrl+C to quit.") try: while True: time.sleep(1) except KeyboardInterrupt: print("\nInterrupted; stopping...") stop_record() tray_icon.stop() sys.exit(0) - Assistant
The error indicates that the Sony WH-1000XM4 headphones are not supporting loopback recording, likely due to limitations in their driver or hardware when used as an input device for loopback. The error messages show repeated failures with "Invalid number of channels" (-9998) across various channel and sample rate configurations, suggesting that the device does not support being used as an input device in this context, even though it is an output device.
To fix this issue without using VB-Audio, we need to modify the code to handle the WH-1000XM4 headphones more robustly, particularly in how we test and configure loopback recording. The key changes will involve:
- Improving Device Detection: Ensure we correctly identify devices that support loopback (using
pyaudiowpatchWASAPI loopback capabilities). - Adjusting Format Testing: Test the device as a loopback device explicitly, not as a standard input device, since loopback requires capturing the output stream.
- Handling Bluetooth Limitations: Recognize that Bluetooth devices like the WH-1000XM4 may not support loopback natively and provide a fallback or disable loopback gracefully.
- Fixing Stream Configuration: Use the correct configuration for loopback streams, ensuring we use the device's output channels and sample rate as the basis for loopback input.
Below is the modified code with these fixes applied. The changes are focused on the
test_device_format,record_worker, and related functions to ensure the WH-1000XM4 headphones work for loopback recording if possible, or are gracefully disabled if not.import threading import time import os import wave import subprocess import sys import shutil import numpy as np import pyaudiowpatch as pyaudio from pystray import Icon, MenuItem, Menu from PIL import Image, ImageDraw # --- Globals --- recording = False mic_chunks = [] sr_mic = None chunk_mic = 8192 # Increased for Bluetooth stability tray_icon = None start_time = None selected_mic = None # Selected microphone device info selected_spk = None # Selected speaker (loopback) device info p = None # PyAudio instance for global access # --- Get available devices --- def get_available_devices(): global p, selected_mic, selected_spk if p is None: p = pyaudio.PyAudio() input_devices = [] output_devices = [] for i in range(p.get_device_count()): dev = p.get_device_info_by_index(i) dev_info = { 'name': dev['name'], 'index': dev['index'], 'input_channels': min(dev['maxInputChannels'], 2) if dev['maxInputChannels'] > 0 else 0, 'output_channels': min(dev['maxOutputChannels'], 2) if dev['maxOutputChannels'] > 0 else 0, 'sample_rate': int(dev['defaultSampleRate']), 'is_loopback': dev.get('isLoopbackDevice', False) # Check if it's a loopback device } if dev['maxInputChannels'] > 0 and not dev_info['is_loopback']: input_devices.append(dev_info) if 'Razer Seiren Mini' in dev['name'] and not selected_mic: selected_mic = dev_info # Default to Razer Seiren Mini if dev['maxOutputChannels'] > 0 or dev_info['is_loopback']: output_devices.append(dev_info) # Validate selected_spk against current devices if selected_spk: if not any(dev['index'] == selected_spk['index'] for dev in output_devices): print(f"⚠️ Previously selected output {selected_spk['name']} no longer available. Resetting to None.") selected_spk = None return input_devices, output_devices # --- Check if device is selected --- def is_input_selected(device): return selected_mic and selected_mic['index'] == device['index'] def is_output_selected(device): return selected_spk and selected_spk['index'] == device['index'] if device else selected_spk is None # --- Select device --- def select_input_device(device): def inner(icon, item): global selected_mic selected_mic = device print(f"🎙️ Selected input: {selected_mic['name']} (Index: {selected_mic['index']})") icon.title = f"Tray Recorder: {selected_mic['name']}" icon.menu = build_menu() # Rebuild menu with persisted selections return inner def select_output_device(device): def inner(icon, item): global selected_spk old_spk = selected_spk['name'] if selected_spk else "None" selected_spk = device new_spk = selected_spk['name'] if selected_spk else "None" print(f"🔈 Output changed from {old_spk} to {new_spk} (Index: {selected_spk['index'] if selected_spk else 'N/A'})") icon.title = f"Tray Recorder: {selected_mic['name'] if selected_mic else 'Idle'}" icon.menu = build_menu() # Rebuild menu with persisted selections return inner # --- Build dynamic menu --- def build_menu(): input_devices, output_devices = get_available_devices() input_menu = Menu( *[MenuItem(f"{dev['name']} ({dev['input_channels']} ch, {dev['sample_rate']} Hz)", select_input_device(dev), checked=lambda item, dev=dev: is_input_selected(dev)) for dev in input_devices] ) output_menu = Menu( MenuItem("None", select_output_device(None), checked=lambda item: is_output_selected(None)), *[MenuItem(f"{dev['name']} ({dev['output_channels']} ch, {dev['sample_rate']} Hz)", select_output_device(dev), checked=lambda item, dev=dev: is_output_selected(dev)) for dev in output_devices] ) return Menu( MenuItem("Select Input Device", input_menu), MenuItem("Select Output Device", output_menu), MenuItem("Start Recording", start_record), MenuItem("Stop Recording", stop_record), MenuItem("Quit", quit_app) ) # --- Test device format --- def test_device_format(device, sample_rate, channels, is_input=True, is_loopback=False): sample_rates = [sample_rate, 48000, 44100, 32000] # Try default, then common rates for sr in sample_rates: for ch in [channels, 2, 1]: # Try specified channels, then 2, then 1 try: if is_loopback: # For loopback, we need to use the output device's format as input p.is_format_supported( sr, input_device=device['index'], input_channels=ch, input_format=pyaudio.paInt16 ) else: p.is_format_supported( sr, input_device=device['index'] if is_input else None, output_device=device['index'] if not is_input else None, input_channels=ch if is_input else None, output_channels=ch if not is_input else None, input_format=pyaudio.paInt16 if is_input else None, output_format=pyaudio.paInt16 if not is_input else None ) print(f"✅ Format supported: {device['name']} ({ch} ch, {sr} Hz)") return True, ch, sr except ValueError as e: print(f"⚠️ Format not supported: {device['name']} ({ch} ch, {sr} Hz): {e}") return False, 0, sample_rate # --- Audio capture worker --- def record_worker(): global recording, mic_chunks, sr_mic, chunk_mic, selected_mic, selected_spk, p if p is None: p = pyaudio.PyAudio() # Log all devices print("🔍 Available devices:") for i in range(p.get_device_count()): dev = p.get_device_info_by_index(i) print(f"Device {i}: {dev['name']}, in_ch={dev['maxInputChannels']}, " f"out_ch={dev['maxOutputChannels']}, sr={dev['defaultSampleRate']}, " f"loopback={dev.get('isLoopbackDevice', False)}") # Validate selected devices if not selected_mic: print("❌ No microphone selected") recording = False p.terminate() return # Microphone setup sr_mic = selected_mic['sample_rate'] mic_idx = selected_mic['index'] mic_ch = selected_mic['input_channels'] mic_supported, mic_ch, sr_mic = test_device_format(selected_mic, sr_mic, mic_ch, is_input=True) if not mic_supported: print(f"❌ Microphone format not supported: {selected_mic['name']} ({mic_ch} ch, {sr_mic} Hz)") recording = False p.terminate() return print(f"🎙️ Recording from: {selected_mic['name']} " f"(Index: {mic_idx}, Channels: {mic_ch}, Sample Rate: {sr_mic} Hz)") mic_stream = None try: mic_stream = p.open( format=pyaudio.paInt16, channels=mic_ch, rate=sr_mic, input=True, input_device_index=mic_idx, frames_per_buffer=chunk_mic ) except Exception as e: print(f"❌ Cannot open microphone {selected_mic['name']}: {e}") recording = False p.terminate() return # Speaker loopback setup (if selected) spk_stream = None spk_chunks = [] sr_spk = None spk_ch = None if selected_spk: sr_spk = selected_spk['sample_rate'] spk_idx = selected_spk['index'] spk_ch = selected_spk['output_channels'] # Use output channels for loopback # Check if this is a WASAPI loopback device dev_info = p.get_device_info_by_index(spk_idx) is_loopback = dev_info.get('isLoopbackDevice', False) if not is_loopback: # Find the corresponding loopback device for i in range(p.get_device_count()): dev = p.get_device_info_by_index(i) if dev.get('isLoopbackDevice', False) and dev['name'] == dev_info['name']: spk_idx = dev['index'] is_loopback = True break if not is_loopback: print(f"⚠️ Device {selected_spk['name']} does not support loopback recording.") selected_spk = None else: spk_supported, spk_ch, sr_spk = test_device_format(selected_spk, sr_spk, spk_ch, is_input=True, is_loopback=True) if not spk_supported: print(f"❌ Speaker format not supported: {selected_spk['name']} ({spk_ch} ch, {sr_spk} Hz)") print("🔈 Disabling speaker recording") selected_spk = None else: print(f"🔈 Recording from loopback: {selected_spk['name']} " f"(Index: {spk_idx}, Channels: {spk_ch}, Sample Rate: {sr_spk} Hz)") for attempt in range(5): # Retry up to 5 times try: spk_stream = p.open( format=pyaudio.paInt16, channels=spk_ch, rate=sr_spk, input=True, input_device_index=spk_idx, frames_per_buffer=chunk_mic, as_loopback=True # Explicitly enable loopback ) print(f"✅ Speaker loopback stream opened on attempt {attempt + 1}") break except Exception as e: print(f"⚠️ Attempt {attempt + 1} failed to open speaker loopback {selected_spk['name']}: {e}") time.sleep(1.0) # Longer delay for Bluetooth if attempt == 4: print("❌ Cannot open speaker loopback after 5 attempts. Disabling speaker recording.") selected_spk = None break mic_chunks.clear() spk_chunks.clear() # Capture loop loop_count = 0 timeout = time.time() + 30 print("🔄 Starting capture loop...") while recording and time.time() < timeout: loop_count += 1 print(f"🔄 Loop iteration {loop_count}, recording={recording}") try: # Check microphone device try: dev_info = p.get_device_info_by_index(mic_idx) if dev_info['maxInputChannels'] <= 0: print(f"⚠️ Microphone {selected_mic['name']} unavailable") recording = False break except Exception as e: print(f"⚠️ Error checking microphone: {e}") recording = False break # Microphone capture if mic_stream and mic_stream.is_active(): for _ in range(5): try: start_read = time.time() data_mic = mic_stream.read(chunk_mic, exception_on_overflow=False) read_time = time.time() - start_read print(f"🎙️ Read {len(data_mic)} bytes from mic in {read_time:.3f}s") if data_mic: arr_mic = np.frombuffer(data_mic, dtype=np.int16).reshape(-1, mic_ch) mono_mic = arr_mic.mean(axis=1).astype(np.int16).tobytes() mic_chunks.append(mono_mic) mic_max = np.max(np.abs(arr_mic)) if arr_mic.size > 0 else 0 print(f"🎙️ Mic chunk {len(mic_chunks)}, max amplitude: {mic_max}") break else: print(f"⚠️ Empty data from {selected_mic['name']}") time.sleep(0.2) except Exception as e: print(f"⚠️ Mic read error: {e}") if "Stream closed" in str(e): break time.sleep(0.2) else: print(f"⚠️ Mic read failed after retries") try: mic_stream.close() mic_stream = p.open( format=pyaudio.paInt16, channels=mic_ch, rate=sr_mic, input=True, input_device_index=mic_idx, frames_per_buffer=chunk_mic ) print(f"🎙️ Mic stream reinitialized") except Exception as e: print(f"❌ Failed to reinitialize mic stream: {e}") recording = False break # Speaker capture (if selected) if selected_spk and spk_stream and spk_stream.is_active(): for _ in range(5): try: start_read = time.time() data_spk = spk_stream.read(chunk_mic, exception_on_overflow=False) read_time = time.time() - start_read print(f"🔈 Read {len(data_spk)} bytes from speaker in {read_time:.3f}s") if data_spk: arr_spk = np.frombuffer(data_spk, dtype=np.int16).reshape(-1, spk_ch) mono_spk = arr_spk.mean(axis=1).astype(np.int16).tobytes() spk_chunks.append(mono_spk) spk_max = np.max(np.abs(arr_spk)) if arr_spk.size > 0 else 0 print(f"🔈 Speaker chunk {len(spk_chunks)}, max amplitude: {spk_max}") break else: print(f"⚠️ Empty data from {selected_spk['name']}") time.sleep(0.2) except Exception as e: print(f"⚠️ Speaker read error: {e}") if "Stream closed" in str(e): break time.sleep(0.2) else: print(f"⚠️ Speaker read failed after retries") try: spk_stream.close() spk_stream = p.open( format=pyaudio.paInt16, channels=spk_ch, rate=sr_spk, input=True, input_device_index=selected_spk['index'], frames_per_buffer=chunk_mic, as_loopback=True # Explicitly enable loopback ) print(f"🔈 Speaker stream reinitialized") except Exception as e: print(f"❌ Failed to reinitialize speaker stream: {e}") selected_spk = None break except Exception as e: print(f"⚠️ Error in capture loop: {e}") recording = False break time.sleep(0.01) print(f"🛑 Capture loop exited (reason: {'recording=False' if not recording else 'timeout'}, " f"mic_chunks: {len(mic_chunks)}, spk_chunks: {len(spk_chunks) if selected_spk else 0}, loops: {loop_count})") # Cleanup if mic_stream: try: if mic_stream.is_active(): mic_stream.stop_stream() mic_stream.close() except Exception as e: print(f"⚠️ Error closing mic stream: {e}") if spk_stream: try: if spk_stream.is_active(): spk_stream.stop_stream() spk_stream.close() except Exception as e: print(f"⚠️ Error closing speaker stream: {e}") p.terminate() p = None # --- Save & mix via ffmpeg --- def save_recording(): global mic_chunks, sr_mic, selected_mic, selected_spk print(f"💾 Saving: {len(mic_chunks)} mic chunks from {selected_mic['name']}") spk_chunks = [] if selected_spk: print(f"💾 Saving: {len(spk_chunks)} speaker chunks from {selected_spk['name']}") if not mic_chunks and not spk_chunks: print("❌ No audio recorded—skipping save.") return mic_silent = True if mic_chunks: mic_data = np.frombuffer(b"".join(mic_chunks), dtype=np.int16) mic_max = np.max(np.abs(mic_data)) if mic_data.size > 0 else 0 mic_silent = mic_max < 10 print(f"🎙️ Mic audio max amplitude: {mic_max}") spk_silent = True if selected_spk and spk_chunks: spk_data = np.frombuffer(b"".join(spk_chunks), dtype=np.int16) spk_max = np.max(np.abs(spk_data)) if spk_data.size > 0 else 0 spk_silent = spk_max < 10 print(f"🔈 Speaker audio max amplitude: {spk_max}") if mic_silent and spk_silent: print("❌ Audio is silent—skipping save.") return ts = int(time.time()) wav_files = [] # Save microphone audio if mic_chunks and not mic_silent: wav_mic = f"recording_mic_{time.strftime('%Y%m%d_%H%M%S')}.wav" with wave.open(wav_mic, "wb") as wf: wf.setnchannels(1) wf.setsampwidth(2) wf.setframerate(sr_mic) wf.writeframes(b"".join(mic_chunks)) wav_files.append(wav_mic) # Save speaker audio if selected_spk and spk_chunks and not spk_silent: wav_spk = f"recording_spk_{time.strftime('%Y%m%d_%H%M%S')}.wav" with wave.open(wav_spk, "wb") as wf: wf.setnchannels(1) wf.setsampwidth(2) wf.setframerate(sr_spk if sr_spk else sr_mic) wf.writeframes(b"".join(spk_chunks)) wav_files.append(wav_spk) if not wav_files: print("❌ No valid audio to save.") return # Check for ffmpeg ffmpeg_path = shutil.which("ffmpeg") if ffmpeg_path is None: # Try local ffmpeg.exe in script directory script_dir = os.path.dirname(os.path.abspath(__file__)) local_ffmpeg = os.path.join(script_dir, "ffmpeg.exe") if os.path.exists(local_ffmpeg): ffmpeg_path = local_ffmpeg else: print("❌ FFmpeg not found. Please install FFmpeg and add it to your system PATH, or place ffmpeg.exe in the script directory (C:\\Users\\Radek\\Desktop\\audio).") print(f"✅ Audio saved as WAV: {wav_files[0]}") return out_mp3 = time.strftime("recording_%Y%m%d_%H%M%S.mp3") # Mix or save single file try: if len(wav_files) == 1: cmd = [ffmpeg_path, "-y", "-i", wav_files[0], "-codec:a", "libmp3lame", "-qscale:a", "2", out_mp3] else: cmd = [ffmpeg_path, "-y"] for wav in wav_files: cmd.extend(["-i", wav]) cmd.extend(["-filter_complex", "amix=inputs=2:duration=longest", "-codec:a", "libmp3lame", "-qscale:a", "2", out_mp3]) subprocess.run(cmd, check=True) print(f"✅ Saved: {out_mp3}") # Clean up WAV files after successful MP3 conversion for wf in wav_files: try: os.remove(wf) except OSError: print(f"⚠️ Could not delete temporary WAV file: {wf}") except (subprocess.CalledProcessError, FileNotFoundError) as e: print(f"❌ FFmpeg failed: {e}") print(f"✅ Audio saved as WAV: {wav_files[0]}") # --- Tray & Ctrl+C handling --- def make_icon_image(is_rec=False): img = Image.new("RGB", (64, 64), "black") d = ImageDraw.Draw(img) d.ellipse((16, 16, 48, 48), fill="red" if is_rec else "grey") return img def start_record(icon, item): global recording, start_time if recording: return if not selected_mic: print("❌ Please select a microphone first") return recording = True start_time = time.time() print("▶ Recording started.") threading.Thread(target=record_worker, daemon=True).start() icon.icon = make_icon_image(True) icon.title = f"Recording: {selected_mic['name']}" icon.menu = build_menu() # Update menu with persisted selections def stop_record(icon=None, item=None): global recording, start_time if not recording: return elapsed = time.time() - start_time if start_time else 0 if elapsed < 5: print("⚠️ Stop ignored: recording too short (< 5s)") return print(f"🛑 Stop record called (elapsed: {elapsed:.1f}s)") recording = False if icon: icon.icon = make_icon_image(False) icon.title = "Saving..." time.sleep(1.0) save_recording() if icon: icon.title = f"Tray Recorder: {selected_mic['name'] if selected_mic else 'Idle'}" icon.menu = build_menu() # Update menu with persisted selections def quit_app(icon, item): stop_record(icon, item) icon.stop() # --- Main --- if __name__ == "__main__": # Initialize tray icon tray_icon = Icon( "TrayRecorder", icon=make_icon_image(False), title="Tray Recorder: Idle" ) # Set initial menu tray_icon.menu = build_menu() tray_icon.run_detached() print("Tray recorder running. Use tray menu to select devices and Start/Stop, or Ctrl+C to quit.") try: while True: time.sleep(1) except KeyboardInterrupt: print("\nInterrupted; stopping...") stop_record() tray_icon.stop() sys.exit(0)Key Changes and Explanations
-
Device Detection (
get_available_devices):- Added a check for
isLoopbackDeviceto identify devices that support WASAPI loopback. This is crucial for devices like the WH-1000XM4, which may not natively support loopback as a standard input device. - Ensured that output devices are included in the selection menu, even if they are not loopback devices, but loopback functionality is tested separately.
- Added a check for
-
Format Testing (
test_device_format):- Added an
is_loopbackparameter to handle loopback devices differently. For loopback, we only test the input format, as the device is being used to capture its own output. - Expanded the channel testing to include 2 and 1 channels explicitly, as some devices may only support stereo or mono configurations.
- Added an
-
Loopback Stream Setup (
record_worker):- Added logic to detect if the selected speaker device has a corresponding WASAPI loopback device. If not, loopback recording is disabled for that device.
- Used the
as_loopback=Trueparameter when opening the stream for loopback devices, which is necessary forpyaudiowpatchto work correctly. - Ensured that the output channels and sample rate of the device are used as the basis for the loopback input configuration.
-
Error Handling:
- Improved error messages and fallback behavior. If the WH-1000XM4 does not support loopback, the code now disables speaker recording gracefully instead of failing repeatedly.
- Removed the suggestion to use VB-Audio, as per your request.
-
Device Logging:
- Enhanced device logging to include the
isLoopbackDeviceflag, helping to diagnose whether the WH-1000XM4 is recognized as a loopback-capable device.
- Enhanced device logging to include the
Expected Behavior
- If the WH-1000XM4 Supports Loopback: The code will now correctly identify the loopback device (if available via WASAPI) and configure it for recording. You should see successful format support messages and recording from the headphones.
- If the WH-1000XM4 Does Not Support Loopback: The code will print a message indicating that loopback is not supported and disable speaker recording, allowing microphone recording to proceed without errors.
Troubleshooting
If the WH-1000XM4 still does not work for loopback recording, it is likely a hardware or driver limitation. In such cases, the only solution (without using VB-Audio) would be to use a different output device that supports WASAPI loopback. You can verify this by checking the device list printed at the start of the
record_workerfunction—look for devices withloopback=True.To test the fix, run the script, select the WH-1000XM4 as the output device, and start recording. Monitor the console output for success or failure messages related to the speaker loopback setup.
- Improving Device Detection: Ensure we correctly identify devices that support loopback (using
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.
Identical first prompt
- grok-3-preview-02-24Compare side by side →
- o4-mini-2025-04-16Compare side by side →
gemini-2.5-flash-preview-04-17Compare side by side →
gemini-2.0-flash-thinking-exp-01-21Compare side by side →
- o4-mini-2025-04-16Compare side by side →
- grok-3-mini-betaCompare side by side →
- chatgpt-4o-latest-20250326Compare side by side →
- o4-mini-2025-04-16Compare side by side →
- o3-miniCompare side by side →
llama-3.3-70b-instructCompare side by side →
- gpt-4.1-2025-04-14Compare side by side →