1berry/wokada-voice-changer
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

server device mode fallback

Browse Source
This commit is contained in:
wataru 2023-05-26 16:26:17 +09:00
parent 7746f3168f
commit 3d8dc7a9bd
10 changed files with 1793 additions and 2807 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
client/demo/dist/index.html vendored
View File

@ -1 +1,10 @@
<!doctype html><html style="width:100%;height:100%;overflow:hidden"><head><meta charset="utf-8"/><title>Voice Changer Client Demo</title><script defer="defer" src="index.js"></script></head><body style="width:100%;height:100%;margin:0"><div id="app" style="width:100%;height:100%"></div></body></html> <!DOCTYPE html>
<html style="width: 100%; height: 100%; overflow: hidden">
<head>
<meta charset="utf-8" />
<title>Voice Changer Client Demo</title>
<script defer src="index.js"></script></head>
<body style="width: 100%; height: 100%; margin: 0px">
<div id="app" style="width: 100%; height: 100%"></div>
</body>
</html>

1706
client/demo/dist/index.js vendored
View File

File diff suppressed because one or more lines are too long

2783
client/demo/dist/index.js.LICENSE.txt vendored
View File

File diff suppressed because it is too large Load Diff

16
client/demo/src/components/demo/components/202_PerformanceRow.tsx
View File

@ -9,14 +9,14 @@ export const PerformanceRow = (_props: PerformanceRowProps) => {
const [showPerformanceDetail, setShowPerformanceDetail] = useState<boolean>(false) const [showPerformanceDetail, setShowPerformanceDetail] = useState<boolean>(false)
const performanceRow = useMemo(() => { const performanceRow = useMemo(() => {
if (appState.serverSetting.serverSetting.enableServerAudio) { // if (appState.serverSetting.serverSetting.enableServerAudio) {
return ( // return (
<div className="body-row split-3-7 left-padding-1 guided"> // <div className="body-row split-3-7 left-padding-1 guided">
<div className="body-item-title left-padding-1">monitor:</div> // <div className="body-item-title left-padding-1">monitor:</div>
<div className="body-item-text">server device mode. refer console.</div> // <div className="body-item-text">server device mode. refer console.</div>
</div> // </div>
) // )
} // }
const performanceDetailLabel = showPerformanceDetail ? "[pre, main, post] <<" : "more >>" const performanceDetailLabel = showPerformanceDetail ? "[pre, main, post] <<" : "more >>"
const performanceData = showPerformanceDetail ? `[${appState.performance.preprocessTime}, ${appState.performance.mainprocessTime},${appState.performance.postprocessTime}]` : "" const performanceData = showPerformanceDetail ? `[${appState.performance.preprocessTime}, ${appState.performance.mainprocessTime},${appState.performance.postprocessTime}]` : ""
return ( return (

1
client/demo/src/components/demo/components/207_ServerOpertationRow.tsx
View File

@ -15,7 +15,6 @@ export const ServerOpertationRow = (props: ServerOpertationRowProps) => {
const guiState = useGuiState() const guiState = useGuiState()
const serverOpertationRow = useMemo(() => { const serverOpertationRow = useMemo(() => {
console.log("UPDATE OPERATION ROW:::::::::::::::::::::::")
const onnxExportButtonAction = async () => { const onnxExportButtonAction = async () => {

2
client/demo/src/components/demo/components/301-k_SampleDownloadControl.tsx
View File

@ -18,7 +18,7 @@ export const SampleDownloadControlRow = (_props: SampleDownloadControlRowProps)
} }
return ( return (
<div className="body-row split-3-3-4 left-padding-1 guided"> <div className="body-row split-3-3-4 left-padding-1 guided">
<div className="body-item-title left-padding-1 ">Advanced Configuration</div> <div className="body-item-title left-padding-2 ">Advanced Configuration</div>
<div> <div>
<input type="checkbox" checked={fileUploadSetting.rvcIndexDownload} onChange={(e) => { <input type="checkbox" checked={fileUploadSetting.rvcIndexDownload} onChange={(e) => {
appState.serverSetting.setFileUploadSetting(slot, { ...fileUploadSetting, rvcIndexDownload: e.target.checked }) appState.serverSetting.setFileUploadSetting(slot, { ...fileUploadSetting, rvcIndexDownload: e.target.checked })

21
client/lib/src/VoiceChangerWorkletNode.ts
View File

@ -74,15 +74,32 @@ export class VoiceChangerWorkletNode extends AudioWorkletNode {
this.listener.notifyException(VOICE_CHANGER_CLIENT_EXCEPTION.ERR_SIO_CONNECT_FAILED, `[SIO] rconnection failed ${err}`) this.listener.notifyException(VOICE_CHANGER_CLIENT_EXCEPTION.ERR_SIO_CONNECT_FAILED, `[SIO] rconnection failed ${err}`)
}) })
this.socket.on('connect', () => { this.socket.on('connect', () => {
console.log(`[SIO] sonnect to ${this.setting.serverUrl}`) console.log(`[SIO] connect to ${this.setting.serverUrl}`)
console.log(`[SIO] ${this.socket?.id}`) console.log(`[SIO] ${this.socket?.id}`)
}); });
this.socket.on('close', function (socket) {
console.log(`[SIO] close ${socket.id}`)
});
this.socket.on('message', (response: any[]) => {
console.log("message:", response)
});
this.socket.on('response', (response: any[]) => { this.socket.on('response', (response: any[]) => {
// console.log("response:", response)
const cur = Date.now() const cur = Date.now()
const responseTime = cur - response[0] const responseTime = cur - response[0]
const result = response[1] as ArrayBuffer const result = response[1] as ArrayBuffer
const perf = response[2] const perf = response[2]
// Quick hack for server device mode
if (response[0] == 0) {
this.listener.notifyResponseTime(Math.round(perf[0] * 1000), perf.slice(1, 4))
return
}
if (result.byteLength < 128 * 2) { if (result.byteLength < 128 * 2) {
this.listener.notifyException(VOICE_CHANGER_CLIENT_EXCEPTION.ERR_SIO_INVALID_RESPONSE, `[SIO] recevied data is too short ${result.byteLength}`) this.listener.notifyException(VOICE_CHANGER_CLIENT_EXCEPTION.ERR_SIO_INVALID_RESPONSE, `[SIO] recevied data is too short ${result.byteLength}`)
} else { } else {

15
server/sio/MMVC_Namespace.py
View File

@ -4,13 +4,27 @@ import numpy as np
import socketio import socketio
from voice_changer.VoiceChangerManager import VoiceChangerManager from voice_changer.VoiceChangerManager import VoiceChangerManager
import asyncio
class MMVC_Namespace(socketio.AsyncNamespace): class MMVC_Namespace(socketio.AsyncNamespace):
sid: int = 0 sid: int = 0
async def emitTo(self, data):
timestamp = 0
audio1 = np.zeros(1).astype(np.int16)
bin = struct.pack("<%sh" % len(audio1), *audio1)
perf = data
await self.emit("response", [timestamp, bin, perf], to=self.sid)
def emit_coroutine(self, data):
asyncio.run(self.emitTo(data))
def __init__(self, namespace: str, voiceChangerManager: VoiceChangerManager): def __init__(self, namespace: str, voiceChangerManager: VoiceChangerManager):
super().__init__(namespace) super().__init__(namespace)
self.voiceChangerManager = voiceChangerManager self.voiceChangerManager = voiceChangerManager
self.voiceChangerManager.voiceChanger.emitTo = self.emit_coroutine
@classmethod @classmethod
def get_instance(cls, voiceChangerManager: VoiceChangerManager): def get_instance(cls, voiceChangerManager: VoiceChangerManager):
@ -29,6 +43,7 @@ class MMVC_Namespace(socketio.AsyncNamespace):
async def on_request_message(self, sid, msg): async def on_request_message(self, sid, msg):
self.sid = sid self.sid = sid
await self.asynctest("on req")
timestamp = int(msg[0]) timestamp = int(msg[0])
data = msg[1] data = msg[1]
if isinstance(data, str): if isinstance(data, str):

7
server/voice_changer/Local/AudioDeviceList.py
View File

@ -11,13 +11,14 @@ class ServerAudioDevice:
hostAPI: str = "" hostAPI: str = ""
maxInputChannels: int = 0 maxInputChannels: int = 0
maxOutputChannels: int = 0 maxOutputChannels: int = 0
default_samplerate: int = 0
def list_audio_device(): def list_audio_device():
audioDeviceList = sd.query_devices() audioDeviceList = sd.query_devices()
inputAudioDeviceList = [d for d in audioDeviceList if d["max_input_channels"] > 0] inputAudioDeviceList = [d for d in audioDeviceList if d["max_input_channels"] > 0]
outputDeviceList = [d for d in audioDeviceList if d["max_output_channels"] > 0] outputAudioDeviceList = [d for d in audioDeviceList if d["max_output_channels"] > 0]
hostapis = sd.query_hostapis() hostapis = sd.query_hostapis()
# print("input:", inputAudioDeviceList) # print("input:", inputAudioDeviceList)
@ -34,9 +35,10 @@ def list_audio_device():
hostAPI=hostapis[d["hostapi"]]["name"], hostAPI=hostapis[d["hostapi"]]["name"],
maxInputChannels=d["max_input_channels"], maxInputChannels=d["max_input_channels"],
maxOutputChannels=d["max_output_channels"], maxOutputChannels=d["max_output_channels"],
default_samplerate=d["default_samplerate"],
) )
serverAudioInputDevices.append(serverInputAudioDevice) serverAudioInputDevices.append(serverInputAudioDevice)
for d in outputDeviceList: for d in outputAudioDeviceList:
serverOutputAudioDevice: ServerAudioDevice = ServerAudioDevice( serverOutputAudioDevice: ServerAudioDevice = ServerAudioDevice(
kind=ServerAudioDeviceTypes.audiooutput, kind=ServerAudioDeviceTypes.audiooutput,
index=d["index"], index=d["index"],
@ -44,6 +46,7 @@ def list_audio_device():
hostAPI=hostapis[d["hostapi"]]["name"], hostAPI=hostapis[d["hostapi"]]["name"],
maxInputChannels=d["max_input_channels"], maxInputChannels=d["max_input_channels"],
maxOutputChannels=d["max_output_channels"], maxOutputChannels=d["max_output_channels"],
default_samplerate=d["default_samplerate"],
) )
serverAudioOutputDevices.append(serverOutputAudioDevice) serverAudioOutputDevices.append(serverOutputAudioDevice)

38
server/voice_changer/VoiceChanger.py
View File

@ -99,6 +99,8 @@ class VoiceChanger:
localPerformanceShowTime = 0.0 localPerformanceShowTime = 0.0
emitTo = None
def audio_callback( def audio_callback(
self, indata: np.ndarray, outdata: np.ndarray, frames, times, status self, indata: np.ndarray, outdata: np.ndarray, frames, times, status
): ):
@ -113,6 +115,8 @@ class VoiceChanger:
) )
all_inference_time = t.secs all_inference_time = t.secs
performance = [all_inference_time] + times performance = [all_inference_time] + times
if self.emitTo is not None:
self.emitTo(performance)
self.settings.performance = [round(x * 1000) for x in performance] self.settings.performance = [round(x * 1000) for x in performance]
except Exception as e: except Exception as e:
print("[Voice Changer] ex:", e) print("[Voice Changer] ex:", e)
@ -131,6 +135,7 @@ class VoiceChanger:
currentInputDeviceId = -1 currentInputDeviceId = -1
currentInputSampleRate = -1 currentInputSampleRate = -1
model_sampling_rate = -1
currentOutputDeviceId = -1 currentOutputDeviceId = -1
currentInputChunkNum = -1 currentInputChunkNum = -1
while True: while True:
@ -166,13 +171,33 @@ class VoiceChanger:
currentInputChannelNum = serverInputAudioDevice.maxInputChannels currentInputChannelNum = serverInputAudioDevice.maxInputChannels
currentOutputChannelNum = serverOutputAudioDevice.maxOutputChannels currentOutputChannelNum = serverOutputAudioDevice.maxOutputChannels
vc.settings.serverInputAudioSampleRate = (
self.voiceChanger.get_processing_sampling_rate()
)
currentInputSampleRate = vc.settings.serverInputAudioSampleRate
currentInputChunkNum = vc.settings.serverReadChunkSize currentInputChunkNum = vc.settings.serverReadChunkSize
block_frame = currentInputChunkNum * 128 block_frame = currentInputChunkNum * 128
# sample rate precheck(alsa cannot use 40000?)
model_sampling_rate = self.voiceChanger.get_processing_sampling_rate()
try:
with sd.Stream(
callback=self.audio_callback,
blocksize=block_frame,
samplerate=model_sampling_rate,
dtype="float32",
channels=[currentInputChannelNum, currentOutputChannelNum],
):
pass
currentInputSampleRate = model_sampling_rate
vc.settings.serverInputAudioSampleRate = model_sampling_rate
print(f"[Voice Changer] sample rate {model_sampling_rate}")
except Exception as e:
print(
"[Voice Changer] ex: fallback to device default samplerate", e
)
currentInputSampleRate = serverInputAudioDevice.default_samplerate
vc.settings.serverInputAudioSampleRate = (
serverInputAudioDevice.default_samplerate
)
# main loop
try: try:
with sd.Stream( with sd.Stream(
callback=self.audio_callback, callback=self.audio_callback,
@ -181,14 +206,13 @@ class VoiceChanger:
dtype="float32", dtype="float32",
channels=[currentInputChannelNum, currentOutputChannelNum], channels=[currentInputChannelNum, currentOutputChannelNum],
): ):
print()
while ( while (
vc.settings.serverAudioStated == 1 vc.settings.serverAudioStated == 1
and currentInputDeviceId == vc.settings.serverInputDeviceId and currentInputDeviceId == vc.settings.serverInputDeviceId
and currentOutputDeviceId and currentOutputDeviceId
== vc.settings.serverOutputDeviceId == vc.settings.serverOutputDeviceId
and currentInputSampleRate and model_sampling_rate
== vc.settings.serverInputAudioSampleRate == self.voiceChanger.get_processing_sampling_rate()
and currentInputChunkNum == vc.settings.serverReadChunkSize and currentInputChunkNum == vc.settings.serverReadChunkSize
): ):
vc.settings.serverInputAudioSampleRate = ( vc.settings.serverInputAudioSampleRate = (