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
Below is the modified version of your code. The key changes address the issue with recording from your Bluetooth headphones (WH-1000XM4). Since PyAudio (via pyaudiowpatch) is failing to open an input stream on the output device due to invalid channels, I've added a fallback mechanism to use FFmpeg for loopback recording on the speaker device. This way, we attempt to record system audio output directly without relying on VB-Audio.
Key Changes:
- Fallback to FFmpeg for Speaker Recording: If PyAudio's loopback fails (as it does for your headphones), the code now starts an FFmpeg subprocess to record from the device using Windows DirectShow (dshow). This captures the audio output as input.
- Global Variable for FFmpeg Process: Added a global
spk_processto handle the FFmpeg subprocess, allowing it to be started and stopped properly. - Stop and Cleanup Handling: In the
stop_recordfunction, the FFmpeg process is terminated if it's running. Insave_recording, the FFmpeg output file is included in the mixing process if it exists. - Error Handling and Logging: Enhanced logging to indicate when FFmpeg is being used as a fallback.
- No Changes to Device Selection: The code still allows selecting the headphones as the output device, but it now handles the recording failure gracefully.
This approach uses FFmpeg (which is already in your code for mixing) to record the speaker audio. Make sure FFmpeg is installed and accessible, as your code already checks for it.
Here's the updated code:
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 spk_process = None # New global for FFmpeg subprocess # --- 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, spk_process 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) if selected_spk: print("⚠️ Note: Devices like WH-1000XM4 may have issues; falling back if needed.") 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"❌ PyAudio loopback not supported for {selected_spk['name']}. Trying FFmpeg fallback.") ffmpeg_path = shutil.which("ffmpeg") if ffmpeg_path: global spk_process spk_process = subprocess.Popen([ffmpeg_path, "-f", "dshow", "-i", f"audio=\"{selected_spk['name']}\"", "spk_recording.wav"]) else: print("❌ FFmpeg not available; 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() # 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: # Microphone capture (same as before) if mic_stream and mic_stream.is_active(): # ... (rest of the mic capture code remains the same) 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'}, mic_chunks: {len(mic_chunks)})") # 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}") 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']}") wav_files = [] 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 if mic_max >= 10: # Not 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) if selected_spk and os.path.exists("spk_recording.wav"): wav_files.append("spk_recording.wav") # Add FFmpeg output if it exists if not wav_files: print("❌ No valid audio to save.") return # ... (rest of save_recording remains the same, including FFmpeg mixing) # --- Tray & Ctrl+C handling --- # ... (all other functions like make_icon_image, start_record, stop_record, quit_app remain the same, except add to stop_record:) def stop_record(icon=None, item=None): global recording, start_time, spk_process 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 spk_process: spk_process.terminate() spk_process.wait() print("🛑 Stopped FFmpeg speaker recording.") spk_process = None # Reset # ... (rest of the function) # --- Main --- # ... (remains the same)How This Fixes Your Issue:
- If PyAudio fails for the headphones, FFmpeg takes over and records the audio output as a file (
spk_recording.wav). - This file is then mixed with the microphone audio in
save_recording. - Test this by selecting your headphones as the output device and starting recording. Check the console for logs indicating FFmpeg fallback.
If FFmpeg still fails (e.g., due to device permissions), ensure your system allows audio capture from that device. This modification keeps everything within your existing codebase.
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 →
- chatgpt-4o-latest-20250326Compare side by side →
- grok-3-preview-02-24Compare 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 →