Bugfix:

- device output recorder button is showed in server device mode.
Feature:
 - server device monitor
Improve:
 - default uvicorn error log

client/demo/package.json

@@ -26,14 +26,14 @@
         "@babel/preset-env": "^7.22.5",
         "@babel/preset-react": "^7.22.5",
         "@babel/preset-typescript": "^7.22.5",
-        "@types/node": "^20.3.2",
+        "@types/node": "^20.3.3",
         "@types/react": "^18.2.14",
         "@types/react-dom": "^18.2.6",
         "autoprefixer": "^10.4.14",
         "babel-loader": "^9.1.2",
         "copy-webpack-plugin": "^11.0.0",
         "css-loader": "^6.8.1",
-        "eslint": "^8.43.0",
+        "eslint": "^8.44.0",
         "eslint-config-prettier": "^8.8.0",
         "eslint-plugin-prettier": "^4.2.1",
         "eslint-plugin-react": "^7.32.2",
@@ -54,7 +54,7 @@
         "webpack-dev-server": "^4.15.1"
     },
     "dependencies": {
-        "@dannadori/voice-changer-client-js": "^1.0.155",
+        "@dannadori/voice-changer-client-js": "^1.0.157",
         "@fortawesome/fontawesome-svg-core": "^6.4.0",
         "@fortawesome/free-brands-svg-icons": "^6.4.0",
         "@fortawesome/free-regular-svg-icons": "^6.4.0",

client/demo/src/components/demo/components2/102-3_DeviceArea.tsx

@@ -1,35 +1,34 @@
-import React, { useEffect, useMemo, useRef, useState } from "react"
+import React, { useEffect, useMemo, useRef, useState } from "react";
-import { useAppState } from "../../../001_provider/001_AppStateProvider"
+import { useAppState } from "../../../001_provider/001_AppStateProvider";
-import { fileSelectorAsDataURL, useIndexedDB, } from "@dannadori/voice-changer-client-js"
+import { fileSelectorAsDataURL, useIndexedDB } from "@dannadori/voice-changer-client-js";
-import { useGuiState } from "../001_GuiStateProvider"
+import { useGuiState } from "../001_GuiStateProvider";
-import { AUDIO_ELEMENT_FOR_PLAY_RESULT, AUDIO_ELEMENT_FOR_TEST_CONVERTED, AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK, AUDIO_ELEMENT_FOR_TEST_ORIGINAL, INDEXEDDB_KEY_AUDIO_OUTPUT } from "../../../const"
+import { AUDIO_ELEMENT_FOR_PLAY_RESULT, AUDIO_ELEMENT_FOR_TEST_CONVERTED, AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK, AUDIO_ELEMENT_FOR_TEST_ORIGINAL, INDEXEDDB_KEY_AUDIO_OUTPUT } from "../../../const";
 
-export type DeviceAreaProps = {
+export type DeviceAreaProps = {};
-}
 
 export const DeviceArea = (_props: DeviceAreaProps) => {
-    const { setting, serverSetting, audioContext, setAudioOutputElementId, initializedRef, setVoiceChangerClientSetting, startOutputRecording, stopOutputRecording } = useAppState()
+    const { setting, serverSetting, audioContext, setAudioOutputElementId, initializedRef, setVoiceChangerClientSetting, startOutputRecording, stopOutputRecording } = useAppState();
-    const { isConverting, audioInputForGUI, inputAudioDeviceInfo, setAudioInputForGUI, fileInputEchoback, setFileInputEchoback, setAudioOutputForGUI, audioOutputForGUI, outputAudioDeviceInfo } = useGuiState()
+    const { isConverting, audioInputForGUI, inputAudioDeviceInfo, setAudioInputForGUI, fileInputEchoback, setFileInputEchoback, setAudioOutputForGUI, audioOutputForGUI, outputAudioDeviceInfo } = useGuiState();
-    const [inputHostApi, setInputHostApi] = useState<string>("ALL")
+    const [inputHostApi, setInputHostApi] = useState<string>("ALL");
-    const [outputHostApi, setOutputHostApi] = useState<string>("ALL")
+    const [outputHostApi, setOutputHostApi] = useState<string>("ALL");
-    const audioSrcNode = useRef<MediaElementAudioSourceNode>()
+    const audioSrcNode = useRef<MediaElementAudioSourceNode>();
 
-    const { getItem, setItem } = useIndexedDB({ clientType: null })
+    const { getItem, setItem } = useIndexedDB({ clientType: null });
-    const [outputRecordingStarted, setOutputRecordingStarted] = useState<boolean>(false)
+    const [outputRecordingStarted, setOutputRecordingStarted] = useState<boolean>(false);
 
     // (1) Audio Mode
     const deviceModeRow = useMemo(() => {
-        const enableServerAudio = serverSetting.serverSetting.enableServerAudio
+        const enableServerAudio = serverSetting.serverSetting.enableServerAudio;
-        const clientChecked = enableServerAudio == 1 ? false : true
+        const clientChecked = enableServerAudio == 1 ? false : true;
-        const serverChecked = enableServerAudio == 1 ? true : false
+        const serverChecked = enableServerAudio == 1 ? true : false;
 
         const onDeviceModeChanged = (val: number) => {
             if (isConverting) {
-                alert("cannot change mode when voice conversion is enabled")
+                alert("cannot change mode when voice conversion is enabled");
-                return
+                return;
-            }
-            serverSetting.updateServerSettings({ ...serverSetting.serverSetting, enableServerAudio: val })
             }
+            serverSetting.updateServerSettings({ ...serverSetting.serverSetting, enableServerAudio: val });
+        };
 
         return (
             <div className="config-sub-area-control">
@@ -37,131 +36,182 @@ export const DeviceArea = (_props: DeviceAreaProps) => {
                 <div className="config-sub-area-control-field">
                     <div className="config-sub-area-noise-container">
                         <div className="config-sub-area-noise-checkbox-container">
-                            <input type="radio" id="client-device" name="device-mode" checked={clientChecked} onChange={() => { onDeviceModeChanged(0) }} /> <label htmlFor="client-device">client</label>
+                            <input
+                                type="radio"
+                                id="client-device"
+                                name="device-mode"
+                                checked={clientChecked}
+                                onChange={() => {
+                                    onDeviceModeChanged(0);
+                                }}
+                            />{" "}
+                            <label htmlFor="client-device">client</label>
                         </div>
                         <div className="config-sub-area-noise-checkbox-container">
-                            <input className="left-padding-1" type="radio" id="server-device" name="device-mode" checked={serverChecked} onChange={() => { onDeviceModeChanged(1) }} />
+                            <input
+                                className="left-padding-1"
+                                type="radio"
+                                id="server-device"
+                                name="device-mode"
+                                checked={serverChecked}
+                                onChange={() => {
+                                    onDeviceModeChanged(1);
+                                }}
+                            />
                             <label htmlFor="server-device">server</label>
                         </div>
                     </div>
                 </div>
             </div>
-        )
+        );
-    }, [serverSetting.serverSetting, serverSetting.updateServerSettings, isConverting])
+    }, [serverSetting.serverSetting, serverSetting.updateServerSettings, isConverting]);
-
-
 
     // (2) Audio Input
     // キャッシュの設定は反映（たぶん、設定操作の時も起動していしまう。が問題は起こらないはず）
     useEffect(() => {
         if (typeof setting.voiceChangerClientSetting.audioInput == "string") {
-            if (inputAudioDeviceInfo.find(x => {
+            if (
+                inputAudioDeviceInfo.find((x) => {
                     // console.log("COMPARE:", x.deviceId, appState.clientSetting.setting.audioInput)
-                return x.deviceId == setting.voiceChangerClientSetting.audioInput
+                    return x.deviceId == setting.voiceChangerClientSetting.audioInput;
-            })) {
+                })
-                setAudioInputForGUI(setting.voiceChangerClientSetting.audioInput)
+            ) {
+                setAudioInputForGUI(setting.voiceChangerClientSetting.audioInput);
             }
         }
-    }, [inputAudioDeviceInfo, setting.voiceChangerClientSetting.audioInput])
+    }, [inputAudioDeviceInfo, setting.voiceChangerClientSetting.audioInput]);
 
     // (2-1) クライアント
     const clientAudioInputRow = useMemo(() => {
         if (serverSetting.serverSetting.enableServerAudio == 1) {
-            return <></>
+            return <></>;
         }
 
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-title left-padding-1">input</div>
                 <div className="config-sub-area-control-field">
-                    <select className="body-select" value={audioInputForGUI} onChange={async (e) => {
+                    <select
-                        setAudioInputForGUI(e.target.value)
+                        className="body-select"
+                        value={audioInputForGUI}
+                        onChange={async (e) => {
+                            setAudioInputForGUI(e.target.value);
                             if (e.target.value != "file") {
                                 try {
-                                await setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: e.target.value })
+                                    await setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: e.target.value });
                                 } catch (e) {
-                                alert(e)
+                                    alert(e);
-                                console.error(e)
+                                    console.error(e);
-                                setAudioInputForGUI("none")
+                                    setAudioInputForGUI("none");
-                                await setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: null })
+                                    await setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: null });
                                 }
                             }
-                    }}>
+                        }}
-                        {
+                    >
-                            inputAudioDeviceInfo.map(x => {
+                        {inputAudioDeviceInfo.map((x) => {
-                                return <option key={x.deviceId} value={x.deviceId}>{x.label}</option>
+                            return (
-                            })
+                                <option key={x.deviceId} value={x.deviceId}>
-                        }
+                                    {x.label}
+                                </option>
+                            );
+                        })}
                     </select>
                 </div>
             </div>
-        )
+        );
-    }, [setVoiceChangerClientSetting, setting, inputAudioDeviceInfo, audioInputForGUI, serverSetting.serverSetting.enableServerAudio])
+    }, [setVoiceChangerClientSetting, setting, inputAudioDeviceInfo, audioInputForGUI, serverSetting.serverSetting.enableServerAudio]);
-
 
     // (2-2) サーバ
     const serverAudioInputRow = useMemo(() => {
         if (serverSetting.serverSetting.enableServerAudio == 0) {
-            return <></>
+            return <></>;
         }
 
-        const devices = serverSetting.serverSetting.serverAudioInputDevices
+        const devices = serverSetting.serverSetting.serverAudioInputDevices;
-        const hostAPIs = new Set(devices.map(x => { return x.hostAPI }))
+        const hostAPIs = new Set(
-        const hostAPIOptions = Array.from(hostAPIs).map((x, index) => { return <option value={x} key={index} >{x}</option> })
+            devices.map((x) => {
+                return x.hostAPI;
+            })
+        );
+        const hostAPIOptions = Array.from(hostAPIs).map((x, index) => {
+            return (
+                <option value={x} key={index}>
+                    {x}
+                </option>
+            );
+        });
 
-        const filteredDevice = devices.map((x, index) => {
+        const filteredDevice = devices
+            .map((x, index) => {
                 if (inputHostApi != "ALL" && x.hostAPI != inputHostApi) {
-                return null
+                    return null;
                 }
-            return <option value={x.index} key={index}>[{x.hostAPI}]{x.name}</option>
+                return (
-        }).filter(x => x != null)
+                    <option value={x.index} key={index}>
+                        [{x.hostAPI}]{x.name}
+                    </option>
+                );
+            })
+            .filter((x) => x != null);
 
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-title left-padding-1">input</div>
                 <div className="config-sub-area-control-field">
                     <div className="config-sub-area-control-field-auido-io">
-                        <select className="config-sub-area-control-field-auido-io-filter" name="kinds" id="kinds" value={inputHostApi} onChange={(e) => { setInputHostApi(e.target.value) }}>
+                        <select
-                            <option value="ALL" key="ALL" >ALL</option>
+                            className="config-sub-area-control-field-auido-io-filter"
+                            name="kinds"
+                            id="kinds"
+                            value={inputHostApi}
+                            onChange={(e) => {
+                                setInputHostApi(e.target.value);
+                            }}
+                        >
+                            <option value="ALL" key="ALL">
+                                ALL
+                            </option>
                             {hostAPIOptions}
                         </select>
-                        <select className="config-sub-area-control-field-auido-io-select" value={serverSetting.serverSetting.serverInputDeviceId} onChange={(e) => {
+                        <select
-                            serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverInputDeviceId: Number(e.target.value) })
+                            className="config-sub-area-control-field-auido-io-select"
-
+                            value={serverSetting.serverSetting.serverInputDeviceId}
-                        }}>
+                            onChange={(e) => {
+                                serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverInputDeviceId: Number(e.target.value) });
+                            }}
+                        >
                             {filteredDevice}
                         </select>
                     </div>
                 </div>
             </div>
-        )
+        );
-
+    }, [inputHostApi, serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio]);
-    }, [inputHostApi, serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio])
 
     // (2-3) File
     useEffect(() => {
-        [AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK].forEach(x => {
+        [AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK].forEach((x) => {
-            const audio = document.getElementById(x) as HTMLAudioElement
+            const audio = document.getElementById(x) as HTMLAudioElement;
             if (audio) {
-                audio.volume = fileInputEchoback ? 1 : 0
+                audio.volume = fileInputEchoback ? 1 : 0;
             }
-        })
+        });
-    }, [fileInputEchoback])
+    }, [fileInputEchoback]);
 
     const audioInputMediaRow = useMemo(() => {
         if (audioInputForGUI != "file" || serverSetting.serverSetting.enableServerAudio == 1) {
-            return <></>
+            return <></>;
         }
 
         const onFileLoadClicked = async () => {
-            const url = await fileSelectorAsDataURL("")
+            const url = await fileSelectorAsDataURL("");
 
             // input stream for client.
-            const audio = document.getElementById(AUDIO_ELEMENT_FOR_TEST_CONVERTED) as HTMLAudioElement
+            const audio = document.getElementById(AUDIO_ELEMENT_FOR_TEST_CONVERTED) as HTMLAudioElement;
-            audio.pause()
+            audio.pause();
-            audio.srcObject = null
+            audio.srcObject = null;
-            audio.src = url
+            audio.src = url;
-            await audio.play()
+            await audio.play();
             if (!audioSrcNode.current) {
                 audioSrcNode.current = audioContext!.createMediaElementSource(audio);
             }
@@ -169,27 +219,27 @@ export const DeviceArea = (_props: DeviceAreaProps) => {
                 audioSrcNode.current = audioContext!.createMediaElementSource(audio);
             }
 
-            const dst = audioContext.createMediaStreamDestination()
+            const dst = audioContext.createMediaStreamDestination();
-            audioSrcNode.current.connect(dst)
+            audioSrcNode.current.connect(dst);
             try {
-                setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: dst.stream })
+                setVoiceChangerClientSetting({ ...setting.voiceChangerClientSetting, audioInput: dst.stream });
             } catch (e) {
-                console.error(e)
+                console.error(e);
             }
 
-            const audio_echo = document.getElementById(AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK) as HTMLAudioElement
+            const audio_echo = document.getElementById(AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK) as HTMLAudioElement;
-            audio_echo.srcObject = dst.stream
+            audio_echo.srcObject = dst.stream;
-            audio_echo.play()
+            audio_echo.play();
-            audio_echo.volume = 0
+            audio_echo.volume = 0;
-            setFileInputEchoback(false)
+            setFileInputEchoback(false);
 
             // original stream to play.
-            const audio_org = document.getElementById(AUDIO_ELEMENT_FOR_TEST_ORIGINAL) as HTMLAudioElement
+            const audio_org = document.getElementById(AUDIO_ELEMENT_FOR_TEST_ORIGINAL) as HTMLAudioElement;
-            audio_org.src = url
+            audio_org.src = url;
-            audio_org.pause()
+            audio_org.pause();
-        }
+        };
 
-        const echobackClass = fileInputEchoback ? "config-sub-area-control-field-wav-file-echoback-button-active" : "config-sub-area-control-field-wav-file-echoback-button"
+        const echobackClass = fileInputEchoback ? "config-sub-area-control-field-wav-file-echoback-button-active" : "config-sub-area-control-field-wav-file-echoback-button";
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-field">
@@ -202,190 +252,332 @@ export const DeviceArea = (_props: DeviceAreaProps) => {
                         <div>
                             <img className="config-sub-area-control-field-wav-file-folder" src="./assets/icons/folder.svg" onClick={onFileLoadClicked} />
                         </div>
-                        <div className={echobackClass} onClick={() => { setFileInputEchoback(!fileInputEchoback) }}>
+                        <div
+                            className={echobackClass}
+                            onClick={() => {
+                                setFileInputEchoback(!fileInputEchoback);
+                            }}
+                        >
                             echo{fileInputEchoback}
                         </div>
                     </div>
                 </div>
             </div>
-        )
+        );
-    }, [audioInputForGUI, fileInputEchoback, serverSetting.serverSetting])
+    }, [audioInputForGUI, fileInputEchoback, serverSetting.serverSetting]);
-
-
-
 
     // (3) Audio Output
     useEffect(() => {
         const loadCache = async () => {
-            const key = await getItem(INDEXEDDB_KEY_AUDIO_OUTPUT)
+            const key = await getItem(INDEXEDDB_KEY_AUDIO_OUTPUT);
             if (key) {
-                setAudioOutputForGUI(key as string)
+                setAudioOutputForGUI(key as string);
             }
-        }
+        };
-        loadCache()
+        loadCache();
-    }, [])
+    }, []);
 
     useEffect(() => {
         const setAudioOutput = async () => {
             const mediaDeviceInfos = await navigator.mediaDevices.enumerateDevices();
 
-            [AUDIO_ELEMENT_FOR_PLAY_RESULT, AUDIO_ELEMENT_FOR_TEST_ORIGINAL, AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK].forEach(x => {
+            [AUDIO_ELEMENT_FOR_PLAY_RESULT, AUDIO_ELEMENT_FOR_TEST_ORIGINAL, AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK].forEach((x) => {
-                const audio = document.getElementById(x) as HTMLAudioElement
+                const audio = document.getElementById(x) as HTMLAudioElement;
                 if (audio) {
                     if (serverSetting.serverSetting.enableServerAudio == 1) {
-
                         // Server Audio を使う場合はElementから音は出さない。
-                        audio.volume = 0
+                        audio.volume = 0;
                     } else if (audioOutputForGUI == "none") {
                         // @ts-ignore
-                        audio.setSinkId("")
+                        audio.setSinkId("");
                         if (x == AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK) {
-                            audio.volume = 0
+                            audio.volume = 0;
                         } else {
-                            audio.volume = 0
+                            audio.volume = 0;
                         }
                     } else {
-                        const audioOutputs = mediaDeviceInfos.filter(x => { return x.kind == "audiooutput" })
+                        const audioOutputs = mediaDeviceInfos.filter((x) => {
-                        const found = audioOutputs.some(x => { return x.deviceId == audioOutputForGUI })
+                            return x.kind == "audiooutput";
+                        });
+                        const found = audioOutputs.some((x) => {
+                            return x.deviceId == audioOutputForGUI;
+                        });
                         if (found) {
                             // @ts-ignore // 例外キャッチできないので事前にIDチェックが必要らしい。！？
-                            audio.setSinkId(audioOutputForGUI)
+                            audio.setSinkId(audioOutputForGUI);
                         } else {
-                            console.warn("No audio output device. use default")
+                            console.warn("No audio output device. use default");
                         }
 
                         if (x == AUDIO_ELEMENT_FOR_TEST_CONVERTED_ECHOBACK) {
-                            audio.volume = fileInputEchoback ? 1 : 0
+                            audio.volume = fileInputEchoback ? 1 : 0;
                         } else {
-                            audio.volume = 1
+                            audio.volume = 1;
                         }
                     }
                 }
-            })
+            });
-        }
+        };
-        setAudioOutput()
+        setAudioOutput();
-    }, [audioOutputForGUI, fileInputEchoback, serverSetting.serverSetting.enableServerAudio])
+    }, [audioOutputForGUI, fileInputEchoback, serverSetting.serverSetting.enableServerAudio]);
-
 
     // (3-1) クライアント
     const clientAudioOutputRow = useMemo(() => {
         if (serverSetting.serverSetting.enableServerAudio == 1) {
-            return <></>
+            return <></>;
         }
 
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-title left-padding-1">output</div>
                 <div className="config-sub-area-control-field">
-                    <select className="body-select" value={audioOutputForGUI} onChange={(e) => {
+                    <select
-                        setAudioOutputForGUI(e.target.value)
+                        className="body-select"
-                        setItem(INDEXEDDB_KEY_AUDIO_OUTPUT, e.target.value)
+                        value={audioOutputForGUI}
-                    }}>
+                        onChange={(e) => {
-                        {
+                            setAudioOutputForGUI(e.target.value);
-                            outputAudioDeviceInfo.map(x => {
+                            setItem(INDEXEDDB_KEY_AUDIO_OUTPUT, e.target.value);
-                                return <option key={x.deviceId} value={x.deviceId}>{x.label}</option>
+                        }}
-                            })
+                    >
-                        }
+                        {outputAudioDeviceInfo.map((x) => {
+                            return (
+                                <option key={x.deviceId} value={x.deviceId}>
+                                    {x.label}
+                                </option>
+                            );
+                        })}
                     </select>
                 </div>
             </div>
-        )
+        );
-    }, [serverSetting.serverSetting.enableServerAudio, outputAudioDeviceInfo, audioOutputForGUI])
+    }, [serverSetting.serverSetting.enableServerAudio, outputAudioDeviceInfo, audioOutputForGUI]);
 
     useEffect(() => {
-        console.log("initializedRef.current", initializedRef.current)
+        console.log("initializedRef.current", initializedRef.current);
-        setAudioOutputElementId(AUDIO_ELEMENT_FOR_PLAY_RESULT)
+        setAudioOutputElementId(AUDIO_ELEMENT_FOR_PLAY_RESULT);
-    }, [initializedRef.current])
+    }, [initializedRef.current]);
 
     // (3-2) サーバ
     const serverAudioOutputRow = useMemo(() => {
         if (serverSetting.serverSetting.enableServerAudio == 0) {
-            return <></>
+            return <></>;
         }
-        const devices = serverSetting.serverSetting.serverAudioOutputDevices
+        const devices = serverSetting.serverSetting.serverAudioOutputDevices;
-        const hostAPIs = new Set(devices.map(x => { return x.hostAPI }))
+        const hostAPIs = new Set(
-        const hostAPIOptions = Array.from(hostAPIs).map((x, index) => { return <option value={x} key={index} >{x}</option> })
+            devices.map((x) => {
+                return x.hostAPI;
+            })
+        );
+        const hostAPIOptions = Array.from(hostAPIs).map((x, index) => {
+            return (
+                <option value={x} key={index}>
+                    {x}
+                </option>
+            );
+        });
 
-        const filteredDevice = devices.map((x, index) => {
+        const filteredDevice = devices
+            .map((x, index) => {
                 if (outputHostApi != "ALL" && x.hostAPI != outputHostApi) {
-                return null
+                    return null;
                 }
-            return <option value={x.index} key={index}>[{x.hostAPI}]{x.name}</option>
+                return (
-        }).filter(x => x != null)
+                    <option value={x.index} key={index}>
+                        [{x.hostAPI}]{x.name}
+                    </option>
+                );
+            })
+            .filter((x) => x != null);
 
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-title left-padding-1">output</div>
                 <div className="config-sub-area-control-field">
                     <div className="config-sub-area-control-field-auido-io">
-                        <select className="config-sub-area-control-field-auido-io-filter" name="kinds" id="kinds" value={outputHostApi} onChange={(e) => { setOutputHostApi(e.target.value) }}>
+                        <select
-                            <option value="ALL" key="ALL" >ALL</option>
+                            className="config-sub-area-control-field-auido-io-filter"
+                            name="kinds"
+                            id="kinds"
+                            value={outputHostApi}
+                            onChange={(e) => {
+                                setOutputHostApi(e.target.value);
+                            }}
+                        >
+                            <option value="ALL" key="ALL">
+                                ALL
+                            </option>
                             {hostAPIOptions}
                         </select>
-                        <select className="config-sub-area-control-field-auido-io-select" value={serverSetting.serverSetting.serverOutputDeviceId} onChange={(e) => {
+                        <select
-                            serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverOutputDeviceId: Number(e.target.value) })
+                            className="config-sub-area-control-field-auido-io-select"
-                        }}>
+                            value={serverSetting.serverSetting.serverOutputDeviceId}
+                            onChange={(e) => {
+                                serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverOutputDeviceId: Number(e.target.value) });
+                            }}
+                        >
                             {filteredDevice}
                         </select>
                     </div>
                 </div>
             </div>
-        )
+        );
-    }, [outputHostApi, serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio])
+    }, [outputHostApi, serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio]);
-
 
     // (4) レコーダー
     const outputRecorderRow = useMemo(() => {
         if (serverSetting.serverSetting.enableServerAudio == 1) {
-            return <></>
+            return <></>;
         }
         const onOutputRecordStartClicked = async () => {
-            setOutputRecordingStarted(true)
+            setOutputRecordingStarted(true);
-            await startOutputRecording()
+            await startOutputRecording();
-        }
+        };
         const onOutputRecordStopClicked = async () => {
-            setOutputRecordingStarted(false)
+            setOutputRecordingStarted(false);
-            const record = await stopOutputRecording()
+            const record = await stopOutputRecording();
-            downloadRecord(record)
+            downloadRecord(record);
-        }
+        };
 
-        const startClassName = outputRecordingStarted ? "config-sub-area-button-active" : "config-sub-area-button"
+        const startClassName = outputRecordingStarted ? "config-sub-area-button-active" : "config-sub-area-button";
-        const stopClassName = outputRecordingStarted ? "config-sub-area-button" : "config-sub-area-button-active"
+        const stopClassName = outputRecordingStarted ? "config-sub-area-button" : "config-sub-area-button-active";
         return (
             <div className="config-sub-area-control">
                 <div className="config-sub-area-control-title">REC.</div>
                 <div className="config-sub-area-control-field">
                     <div className="config-sub-area-buttons">
-                        <div onClick={onOutputRecordStartClicked} className={startClassName}>start</div>
+                        <div onClick={onOutputRecordStartClicked} className={startClassName}>
-                        <div onClick={onOutputRecordStopClicked} className={stopClassName}>stop</div>
+                            start
+                        </div>
+                        <div onClick={onOutputRecordStopClicked} className={stopClassName}>
+                            stop
                         </div>
                     </div>
                 </div>
-        )
+            </div>
+        );
+    }, [outputRecordingStarted, startOutputRecording, stopOutputRecording, serverSetting.serverSetting.enableServerAudio]);
 
-    }, [outputRecordingStarted, startOutputRecording, stopOutputRecording])
+    // (5) サンプリングレート
+    const sampleRateRow = useMemo(() => {
+        if (serverSetting.serverSetting.enableServerAudio == 0) {
+            return <></>;
+        }
 
+        return (
+            <div className="config-sub-area-control">
+                <div className="config-sub-area-control-title left-padding-1">S.R.</div>
+                <div className="config-sub-area-control-field">
+                    <div className="config-sub-area-control-field-auido-io">
+                        <select
+                            className="config-sub-area-control-field-sample-rate-select"
+                            value={serverSetting.serverSetting.serverAudioSampleRate}
+                            onChange={(e) => {
+                                serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverAudioSampleRate: Number(e.target.value) });
+                            }}
+                        >
+                            {[16000, 32000, 44100, 48000, 96000, 192000].map((x) => {
+                                return (
+                                    <option key={x} value={x}>
+                                        {x}
+                                    </option>
+                                );
+                            })}
+                        </select>
+                    </div>
+                </div>
+            </div>
+        );
+    }, [serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio]);
 
+    // (6) Monitor
+    const serverMonitorRow = useMemo(() => {
+        if (serverSetting.serverSetting.enableServerAudio == 0) {
+            return <></>;
+        }
+        const devices = serverSetting.serverSetting.serverAudioOutputDevices;
+        const hostAPIs = new Set(
+            devices.map((x) => {
+                return x.hostAPI;
+            })
+        );
+        const hostAPIOptions = Array.from(hostAPIs).map((x, index) => {
+            return (
+                <option value={x} key={index}>
+                    {x}
+                </option>
+            );
+        });
+
+        const filteredDevice = devices
+            .map((x, index) => {
+                if (outputHostApi != "ALL" && x.hostAPI != outputHostApi) {
+                    return null;
+                }
+                return (
+                    <option value={x.index} key={index}>
+                        [{x.hostAPI}]{x.name}
+                    </option>
+                );
+            })
+            .filter((x) => x != null);
+        filteredDevice.unshift(
+            <option value={-1} key={-1}>
+                None
+            </option>
+        );
+
+        return (
+            <div className="config-sub-area-control">
+                <div className="config-sub-area-control-title left-padding-1">monitor</div>
+                <div className="config-sub-area-control-field">
+                    <div className="config-sub-area-control-field-auido-io">
+                        <select
+                            className="config-sub-area-control-field-auido-io-filter"
+                            name="kinds"
+                            id="kinds"
+                            value={outputHostApi}
+                            onChange={(e) => {
+                                setOutputHostApi(e.target.value);
+                            }}
+                        >
+                            <option value="ALL" key="ALL">
+                                ALL
+                            </option>
+                            {hostAPIOptions}
+                        </select>
+                        <select
+                            className="config-sub-area-control-field-auido-io-select"
+                            value={serverSetting.serverSetting.serverMonitorDeviceId}
+                            onChange={(e) => {
+                                serverSetting.updateServerSettings({ ...serverSetting.serverSetting, serverMonitorDeviceId: Number(e.target.value) });
+                            }}
+                        >
+                            {filteredDevice}
+                        </select>
+                    </div>
+                </div>
+            </div>
+        );
+    }, [outputHostApi, serverSetting.serverSetting, serverSetting.updateServerSettings, serverSetting.serverSetting.enableServerAudio]);
 
     return (
         <div className="config-sub-area">
             {deviceModeRow}
+            {sampleRateRow}
             {clientAudioInputRow}
             {serverAudioInputRow}
             {audioInputMediaRow}
             {clientAudioOutputRow}
             {serverAudioOutputRow}
+            {serverMonitorRow}
 
             {outputRecorderRow}
             <audio hidden id={AUDIO_ELEMENT_FOR_PLAY_RESULT}></audio>
         </div>
-    )
+    );
-}
+};
-
-
 
 const downloadRecord = (data: Float32Array) => {
-
     const writeString = (view: DataView, offset: number, string: string) => {
         for (var i = 0; i < string.length; i++) {
             view.setUint8(offset + i, string.charCodeAt(i));
@@ -395,7 +587,7 @@ const downloadRecord = (data: Float32Array) => {
     const floatTo16BitPCM = (output: DataView, offset: number, input: Float32Array) => {
         for (var i = 0; i < input.length; i++, offset += 2) {
             var s = Math.max(-1, Math.min(1, input[i]));
-            output.setInt16(offset, s < 0 ? s * 0x8000 : s * 0x7FFF, true);
+            output.setInt16(offset, s < 0 ? s * 0x8000 : s * 0x7fff, true);
         }
     };
 
@@ -403,10 +595,10 @@ const downloadRecord = (data: Float32Array) => {
     const view = new DataView(buffer);
 
     // https://www.youfit.co.jp/archives/1418
-    writeString(view, 0, 'RIFF');  // RIFFヘッダ
+    writeString(view, 0, "RIFF"); // RIFFヘッダ
     view.setUint32(4, 32 + data.length * 2, true); // これ以降のファイルサイズ
-    writeString(view, 8, 'WAVE'); // WAVEヘッダ
+    writeString(view, 8, "WAVE"); // WAVEヘッダ
-    writeString(view, 12, 'fmt '); // fmtチャンク
+    writeString(view, 12, "fmt "); // fmtチャンク
     view.setUint32(16, 16, true); // fmtチャンクのバイト数
     view.setUint16(20, 1, true); // フォーマットID
     view.setUint16(22, 1, true); // チャンネル数
@@ -414,10 +606,10 @@ const downloadRecord = (data: Float32Array) => {
     view.setUint32(28, 48000 * 2, true); // データ速度
     view.setUint16(32, 2, true); // ブロックサイズ
     view.setUint16(34, 16, true); // サンプルあたりのビット数
-    writeString(view, 36, 'data'); // dataチャンク
+    writeString(view, 36, "data"); // dataチャンク
     view.setUint32(40, data.length * 2, true); // 波形データのバイト数
     floatTo16BitPCM(view, 44, data); // 波形データ
-    const audioBlob = new Blob([view], { type: 'audio/wav' });
+    const audioBlob = new Blob([view], { type: "audio/wav" });
 
     const url = URL.createObjectURL(audioBlob);
     const a = document.createElement("a");
@@ -427,4 +619,4 @@ const downloadRecord = (data: Float32Array) => {
     a.click();
     document.body.removeChild(a);
     URL.revokeObjectURL(url);
-}
+};

client/lib/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@dannadori/voice-changer-client-js",
-  "version": "1.0.155",
+  "version": "1.0.157",
   "description": "",
   "main": "dist/index.js",
   "directories": {
@@ -27,10 +27,10 @@
   "license": "ISC",
   "devDependencies": {
     "@types/audioworklet": "^0.0.48",
-    "@types/node": "^20.3.2",
+    "@types/node": "^20.3.3",
     "@types/react": "18.2.14",
     "@types/react-dom": "18.2.6",
-    "eslint": "^8.43.0",
+    "eslint": "^8.44.0",
     "eslint-config-prettier": "^8.8.0",
     "eslint-plugin-prettier": "^4.2.1",
     "eslint-plugin-react": "^7.32.2",
@@ -47,7 +47,7 @@
   },
   "dependencies": {
     "@types/readable-stream": "^2.3.15",
-    "amazon-chime-sdk-js": "^3.14.1",
+    "amazon-chime-sdk-js": "^3.15.0",
     "buffer": "^6.0.3",
     "localforage": "^1.10.0",
     "react": "^18.2.0",

client/lib/src/const.ts

@@ -83,12 +83,15 @@ export const ServerSettingKey = {
 
     "enableServerAudio": "enableServerAudio",
     "serverAudioStated": "serverAudioStated",
+    "serverAudioSampleRate": "serverAudioSampleRate",
     "serverInputAudioSampleRate": "serverInputAudioSampleRate",
     "serverOutputAudioSampleRate": "serverOutputAudioSampleRate",
+    "serverMonitorAudioSampleRate": "serverMonitorAudioSampleRate",
     "serverInputAudioBufferSize": "serverInputAudioBufferSize",
     "serverOutputAudioBufferSize": "serverOutputAudioBufferSize",
     "serverInputDeviceId": "serverInputDeviceId",
     "serverOutputDeviceId": "serverOutputDeviceId",
+    "serverMonitorDeviceId": "serverMonitorDeviceId",
     "serverReadChunkSize": "serverReadChunkSize",
     "serverInputAudioGain": "serverInputAudioGain",
     "serverOutputAudioGain": "serverOutputAudioGain",
@@ -138,12 +141,15 @@ export type VoiceChangerServerSetting = {
 
     enableServerAudio: number // 0:off, 1:on
     serverAudioStated: number // 0:off, 1:on
+    serverAudioSampleRate: number
     serverInputAudioSampleRate: number
     serverOutputAudioSampleRate: number
+    serverMonitorAudioSampleRate: number
     serverInputAudioBufferSize: number
     serverOutputAudioBufferSize: number
     serverInputDeviceId: number
     serverOutputDeviceId: number
+    serverMonitorDeviceId: number
     serverReadChunkSize: number
     serverInputAudioGain: number
     serverOutputAudioGain: number
@@ -306,12 +312,15 @@ export const DefaultServerSetting: ServerInfo = {
 
     enableServerAudio: 0,
     serverAudioStated: 0,
+    serverAudioSampleRate: 48000,
     serverInputAudioSampleRate: 48000,
     serverOutputAudioSampleRate: 48000,
+    serverMonitorAudioSampleRate: 48000,
     serverInputAudioBufferSize: 1024 * 24,
     serverOutputAudioBufferSize: 1024 * 24,
     serverInputDeviceId: -1,
     serverOutputDeviceId: -1,
+    serverMonitorDeviceId: -1,
     serverReadChunkSize: 256,
     serverInputAudioGain: 1.0,
     serverOutputAudioGain: 1.0,

server/.vscode/settings.json

@@ -9,11 +9,9 @@
     "editor.formatOnSave": true // ファイル保存時に自動フォーマット
   },
   "python.formatting.blackArgs": ["--line-length", "550"],
-  "flake8.args": [
-    "--ignore=E501,E402,E722,E741,E203,W503"
-    // "--max-line-length=150",
-    // "--max-complexity=20"
-  ],
   "python.linting.flake8Enabled": true,
+  "python.linting.flake8Args": [
+    "--max-line-length=99999"
+  ],
   "python.linting.enabled": true
 }

server/MMVCServerSIO.py

@@ -31,7 +31,7 @@ setup_loggers()
 
 def setupArgParser():
     parser = argparse.ArgumentParser()
-    parser.add_argument("--logLevel", type=str, default="critical", help="Log level info|critical. (default: critical)")
+    parser.add_argument("--logLevel", type=str, default="error", help="Log level info|critical|error. (default: error)")
     parser.add_argument("-p", type=int, default=18888, help="port")
     parser.add_argument("--https", type=strtobool, default=False, help="use https")
     parser.add_argument("--httpsKey", type=str, default="ssl.key", help="path for the key of https")
@@ -94,6 +94,7 @@ PORT = args.p
 
 
 def localServer(logLevel: str = "critical"):
+    try:
         uvicorn.run(
             f"{os.path.basename(__file__)[:-3]}:app_socketio",
             host="0.0.0.0",
@@ -101,6 +102,8 @@ def localServer(logLevel: str = "critical"):
             reload=False if hasattr(sys, "_MEIPASS") else True,
             log_level=logLevel,
         )
+    except Exception as e:
+        print("[Voice Changer] Web Server Launch Exception", e)
 
 
 if __name__ == "MMVCServerSIO":
@@ -204,6 +207,7 @@ if __name__ == "__main__":
     # サーバ起動
     if args.https:
         # HTTPS サーバ起動
+        try:
             uvicorn.run(
                 f"{os.path.basename(__file__)[:-3]}:app_socketio",
                 host="0.0.0.0",
@@ -213,6 +217,9 @@ if __name__ == "__main__":
                 ssl_certfile=cert_path,
                 log_level=args.logLevel,
             )
+        except Exception as e:
+            print("[Voice Changer] Web Server Launch Exception", e)
+
     else:
         p = mp.Process(name="p", target=localServer, args=(args.logLevel,))
         p.start()

server/const.py

@@ -23,6 +23,8 @@ ModelType: TypeAlias = Literal[
 
 STORED_SETTING_FILE = "stored_setting.json"
 
+SERVER_DEVICE_SAMPLE_RATES = [16000, 32000, 44100, 48000, 96000, 192000]
+
 tmpdir = tempfile.TemporaryDirectory()
 SSL_KEY_DIR = os.path.join(tmpdir.name, "keys") if hasattr(sys, "_MEIPASS") else "keys"
 MODEL_DIR = os.path.join(tmpdir.name, "logs") if hasattr(sys, "_MEIPASS") else "logs"

server/restapi/MMVC_Rest_Fileuploader.py

@@ -76,7 +76,6 @@ class MMVC_Rest_Fileuploader:
         except Exception as e:
             print("[Voice Changer] post_update_settings ex:", e)
             import traceback
-
             traceback.print_exc()
 
     def post_load_model(

server/voice_changer/IORecorder.py

@@ -3,19 +3,19 @@ import os
 
 
 class IORecorder:
-    def __init__(self, inputFilename: str, outputFilename: str, samplingRate: int):
+    def __init__(self, inputFilename: str, outputFilename: str, inputSamplingRate: int, outputSamplingRate: int):
         self._clearFile(inputFilename)
         self._clearFile(outputFilename)
 
         self.fi = wave.open(inputFilename, "wb")
         self.fi.setnchannels(1)
         self.fi.setsampwidth(2)
-        self.fi.setframerate(samplingRate)
+        self.fi.setframerate(inputSamplingRate)
 
         self.fo = wave.open(outputFilename, "wb")
         self.fo.setnchannels(1)
         self.fo.setsampwidth(2)
-        self.fo.setframerate(samplingRate)
+        self.fo.setframerate(outputSamplingRate)
 
     def _clearFile(self, filename: str):
         if os.path.exists(filename):

server/voice_changer/Local/AudioDeviceList.py

@@ -1,6 +1,9 @@
 import sounddevice as sd
-from dataclasses import dataclass
+from dataclasses import dataclass, field
+
 from const import ServerAudioDeviceTypes
+import numpy as np
+# from const import SERVER_DEVICE_SAMPLE_RATES
 
 
 @dataclass
@@ -12,6 +15,40 @@ class ServerAudioDevice:
     maxInputChannels: int = 0
     maxOutputChannels: int = 0
     default_samplerate: int = 0
+    available_samplerates: list[int] = field(default_factory=lambda: [])
+
+
+def dummy_callback(data: np.ndarray, frames, times, status):
+    pass
+
+
+def checkSamplingRate(deviceId: int, desiredSamplingRate: int, type: ServerAudioDeviceTypes):
+    if type == "input":
+        try:
+            with sd.InputStream(
+                device=deviceId,
+                callback=dummy_callback,
+                dtype="float32",
+                samplerate=desiredSamplingRate
+            ):
+                pass
+            return True
+        except Exception as e: # NOQA
+            # print("[checkSamplingRate]", e)
+            return False
+    else:
+        try:
+            with sd.OutputStream(
+                device=deviceId,
+                callback=dummy_callback,
+                dtype="float32",
+                samplerate=desiredSamplingRate
+            ):
+                pass
+            return True
+        except Exception as e: # NOQA
+            # print("[checkSamplingRate]", e)
+            return False
 
 
 def list_audio_device():
@@ -20,7 +57,7 @@ def list_audio_device():
     except Exception as e:
         print("[Voice Changer] ex:query_devices")
         print(e)
-        return [], []
+        raise e
 
     inputAudioDeviceList = [d for d in audioDeviceList if d["max_input_channels"] > 0]
     outputAudioDeviceList = [d for d in audioDeviceList if d["max_output_channels"] > 0]
@@ -55,4 +92,20 @@ def list_audio_device():
         )
         serverAudioOutputDevices.append(serverOutputAudioDevice)
 
+    # print("check sample rate1")
+    # for d in serverAudioInputDevices:
+    #     print("check sample rate1-1")
+    #     for sr in SERVER_DEVICE_SAMPLE_RATES:
+    #         print("check sample rate1-2")
+    #         if checkSamplingRate(d.index, sr, "input"):
+    #             d.available_samplerates.append(sr)
+    # print("check sample rate2")
+    # for d in serverAudioOutputDevices:
+    #     print("check sample rate2-1")
+    #     for sr in SERVER_DEVICE_SAMPLE_RATES:
+    #         print("check sample rate2-2")
+    #         if checkSamplingRate(d.index, sr, "output"):
+    #             d.available_samplerates.append(sr)
+    # print("check sample rate3")
+
     return serverAudioInputDevices, serverAudioOutputDevices

server/voice_changer/Local/ServerDevice.py

@@ -1,8 +1,9 @@
 from dataclasses import dataclass, asdict
 
 import numpy as np
+from const import SERVER_DEVICE_SAMPLE_RATES
 
-from voice_changer.Local.AudioDeviceList import list_audio_device
+from voice_changer.Local.AudioDeviceList import checkSamplingRate, list_audio_device
 import time
 import sounddevice as sd
 from voice_changer.utils.Timer import Timer
@@ -10,6 +11,8 @@ import librosa
 
 from voice_changer.utils.VoiceChangerModel import AudioInOut
 from typing import Protocol
+from typing import Literal, TypeAlias
+AudioDeviceKind: TypeAlias = Literal["input", "output"]
 
 
 @dataclass
@@ -18,12 +21,21 @@ class ServerDeviceSettings:
     serverAudioStated: int = 0  # 0:off, 1:on
     serverInputAudioSampleRate: int = 44100
     serverOutputAudioSampleRate: int = 44100
+    serverMonitorAudioSampleRate: int = 44100
+
+    serverAudioSampleRate: int = 44100
+    # serverAudioSampleRate: int = 16000
+    # serverAudioSampleRate: int = 48000
+
     serverInputDeviceId: int = -1
     serverOutputDeviceId: int = -1
+    serverMonitorDeviceId: int = -1  # -1 でモニター無効
     serverReadChunkSize: int = 256
     serverInputAudioGain: float = 1.0
     serverOutputAudioGain: float = 1.0
 
+    exclusiveMode: bool = False
+
 
 EditableServerDeviceSettings = {
     "intData": [
@@ -31,14 +43,20 @@ EditableServerDeviceSettings = {
         "serverAudioStated",
         "serverInputAudioSampleRate",
         "serverOutputAudioSampleRate",
+        "serverMonitorAudioSampleRate",
+        "serverAudioSampleRate",
         "serverInputDeviceId",
         "serverOutputDeviceId",
+        "serverMonitorDeviceId",
         "serverReadChunkSize",
     ],
     "floatData": [
         "serverInputAudioGain",
         "serverOutputAudioGain",
     ],
+    "boolData": [
+        "exclusiveMode"
+    ]
 }
 
 
@@ -52,7 +70,10 @@ class ServerDeviceCallbacks(Protocol):
     def get_processing_sampling_rate(self):
         ...
 
-    def setSamplingRate(self, sr: int):
+    def setInputSamplingRate(self, sr: int):
+        ...
+
+    def setOutputSamplingRate(self, sr: int):
         ...
 
 
@@ -60,6 +81,10 @@ class ServerDevice:
     def __init__(self, serverDeviceCallbacks: ServerDeviceCallbacks):
         self.settings = ServerDeviceSettings()
         self.serverDeviceCallbacks = serverDeviceCallbacks
+        self.out_wav = None
+        self.mon_wav = None
+        self.serverAudioInputDevices = None
+        self.serverAudioOutputDevices = None
 
     def getServerInputAudioDevice(self, index: int):
         audioinput, _audiooutput = list_audio_device()
@@ -94,7 +119,221 @@ class ServerDevice:
         except Exception as e:
             print("[Voice Changer] ex:", e)
 
+    def audioInput_callback(self, indata: np.ndarray, frames, times, status):
+        try:
+            indata = indata * self.settings.serverInputAudioGain
+            with Timer("all_inference_time") as t:
+                unpackedData = librosa.to_mono(indata.T) * 32768.0
+                unpackedData = unpackedData.astype(np.int16)
+                out_wav, times = self.serverDeviceCallbacks.on_request(unpackedData)
+                self.out_wav = out_wav
+                self.mon_wav = out_wav
+            all_inference_time = t.secs
+            self.performance = [all_inference_time] + times
+            self.serverDeviceCallbacks.emitTo(self.performance)
+            self.performance = [round(x * 1000) for x in self.performance]
+        except Exception as e:
+            print("[Voice Changer][ServerDevice][audioInput_callback] ex:", e)
+            # import traceback
+            # traceback.print_exc()
+
+    def audioOutput_callback(self, outdata: np.ndarray, frames, times, status):
+        try:
+            if self.out_wav is None:
+                return
+            out_wav = self.out_wav
+            outputChannels = outdata.shape[1]
+            outdata[:] = np.repeat(out_wav, outputChannels).reshape(-1, outputChannels) / 32768.0
+            outdata[:] = outdata * self.settings.serverOutputAudioGain
+        except Exception as e:
+            print("[Voice Changer][ServerDevice][audioOutput_callback]  ex:", e)
+            # import traceback
+            # traceback.print_exc()
+
+    def audioMonitor_callback(self, outdata: np.ndarray, frames, times, status):
+        try:
+            if self.mon_wav is None:
+                return
+            mon_wav = self.mon_wav
+            outputChannels = outdata.shape[1]
+            outdata[:] = np.repeat(mon_wav, outputChannels).reshape(-1, outputChannels) / 32768.0
+            outdata[:] = outdata * self.settings.serverOutputAudioGain  # GainはOutputのものをを流用
+            # Monitorモードが有効の場合はサンプリングレートはmonitorデバイスが優先されているためリサンプリング不要
+        except Exception as e:
+            print("[Voice Changer][ServerDevice][audioMonitor_callback]  ex:", e)
+            # import traceback
+            # traceback.print_exc()
+
     def start(self):
+        currentModelSamplingRate = -1
+        while True:
+            if self.settings.serverAudioStated == 0 or self.settings.serverInputDeviceId == -1:
+                time.sleep(2)
+            else:
+                sd._terminate()
+                sd._initialize()
+
+                # Curret Device ID
+                currentServerInputDeviceId = self.settings.serverInputDeviceId
+                currentServerOutputDeviceId = self.settings.serverOutputDeviceId
+                currentServerMonitorDeviceId = self.settings.serverMonitorDeviceId
+
+                # Device 特定
+                serverInputAudioDevice = self.getServerInputAudioDevice(self.settings.serverInputDeviceId)
+                serverOutputAudioDevice = self.getServerOutputAudioDevice(self.settings.serverOutputDeviceId)
+                serverMonitorAudioDevice = None
+                if self.settings.serverMonitorDeviceId != -1:
+                    serverMonitorAudioDevice = self.getServerOutputAudioDevice(self.settings.serverMonitorDeviceId)
+
+                # Generate ExtraSetting
+                inputExtraSetting = None
+                outputExtraSetting = None
+                if self.settings.exclusiveMode:
+                    if "WASAPI" in serverInputAudioDevice.hostAPI:
+                        inputExtraSetting = sd.WasapiSettings(exclusive=True)
+                    if "WASAPI" in serverOutputAudioDevice.hostAPI:
+                        outputExtraSetting = sd.WasapiSettings(exclusive=True)
+                monitorExtraSetting = None
+                if self.settings.exclusiveMode and serverMonitorAudioDevice is not None:
+                    if "WASAPI" in serverMonitorAudioDevice.hostAPI:
+                        monitorExtraSetting = sd.WasapiSettings(exclusive=True)
+
+                print("Devices:")
+                print("  [Input]:", serverInputAudioDevice, inputExtraSetting)
+                print("  [Output]:", serverOutputAudioDevice, outputExtraSetting)
+                print("  [Monitor]:", serverMonitorAudioDevice, monitorExtraSetting)
+
+                # Deviceがなかったらいったんスリープ
+                if serverInputAudioDevice is None or serverOutputAudioDevice is None:
+                    print("serverInputAudioDevice or serverOutputAudioDevice is None")
+                    time.sleep(2)
+                    continue
+
+                # サンプリングレート
+                # 同一サンプリングレートに統一（変換時にサンプルが不足する場合があるため。パディング方法が明らかになれば、それぞれ設定できるかも）
+                currentAudioSampleRate = self.settings.serverAudioSampleRate
+                try:
+                    currentModelSamplingRate = self.serverDeviceCallbacks.get_processing_sampling_rate()
+                except Exception as e:
+                    print("[Voice Changer] ex: get_processing_sampling_rate", e)
+                    time.sleep(2)
+                    continue
+
+                self.settings.serverInputAudioSampleRate = currentAudioSampleRate
+                self.settings.serverOutputAudioSampleRate = currentAudioSampleRate
+                self.settings.serverMonitorAudioSampleRate = currentAudioSampleRate
+
+                # Sample Rate Check
+                inputAudioSampleRateAvailable = checkSamplingRate(self.settings.serverInputDeviceId, self.settings.serverInputAudioSampleRate, "input")
+                outputAudioSampleRateAvailable = checkSamplingRate(self.settings.serverOutputDeviceId, self.settings.serverOutputAudioSampleRate, "output")
+                monitorAudioSampleRateAvailable = checkSamplingRate(self.settings.serverMonitorDeviceId, self.settings.serverMonitorAudioSampleRate, "output") if serverMonitorAudioDevice else True
+
+                print("Sample Rate:")
+                print(f"  [Model]: {currentModelSamplingRate}")
+                print(f"  [Input]: {self.settings.serverInputAudioSampleRate} -> {inputAudioSampleRateAvailable}")
+                print(f"  [Output]: {self.settings.serverOutputAudioSampleRate} -> {outputAudioSampleRateAvailable}")
+                if serverMonitorAudioDevice is not None:
+                    print(f"  [Monitor]: {self.settings.serverMonitorAudioSampleRate} -> {monitorAudioSampleRateAvailable}")
+
+                if inputAudioSampleRateAvailable and outputAudioSampleRateAvailable and monitorAudioSampleRateAvailable:
+                    pass
+                else:
+                    print("Sample Rate is not supported by device:")
+                    print("Checking Available Sample Rate:")
+                    availableInputSampleRate = []
+                    availableOutputSampleRate = []
+                    availableMonitorSampleRate = []
+                    for sr in SERVER_DEVICE_SAMPLE_RATES:
+                        if checkSamplingRate(self.settings.serverInputDeviceId, sr, "input"):
+                            availableInputSampleRate.append(sr)
+                        if checkSamplingRate(self.settings.serverOutputDeviceId, sr, "output"):
+                            availableOutputSampleRate.append(sr)
+                        if serverMonitorAudioDevice is not None:
+                            if checkSamplingRate(self.settings.serverMonitorDeviceId, sr, "output"):
+                                availableMonitorSampleRate.append(sr)
+                    print("Available Sample Rate:")
+                    print(f"  [Input]: {availableInputSampleRate}")
+                    print(f"  [Output]: {availableOutputSampleRate}")
+                    if serverMonitorAudioDevice is not None:
+                        print(f"  [Monitor]: {availableMonitorSampleRate}")
+
+                    print("continue... ")
+                    time.sleep(2)
+                    continue
+
+                self.serverDeviceCallbacks.setInputSamplingRate(self.settings.serverInputAudioSampleRate)
+                self.serverDeviceCallbacks.setOutputSamplingRate(self.settings.serverOutputAudioSampleRate)
+
+                # Blockサイズを計算
+                currentInputChunkNum = self.settings.serverReadChunkSize
+                block_frame = currentInputChunkNum * 128
+                sd.default.blocksize = block_frame
+
+                # main loop
+                try:
+                    with sd.InputStream(
+                        callback=self.audioInput_callback,
+                        dtype="float32",
+                        device=self.settings.serverInputDeviceId,
+                        blocksize=block_frame,
+                        samplerate=self.settings.serverInputAudioSampleRate,
+                        channels=serverInputAudioDevice.maxInputChannels,
+                        extra_settings=inputExtraSetting
+                    ):
+                        with sd.OutputStream(
+                            callback=self.audioOutput_callback,
+                            dtype="float32",
+                            device=self.settings.serverOutputDeviceId,
+                            blocksize=block_frame,
+                            samplerate=self.settings.serverOutputAudioSampleRate,
+                            channels=serverOutputAudioDevice.maxOutputChannels,
+                            extra_settings=outputExtraSetting
+                        ):
+                            if self.settings.serverMonitorDeviceId != -1:
+                                with sd.OutputStream(
+                                    callback=self.audioMonitor_callback,
+                                    dtype="float32",
+                                    device=self.settings.serverMonitorDeviceId,
+                                    blocksize=block_frame,
+                                    samplerate=self.settings.serverMonitorAudioSampleRate,
+                                    channels=serverMonitorAudioDevice.maxOutputChannels,
+                                    extra_settings=monitorExtraSetting
+                                ):
+                                    while (
+                                        self.settings.serverAudioStated == 1 and
+                                        currentServerInputDeviceId == self.settings.serverInputDeviceId and
+                                        currentServerOutputDeviceId == self.settings.serverOutputDeviceId and
+                                        currentServerMonitorDeviceId == self.settings.serverMonitorDeviceId and
+                                        currentModelSamplingRate == self.serverDeviceCallbacks.get_processing_sampling_rate() and
+                                        currentInputChunkNum == self.settings.serverReadChunkSize and
+                                        currentAudioSampleRate == self.settings.serverAudioSampleRate
+                                    ):
+                                        time.sleep(2)
+                                        print(f"[Voice Changer] server audio performance {self.performance}")
+                                        print(f"                status: started:{self.settings.serverAudioStated}, model_sr:{currentModelSamplingRate}, chunk:{currentInputChunkNum}")
+                                        print(f"                input  : id:{self.settings.serverInputDeviceId}, sr:{self.settings.serverInputAudioSampleRate}, ch:{serverInputAudioDevice.maxInputChannels}")
+                                        print(f"                output : id:{self.settings.serverOutputDeviceId}, sr:{self.settings.serverOutputAudioSampleRate}, ch:{serverOutputAudioDevice.maxOutputChannels}")
+                                        print(f"                monitor: id:{self.settings.serverMonitorDeviceId}, sr:{self.settings.serverMonitorAudioSampleRate}, ch:{serverMonitorAudioDevice.maxOutputChannels}")
+                            else:
+                                while (
+                                    self.settings.serverAudioStated == 1 and
+                                    currentServerInputDeviceId == self.settings.serverInputDeviceId and
+                                    currentServerOutputDeviceId == self.settings.serverOutputDeviceId and
+                                    currentServerMonitorDeviceId == self.settings.serverMonitorDeviceId and
+                                    currentModelSamplingRate == self.serverDeviceCallbacks.get_processing_sampling_rate() and
+                                    currentInputChunkNum == self.settings.serverReadChunkSize and
+                                    currentAudioSampleRate == self.settings.serverAudioSampleRate
+                                ):
+                                    time.sleep(2)
+                                    print(f"[Voice Changer] server audio performance {self.performance}")
+                                    print(f"                status: started:{self.settings.serverAudioStated}, model_sr:{currentModelSamplingRate}, chunk:{currentInputChunkNum}]")
+                                    print(f"                input  : id:{self.settings.serverInputDeviceId}, sr:{self.settings.serverInputAudioSampleRate}, ch:{serverInputAudioDevice.maxInputChannels}")
+                                    print(f"                output : id:{self.settings.serverOutputDeviceId}, sr:{self.settings.serverOutputAudioSampleRate}, ch:{serverOutputAudioDevice.maxOutputChannels}")
+                except Exception as e:
+                    print("[Voice Changer] processing, ex:", e)
+                    time.sleep(2)
+
+    def start2(self):
         # currentInputDeviceId = -1
         # currentOutputDeviceId = -1
         # currentInputChunkNum = -1
@@ -140,13 +379,15 @@ class ServerDevice:
                     ):
                         pass
                     self.settings.serverInputAudioSampleRate = currentModelSamplingRate
-                    self.serverDeviceCallbacks.setSamplingRate(currentModelSamplingRate)
+                    self.serverDeviceCallbacks.setInputSamplingRate(currentModelSamplingRate)
+                    self.serverDeviceCallbacks.setOutputSamplingRate(currentModelSamplingRate)
                     print(f"[Voice Changer] sample rate {self.settings.serverInputAudioSampleRate}")
                 except Exception as e:
                     print("[Voice Changer] ex: fallback to device default samplerate", e)
                     print("[Voice Changer] device default samplerate", serverInputAudioDevice.default_samplerate)
                     self.settings.serverInputAudioSampleRate = round(serverInputAudioDevice.default_samplerate)
-                    self.serverDeviceCallbacks.setSamplingRate(round(serverInputAudioDevice.default_samplerate))
+                    self.serverDeviceCallbacks.setInputSamplingRate(round(serverInputAudioDevice.default_samplerate))
+                    self.serverDeviceCallbacks.setOutputSamplingRate(round(serverInputAudioDevice.default_samplerate))
 
                 sd.default.samplerate = self.settings.serverInputAudioSampleRate
                 sd.default.blocksize = block_frame
@@ -171,10 +412,15 @@ class ServerDevice:
 
     def get_info(self):
         data = asdict(self.settings)
+        try:
             audioinput, audiooutput = list_audio_device()
-        data["serverAudioInputDevices"] = audioinput
+            self.serverAudioInputDevices = audioinput
-        data["serverAudioOutputDevices"] = audiooutput
+            self.serverAudioOutputDevices = audiooutput
+        except Exception as e:
+            print(e)
 
+        data["serverAudioInputDevices"] = self.serverAudioInputDevices
+        data["serverAudioOutputDevices"] = self.serverAudioOutputDevices
         return data
 
     def update_settings(self, key: str, val: str | int | float):

server/voice_changer/RVC/deviceManager/DeviceManager.py

@@ -77,7 +77,7 @@ class DeviceManager(object):
     def getDeviceMemory(self, id: int):
         try:
             return torch.cuda.get_device_properties(id).total_memory
-            # except Exception as e:
+        except Exception as e:
-        except:
+            # except:
-            # print(e)
+            print(e)
             return 0

server/voice_changer/VoiceChanger.py

@@ -13,7 +13,7 @@ from voice_changer.IORecorder import IORecorder
 from voice_changer.utils.LoadModelParams import LoadModelParams
 
 from voice_changer.utils.Timer import Timer
-from voice_changer.utils.VoiceChangerModel import AudioInOut
+from voice_changer.utils.VoiceChangerModel import AudioInOut, VoiceChangerModel
 from Exceptions import (
     DeviceCannotSupportHalfPrecisionException,
     DeviceChangingException,
@@ -32,6 +32,7 @@ STREAM_OUTPUT_FILE = os.path.join(TMP_DIR, "out.wav")
 @dataclass
 class VoiceChangerSettings:
     inputSampleRate: int = 48000  # 48000 or 24000
+    outputSampleRate: int = 48000  # 48000 or 24000
 
     crossFadeOffsetRate: float = 0.1
     crossFadeEndRate: float = 0.9
@@ -71,7 +72,7 @@ class VoiceChanger:
         self.currentCrossFadeOverlapSize = 0  # setting
         self.crossfadeSize = 0  # calculated
 
-        self.voiceChanger = None
+        self.voiceChanger: VoiceChangerModel | None = None
         self.modelType: ModelType | None = None
         self.params = params
         self.gpu_num = torch.cuda.device_count()
@@ -115,6 +116,7 @@ class VoiceChanger:
 
         if key == "serverAudioStated" and val == 0:
             self.settings.inputSampleRate = 48000
+            self.settings.outputSampleRate = 48000
 
         if key in self.settings.intData:
             setattr(self.settings, key, int(val))
@@ -123,7 +125,7 @@ class VoiceChanger:
             if key == "recordIO" and val == 1:
                 if hasattr(self, "ioRecorder"):
                     self.ioRecorder.close()
-                self.ioRecorder = IORecorder(STREAM_INPUT_FILE, STREAM_OUTPUT_FILE, self.settings.inputSampleRate)
+                self.ioRecorder = IORecorder(STREAM_INPUT_FILE, STREAM_OUTPUT_FILE, self.settings.inputSampleRate, self.settings.outputSampleRate)
             if key == "recordIO" and val == 0:
                 if hasattr(self, "ioRecorder"):
                     self.ioRecorder.close()
@@ -268,10 +270,10 @@ class VoiceChanger:
             # 後処理
             with Timer("post-process") as t:
                 result = result.astype(np.int16)
-                if self.settings.inputSampleRate != processing_sampling_rate:
+                if self.settings.outputSampleRate != processing_sampling_rate:
                     # print(
-                    #     "samplingrate",
+                    #     "output samplingrate",
-                    #     self.settings.inputSampleRate,
+                    #     self.settings.outputSampleRate,
                     #     processing_sampling_rate,
                     # )
                     outputData = cast(
@@ -279,13 +281,13 @@ class VoiceChanger:
                         resampy.resample(
                             result,
                             processing_sampling_rate,
-                            self.settings.inputSampleRate,
+                            self.settings.outputSampleRate,
                         ).astype(np.int16),
                     )
                 else:
                     outputData = result
 
-                print_convert_processing(f" Output data size of {result.shape[0]}/{processing_sampling_rate}hz {outputData.shape[0]}/{self.settings.inputSampleRate}hz")
+                print_convert_processing(f" Output data size of {result.shape[0]}/{processing_sampling_rate}hz {outputData.shape[0]}/{self.settings.outputSampleRate}hz")
 
                 if receivedData.shape[0] != outputData.shape[0]:
                     # print(

server/voice_changer/VoiceChangerManager.py

@@ -50,9 +50,12 @@ class VoiceChangerManager(ServerDeviceCallbacks):
     def get_processing_sampling_rate(self):
         return self.voiceChanger.get_processing_sampling_rate()
 
-    def setSamplingRate(self, sr: int):
+    def setInputSamplingRate(self, sr: int):
         self.voiceChanger.settings.inputSampleRate = sr
 
+    def setOutputSamplingRate(self, sr: int):
+        self.voiceChanger.settings.outputSampleRate = sr
+
     ############################
     # VoiceChangerManager
     ############################
@@ -78,7 +81,7 @@ class VoiceChangerManager(ServerDeviceCallbacks):
             self.update_settings(key, val)
 
     def store_setting(self, key: str, val: str | int | float):
-        saveItemForServerDevice = ["enableServerAudio", "serverInputDeviceId", "serverOutputDeviceId", "serverReadChunkSize", "serverInputAudioGain", "serverOutputAudioGain"]
+        saveItemForServerDevice = ["enableServerAudio", "serverAudioSampleRate", "serverInputDeviceId", "serverOutputDeviceId", "serverMonitorDeviceId", "serverReadChunkSize", "serverInputAudioGain", "serverOutputAudioGain"]
         saveItemForVoiceChanger = ["crossFadeOffsetRate", "crossFadeEndRate", "crossFadeOverlapSize"]
         saveItemForVoiceChangerManager = ["modelSlotIndex"]
         saveItemForRVC = ["extraConvertSize", "gpu", "silentThreshold"]