luke/ai-podcast
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add live caller channel to audio settings

Browse Source 
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
tcpsyn
2026-02-05 16:03:52 -07:00
parent edcd5ebb1b
commit cca8eaad84
4 changed files with 21 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
backend/main.py
View File

@@ -436,6 +436,7 @@ class AudioDeviceSettings(BaseModel):
input_channel: Optional[int] = None input_channel: Optional[int] = None
output_device: Optional[int] = None output_device: Optional[int] = None
caller_channel: Optional[int] = None caller_channel: Optional[int] = None
live_caller_channel: Optional[int] = None
music_channel: Optional[int] = None music_channel: Optional[int] = None
sfx_channel: Optional[int] = None sfx_channel: Optional[int] = None
phone_filter: Optional[bool] = None phone_filter: Optional[bool] = None
@@ -471,6 +472,7 @@ async def set_audio_settings(settings: AudioDeviceSettings):
input_channel=settings.input_channel, input_channel=settings.input_channel,
output_device=settings.output_device, output_device=settings.output_device,
caller_channel=settings.caller_channel, caller_channel=settings.caller_channel,
live_caller_channel=settings.live_caller_channel,
music_channel=settings.music_channel, music_channel=settings.music_channel,
sfx_channel=settings.sfx_channel, sfx_channel=settings.sfx_channel,
phone_filter=settings.phone_filter phone_filter=settings.phone_filter
@@ -832,9 +834,8 @@ async def caller_audio_stream(websocket: WebSocket):
pcm_data = message["bytes"] pcm_data = message["bytes"]
audio_buffer.extend(pcm_data) audio_buffer.extend(pcm_data)
# Route to Loopback channel # Route to configured live caller Loopback channel
channel = call_info["channel"] audio_service.route_real_caller_audio(pcm_data, SAMPLE_RATE)
audio_service.route_real_caller_audio(pcm_data, channel, SAMPLE_RATE)
# Transcribe when we have enough audio # Transcribe when we have enough audio
if len(audio_buffer) >= chunk_samples * 2: if len(audio_buffer) >= chunk_samples * 2:

15
backend/services/audio.py
View File

@@ -24,6 +24,7 @@ class AudioService:
self.output_device: Optional[int] = None # Single output device (multi-channel) self.output_device: Optional[int] = None # Single output device (multi-channel)
self.caller_channel: int = 1 # Channel for caller TTS self.caller_channel: int = 1 # Channel for caller TTS
self.live_caller_channel: int = 4 # Channel for live caller audio
self.music_channel: int = 2 # Channel for music self.music_channel: int = 2 # Channel for music
self.sfx_channel: int = 3 # Channel for SFX self.sfx_channel: int = 3 # Channel for SFX
self.phone_filter: bool = False # Phone filter on caller voices self.phone_filter: bool = False # Phone filter on caller voices
@@ -69,10 +70,11 @@ class AudioService:
self.input_channel = data.get("input_channel", 1) self.input_channel = data.get("input_channel", 1)
self.output_device = data.get("output_device") self.output_device = data.get("output_device")
self.caller_channel = data.get("caller_channel", 1) self.caller_channel = data.get("caller_channel", 1)
self.live_caller_channel = data.get("live_caller_channel", 4)
self.music_channel = data.get("music_channel", 2) self.music_channel = data.get("music_channel", 2)
self.sfx_channel = data.get("sfx_channel", 3) self.sfx_channel = data.get("sfx_channel", 3)
self.phone_filter = data.get("phone_filter", False) self.phone_filter = data.get("phone_filter", False)
print(f"Loaded audio settings: output={self.output_device}, channels={self.caller_channel}/{self.music_channel}/{self.sfx_channel}, phone_filter={self.phone_filter}") print(f"Loaded audio settings: output={self.output_device}, channels={self.caller_channel}/{self.live_caller_channel}/{self.music_channel}/{self.sfx_channel}, phone_filter={self.phone_filter}")
except Exception as e: except Exception as e:
print(f"Failed to load audio settings: {e}") print(f"Failed to load audio settings: {e}")
@@ -84,6 +86,7 @@ class AudioService:
"input_channel": self.input_channel, "input_channel": self.input_channel,
"output_device": self.output_device, "output_device": self.output_device,
"caller_channel": self.caller_channel, "caller_channel": self.caller_channel,
"live_caller_channel": self.live_caller_channel,
"music_channel": self.music_channel, "music_channel": self.music_channel,
"sfx_channel": self.sfx_channel, "sfx_channel": self.sfx_channel,
"phone_filter": self.phone_filter, "phone_filter": self.phone_filter,
@@ -114,6 +117,7 @@ class AudioService:
input_channel: Optional[int] = None, input_channel: Optional[int] = None,
output_device: Optional[int] = None, output_device: Optional[int] = None,
caller_channel: Optional[int] = None, caller_channel: Optional[int] = None,
live_caller_channel: Optional[int] = None,
music_channel: Optional[int] = None, music_channel: Optional[int] = None,
sfx_channel: Optional[int] = None, sfx_channel: Optional[int] = None,
phone_filter: Optional[bool] = None phone_filter: Optional[bool] = None
@@ -127,6 +131,8 @@ class AudioService:
self.output_device = output_device self.output_device = output_device
if caller_channel is not None: if caller_channel is not None:
self.caller_channel = caller_channel self.caller_channel = caller_channel
if live_caller_channel is not None:
self.live_caller_channel = live_caller_channel
if music_channel is not None: if music_channel is not None:
self.music_channel = music_channel self.music_channel = music_channel
if sfx_channel is not None: if sfx_channel is not None:
@@ -144,6 +150,7 @@ class AudioService:
"input_channel": self.input_channel, "input_channel": self.input_channel,
"output_device": self.output_device, "output_device": self.output_device,
"caller_channel": self.caller_channel, "caller_channel": self.caller_channel,
"live_caller_channel": self.live_caller_channel,
"music_channel": self.music_channel, "music_channel": self.music_channel,
"sfx_channel": self.sfx_channel, "sfx_channel": self.sfx_channel,
"phone_filter": self.phone_filter, "phone_filter": self.phone_filter,
@@ -313,8 +320,8 @@ class AudioService:
"""Stop any playing caller audio""" """Stop any playing caller audio"""
self._caller_stop_event.set() self._caller_stop_event.set()
def route_real_caller_audio(self, pcm_data: bytes, channel: int, sample_rate: int): def route_real_caller_audio(self, pcm_data: bytes, sample_rate: int):
"""Route real caller PCM audio to a specific Loopback channel""" """Route real caller PCM audio to the configured live caller Loopback channel"""
import librosa import librosa
if self.output_device is None: if self.output_device is None:
@@ -327,7 +334,7 @@ class AudioService:
device_info = sd.query_devices(self.output_device) device_info = sd.query_devices(self.output_device)
num_channels = device_info['max_output_channels'] num_channels = device_info['max_output_channels']
device_sr = int(device_info['default_samplerate']) device_sr = int(device_info['default_samplerate'])
channel_idx = min(channel, num_channels) - 1 channel_idx = min(self.live_caller_channel, num_channels) - 1
# Resample to device sample rate if needed # Resample to device sample rate if needed
if sample_rate != device_sr: if sample_rate != device_sr:

3
frontend/index.html
View File

@@ -128,6 +128,7 @@
</div> </div>
<div class="channel-row"> <div class="channel-row">
<label>Caller Ch <input type="number" id="caller-channel" value="1" min="1" max="16" class="channel-input"></label> <label>Caller Ch <input type="number" id="caller-channel" value="1" min="1" max="16" class="channel-input"></label>
<label>Live Ch <input type="number" id="live-caller-channel" value="4" min="1" max="16" class="channel-input"></label>
<label>Music Ch <input type="number" id="music-channel" value="2" min="1" max="16" class="channel-input"></label> <label>Music Ch <input type="number" id="music-channel" value="2" min="1" max="16" class="channel-input"></label>
<label>SFX Ch <input type="number" id="sfx-channel" value="3" min="1" max="16" class="channel-input"></label> <label>SFX Ch <input type="number" id="sfx-channel" value="3" min="1" max="16" class="channel-input"></label>
</div> </div>
@@ -206,6 +207,6 @@
</div> </div>
</div> </div>
<script src="/js/app.js?v=10"></script> <script src="/js/app.js?v=11"></script>
</body> </body>
</html> </html>

4
frontend/js/app.js
View File

@@ -223,11 +223,13 @@ async function loadAudioDevices() {
// Channel settings // Channel settings
const inputCh = document.getElementById('input-channel'); const inputCh = document.getElementById('input-channel');
const callerCh = document.getElementById('caller-channel'); const callerCh = document.getElementById('caller-channel');
const liveCallerCh = document.getElementById('live-caller-channel');
const musicCh = document.getElementById('music-channel'); const musicCh = document.getElementById('music-channel');
const sfxCh = document.getElementById('sfx-channel'); const sfxCh = document.getElementById('sfx-channel');
if (inputCh) inputCh.value = settings.input_channel || 1; if (inputCh) inputCh.value = settings.input_channel || 1;
if (callerCh) callerCh.value = settings.caller_channel || 1; if (callerCh) callerCh.value = settings.caller_channel || 1;
if (liveCallerCh) liveCallerCh.value = settings.live_caller_channel || 4;
if (musicCh) musicCh.value = settings.music_channel || 2; if (musicCh) musicCh.value = settings.music_channel || 2;
if (sfxCh) sfxCh.value = settings.sfx_channel || 3; if (sfxCh) sfxCh.value = settings.sfx_channel || 3;
@@ -250,6 +252,7 @@ async function saveAudioDevices() {
const outputDevice = document.getElementById('output-device')?.value; const outputDevice = document.getElementById('output-device')?.value;
const inputChannel = document.getElementById('input-channel')?.value; const inputChannel = document.getElementById('input-channel')?.value;
const callerChannel = document.getElementById('caller-channel')?.value; const callerChannel = document.getElementById('caller-channel')?.value;
const liveCallerChannel = document.getElementById('live-caller-channel')?.value;
const musicChannel = document.getElementById('music-channel')?.value; const musicChannel = document.getElementById('music-channel')?.value;
const sfxChannel = document.getElementById('sfx-channel')?.value; const sfxChannel = document.getElementById('sfx-channel')?.value;
const phoneFilterChecked = document.getElementById('phone-filter')?.checked ?? false; const phoneFilterChecked = document.getElementById('phone-filter')?.checked ?? false;
@@ -262,6 +265,7 @@ async function saveAudioDevices() {
input_channel: inputChannel ? parseInt(inputChannel) : 1, input_channel: inputChannel ? parseInt(inputChannel) : 1,
output_device: outputDevice ? parseInt(outputDevice) : null, output_device: outputDevice ? parseInt(outputDevice) : null,
caller_channel: callerChannel ? parseInt(callerChannel) : 1, caller_channel: callerChannel ? parseInt(callerChannel) : 1,
live_caller_channel: liveCallerChannel ? parseInt(liveCallerChannel) : 4,
music_channel: musicChannel ? parseInt(musicChannel) : 2, music_channel: musicChannel ? parseInt(musicChannel) : 2,
sfx_channel: sfxChannel ? parseInt(sfxChannel) : 3, sfx_channel: sfxChannel ? parseInt(sfxChannel) : 3,
phone_filter: phoneFilterChecked phone_filter: phoneFilterChecked