优化 TTS 新增 proxy 配置，新增语调配置

app/models/schema.py

@@ -347,6 +347,7 @@ class VideoClipParams(BaseModel):
     voice_name: Optional[str] = Field(default="zh-CN-YunjianNeural", description="语音名称")
     voice_volume: Optional[float] = Field(default=1.0, description="语音音量")
     voice_rate: Optional[float] = Field(default=1.0, description="语速")
+    voice_pitch: Optional[float] = Field(default=1.0, description="语调")
 
     bgm_name: Optional[str] = Field(default="random", description="背景音乐名称")
     bgm_type: Optional[str] = Field(default="random", description="背景音乐类型")

app/services/voice.py

@@ -1032,11 +1032,11 @@ def is_azure_v2_voice(voice_name: str):
 
 
 def tts(
-    text: str, voice_name: str, voice_rate: float, voice_file: str
+    text: str, voice_name: str, voice_rate: float, voice_pitch: float, voice_file: str
 ) -> [SubMaker, None]:
     # if is_azure_v2_voice(voice_name):
     #     return azure_tts_v2(text, voice_name, voice_file)
-    return azure_tts_v1(text, voice_name, voice_rate, voice_file)
+    return azure_tts_v1(text, voice_name, voice_rate, voice_pitch, voice_file)
 
 
 def convert_rate_to_percent(rate: float) -> str:
@@ -1049,18 +1049,29 @@ def convert_rate_to_percent(rate: float) -> str:
         return f"{percent}%"
 
 
+def convert_pitch_to_percent(rate: float) -> str:
+    if rate == 1.0:
+        return "+0Hz"
+    percent = round((rate - 1.0) * 100)
+    if percent > 0:
+        return f"+{percent}Hz"
+    else:
+        return f"{percent}Hz"
+
+
 def azure_tts_v1(
-    text: str, voice_name: str, voice_rate: float, voice_file: str
+    text: str, voice_name: str, voice_rate: float, voice_pitch: float, voice_file: str
 ) -> [SubMaker, None]:
     voice_name = parse_voice_name(voice_name)
     text = text.strip()
     rate_str = convert_rate_to_percent(voice_rate)
+    pitch_str = convert_pitch_to_percent(voice_pitch)
     for i in range(3):
         try:
             logger.info(f"start, voice name: {voice_name}, try: {i + 1}")
 
             async def _do() -> SubMaker:
-                communicate = edge_tts.Communicate(text, voice_name, rate=rate_str, proxy="http://127.0.0.1:7890")
+                communicate = edge_tts.Communicate(text, voice_name, rate=rate_str, pitch=pitch_str, proxy=config.proxy.get("http"))
                 sub_maker = edge_tts.SubMaker()
                 with open(voice_file, "wb") as file:
                     async for chunk in communicate.stream():

config.example.toml

@@ -1,11 +1,12 @@
 [app]
-    project_version="0.2.0"
+    project_version="0.2.2"
-    # 如果你没有 OPENAI API Key，可以使用 g4f 代替，或者使用国内的 Moonshot API
+    # 支持视频理解的大模型提供商
-    # If you don't have an OPENAI API Key, you can use g4f instead
+    #   gemini
+    #   qwen2-vl (待增加)
     video_llm_provider="gemini"
 
-    # 支持的提供商 (Supported providers):
+    # 用于生成文案的大模型支持的提供商 (Supported providers):
-    #   openai
+    #   openai (默认)
     #   moonshot (月之暗面)
     #   oneapi
     #   g4f
@@ -13,8 +14,6 @@
     #   qwen (通义千问)
     #   gemini
     llm_provider="openai"
-    # 支持多模态视频理解能力的大模型
-
     ########## Ollama Settings
     # No need to set it unless you want to use your own proxy
     ollama_base_url = ""
@@ -27,7 +26,7 @@
     # No need to set it unless you want to use your own proxy
     openai_base_url = ""
     # Check your available models at https://platform.openai.com/account/limits
-    openai_model_name = "gpt-4-turbo"
+    openai_model_name = "gpt-4o"
 
     ########## Moonshot API Key
     # Visit https://platform.moonshot.cn/console/api-keys to get your API key.
@@ -56,7 +55,7 @@
 
     ########## Gemini API Key
     gemini_api_key=""
-    gemini_model_name = "gemini-1.5-flash"
+    gemini_model_name = "gemini-1.5-pro"
 
     ########## Qwen API Key
     # Visit https://dashscope.console.aliyun.com/apiKey to get your API key
@@ -66,29 +65,23 @@
     qwen_api_key = ""
     qwen_model_name = "qwen-max"
 
-
     ########## DeepSeek API Key
     # Visit https://platform.deepseek.com/api_keys to get your API key
     deepseek_api_key = ""
     deepseek_base_url = "https://api.deepseek.com"
     deepseek_model_name = "deepseek-chat"
 
-    # Subtitle Provider, "whisper"
+    # 字幕提供商、可选，支持 whisper 和 faster-whisper-large-v2"whisper"
-    # If empty, the subtitle will not be generated
+    # 默认为 faster-whisper-large-v2 模型地址：https://huggingface.co/guillaumekln/faster-whisper-large-v2
     subtitle_provider = "faster-whisper-large-v2"
     subtitle_enabled = true
 
-    #
     # ImageMagick
-    #
+    # 安装后，将自动检测到 ImageMagick，Windows 除外！
-    # Once you have installed it, ImageMagick will be automatically detected, except on Windows!
+    # 例如，在 Windows 上 "C:\Program Files (x86)\ImageMagick-7.1.1-Q16-HDRI\magick.exe"
-    # On Windows, for example "C:\Program Files (x86)\ImageMagick-7.1.1-Q16-HDRI\magick.exe"
+    # 下载位置 https://imagemagick.org/archive/binaries/ImageMagick-7.1.1-29-Q16-x64-static.exe
-    # Download from https://imagemagick.org/archive/binaries/ImageMagick-7.1.1-29-Q16-x64-static.exe
-
     # imagemagick_path = "C:\\Program Files (x86)\\ImageMagick-7.1.1-Q16\\magick.exe"
 
-
-    #
     # FFMPEG
     #
     # 通常情况下，ffmpeg 会被自动下载，并且会被自动检测到。
@@ -97,12 +90,6 @@
     #   Install ffmpeg on your system, or set the IMAGEIO_FFMPEG_EXE environment variable.
     # 此时你可以手动下载 ffmpeg 并设置 ffmpeg_path，下载地址：https://www.gyan.dev/ffmpeg/builds/
 
-    # Under normal circumstances, ffmpeg is downloaded automatically and detected automatically.
-    # However, if there is an issue with your environment that prevents automatic downloading, you might encounter the following error:
-    #   RuntimeError: No ffmpeg exe could be found.
-    #   Install ffmpeg on your system, or set the IMAGEIO_FFMPEG_EXE environment variable.
-    # In such cases, you can manually download ffmpeg and set the ffmpeg_path, download link: https://www.gyan.dev/ffmpeg/builds/
-
     # ffmpeg_path = "C:\\Users\\harry\\Downloads\\ffmpeg.exe"
     #########################################################################################
 
@@ -132,7 +119,7 @@
 
     material_directory = ""
 
-    # Used for state management of the task
+    # 用于任务的状态管理
     enable_redis = false
     redis_host = "localhost"
     redis_port = 6379
@@ -143,7 +130,6 @@
     max_concurrent_tasks = 5
 
     # webui界面是否显示配置项
-    # webui hide baisc config panel
     hide_config = false
 
 
@@ -161,7 +147,7 @@
 
     # recommended model_size: "large-v3"
     model_size="faster-whisper-large-v2"
-    # if you want to use GPU, set device="cuda"
+    # 如果要使用 GPU，请设置 device=“cuda”
     device="CPU"
     compute_type="int8"
 

webui.py

@@ -549,37 +549,6 @@ with middle_panel:
         params.voice_name = voice_name
         config.ui["voice_name"] = voice_name
 
-        # 试听语言合成
-        if st.button(tr("Play Voice")):
-            play_content = "感谢关注 NarratoAI，有任何问题或建议，可以关注微信公众号，求助或讨论"
-            if not play_content:
-                play_content = params.video_script
-            if not play_content:
-                play_content = tr("Voice Example")
-            with st.spinner(tr("Synthesizing Voice")):
-                temp_dir = utils.storage_dir("temp", create=True)
-                audio_file = os.path.join(temp_dir, f"tmp-voice-{str(uuid4())}.mp3")
-                sub_maker = voice.tts(
-                    text=play_content,
-                    voice_name=voice_name,
-                    voice_rate=params.voice_rate,
-                    voice_file=audio_file,
-                )
-                # 如果语音文件生成失败，请使用默认内容重试。
-                if not sub_maker:
-                    play_content = "This is a example voice. if you hear this, the voice synthesis failed with the original content."
-                    sub_maker = voice.tts(
-                        text=play_content,
-                        voice_name=voice_name,
-                        voice_rate=params.voice_rate,
-                        voice_file=audio_file,
-                    )
-
-                if sub_maker and os.path.exists(audio_file):
-                    st.audio(audio_file, format="audio/mp3")
-                    if os.path.exists(audio_file):
-                        os.remove(audio_file)
-
         if voice.is_azure_v2_voice(voice_name):
             saved_azure_speech_region = config.azure.get("speech_region", "")
             saved_azure_speech_key = config.azure.get("speech_key", "")
@@ -604,6 +573,45 @@ with middle_panel:
             index=2,
         )
 
+        params.voice_pitch = st.selectbox(
+            tr("Speech Pitch"),
+            options=[0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0],
+            index=2,
+        )
+
+        # 试听语言合成
+        if st.button(tr("Play Voice")):
+            play_content = "感谢关注 NarratoAI，有任何问题或建议，可以关注微信公众号，求助或讨论"
+            if not play_content:
+                play_content = params.video_script
+            if not play_content:
+                play_content = tr("Voice Example")
+            with st.spinner(tr("Synthesizing Voice")):
+                temp_dir = utils.storage_dir("temp", create=True)
+                audio_file = os.path.join(temp_dir, f"tmp-voice-{str(uuid4())}.mp3")
+                sub_maker = voice.tts(
+                    text=play_content,
+                    voice_name=voice_name,
+                    voice_rate=params.voice_rate,
+                    voice_pitch=params.voice_pitch,
+                    voice_file=audio_file,
+                )
+                # 如果语音文件生成失败，请使用默认内容重试。
+                if not sub_maker:
+                    play_content = "This is a example voice. if you hear this, the voice synthesis failed with the original content."
+                    sub_maker = voice.tts(
+                        text=play_content,
+                        voice_name=voice_name,
+                        voice_rate=params.voice_rate,
+                        voice_pitch=params.voice_pitch,
+                        voice_file=audio_file,
+                    )
+
+                if sub_maker and os.path.exists(audio_file):
+                    st.audio(audio_file, format="audio/mp3")
+                    if os.path.exists(audio_file):
+                        os.remove(audio_file)
+
         bgm_options = [
             (tr("No Background Music"), ""),
             (tr("Random Background Music"), "random"),

webui/i18n/zh.json

@@ -91,6 +91,7 @@
     "Picture description": "图片描述",
     "Narration": "视频文案",
     "Rebuild": "重新生成",
-    "Video Script Load": "加载视频脚本"
+    "Video Script Load": "加载视频脚本",
+    "Speech Pitch": "语调"
   }
 }