语音播报 tts espeak

目录

chattts 对话式，可以生成语气，笑声

edge_tts 麒麟系统效果不错

男播音员：

播报人列表：

播报人代码：

 gtts语音播报，Windows正常，麒麟系统卡死

espeak语音tts

安装：sudo apt install espeak

Full dictionary is not installed for ‘zh’

报错：Can't access (r) file 'zh_rules'

麒麟系统 解决

2、pyttsx3 ubuntu使用

---

chattts 对话式，可以生成语气，笑声

GitHub - 2noise/ChatTTS: ChatTTS is a generative speech model for daily dialogue.

edge_tts 麒麟系统效果不错

edge-tts是github上的一个开源项目，可以免费将文本转为语音，别看它只有2.8k star，替代科大讯飞的收费TTS服务完全没问题，因为这个项目实际是调用的微软edge的在线语音合成服务，支持40多种语言，300多种声音，效果毋容置疑。
https://github.com/rany2/edge-tts

pip install edge-tts

import edge_tts
import asyncio

async def text_to_speech(text, output_file):
    communicate = edge_tts.Communicate(text, "zh-CN-XiaoxiaoNeural")
    await communicate.save(output_file)

text = "赵晓红，检录通过"
output_file = "hello1.mp3"

asyncio.run(text_to_speech(text, output_file))

男播音员：

async def text_to_speech(text, output_file):
    # communicate = edge_tts.Communicate(text, "zh-CN-YunjianNeural")
    # communicate = edge_tts.Communicate(text, "zh-CN-YunxiNeural")
    communicate = edge_tts.Communicate(text, "zh-CN-YunyangNeural")
    await communicate.save(output_file)

播报人列表：

Text to Speech Voice (zh-HK, HiuGaaiNeural), Gender: Female, Locale: zh-HK
Text to Speech Voice (zh-HK, HiuMaanNeural), Gender: Female, Locale: zh-HK
Text to Speech Voice (zh-HK, WanLungNeural), Gender: Male, Locale: zh-HK
Text to Speech Voice (zh-CN, XiaoxiaoNeural), Gender: Female, Locale: zh-CN
Text to Speech Voice (zh-CN, XiaoyiNeural), Gender: Female, Locale: zh-CN
Text to Speech Voice (zh-CN, YunjianNeural), Gender: Male, Locale: zh-CN
Text to Speech Voice (zh-CN, YunxiNeural), Gender: Male, Locale: zh-CN
Text to Speech Voice (zh-CN, YunxiaNeural), Gender: Male, Locale: zh-CN
Text to Speech Voice (zh-CN, YunyangNeural), Gender: Male, Locale: zh-CN
Text to Speech Voice (zh-CN-liaoning, XiaobeiNeural), Gender: Female, Locale: zh-CN-liaoning
Text to Speech Voice (zh-TW, HsiaoChenNeural), Gender: Female, Locale: zh-TW
Text to Speech Voice (zh-TW, YunJheNeural), Gender: Male, Locale: zh-TW
Text to Speech Voice (zh-TW, HsiaoYuNeural), Gender: Female, Locale: zh-TW
Text to Speech Voice (zh-CN-shaanxi, XiaoniNeural), Gender: Female, Locale: zh-CN-shaanxi

播报人代码：

import sys
from gtts import gTTS
import edge_tts
import asyncio

def text_to_speech(text, lang='zh'):
    tts = gTTS(text=text, lang=lang)
    tts.save("output.mp3")

def edge_tts_demo():
    import edge_tts
    import asyncio

    async def text_to_speech(text, output_file):
        communicate = edge_tts.Communicate(text, "zh-CN-XiaoxiaoNeural")
        await communicate.save(output_file)

    text = "赵晓红，检录通过"
    output_file = "hello1.mp3"

    asyncio.run(text_to_speech(text, output_file))

async def list_voices():
    voices = await edge_tts.list_voices()
    for voice in voices:
        if 'zh' in voice['Name']:
            print(f"Name: {voice['Name']}, Gender: {voice['Gender']}, Locale: {voice['Locale']}")

async def text_to_speech(text, voice_name, output_file):
    communicate = edge_tts.Communicate(text, voice_name)
    await communicate.save(output_file)
    print(f'Speech saved to {output_file}')


if __name__ == "__main__":
   # tts = gTTS(text="赵晓红，检录通过", lang="zh")
   # tts.save("hello1.mp3")

    # List available Chinese voices
    asyncio.run(list_voices())

    # Example usage with a specific voice
    text = "赵晓红，检录通过"
    voice_name = "zh-CN-XiaoxiaoNeural"  # Replace with the desired voice name
    output_file = "hello2.mp3"

    asyncio.run(text_to_speech(text, voice_name, output_file))

    # text_to_speech("赵晓红，检录通过")

 gtts语音播报，Windows正常，麒麟系统卡死

import sys
from gtts import gTTS

def text_to_speech(text, lang='zh'):
    tts = gTTS(text=text, lang=lang)
    tts.save("output.mp3")

if __name__ == "__main__":
    tts = gTTS(text="赵晓红，检录通过", lang="zh-tw")
    tts.save("hello1.mp3")

    # text_to_speech("赵晓红，检录通过")

espeak语音tts

linux espeak语音tts

安装：sudo apt install espeak

使用：
#中文男声
espeak -v zh 你好  
#中文女声
espeak -v zh+f3 你好
#粤语男声
espeak -v zhy 你好
 

Full dictionary is not installed for ‘zh’

注意：espeak -v zh 你好 （Full dictionary is not installed for ‘zh’）
解决方法：https://www.cnblogs.com/Hfolsvh/p/15057694.html

##1、重新下载语音包https://github.com/caixxiong/espeak-data/
git clone https://github.com/caixxiong/espeak-data/
cd espeak-data/
unzip espeak-data.zip
cd espeak-data/

##2、拷贝解压文件到原安装包对应目录
sudo cp -r * /usr/lib/x86_64-linux-gnu/espeak-data 

##3、编译中文、粤语包
sudo espeak --compile=zh
sudo espeak --compile=zhy

报错：Can't access (r) file 'zh_rules'

麒麟系统 解决

下载

https://github.com/nvaccess/espeak

把dictsource目录拷贝到：

/usr/lib/aarch64-linux-gnu/espeak-data/

cd /usr/lib/aarch64-linux-gnu/espeak-data/

espeak –compile=zh

然后编译通过了。

##4、编译成功后需要把原来wget下载的espeak-data目录全部整体删除，不然可能报espeak -v zhy 你好
Can't read data file: '/home/loong/espeak-data/phontab'
Failed to load espeak-data

解决方法：rm -rf espeak-data
 

2、pyttsx3 ubuntu使用

参考：https://blog.youkuaiyun.com/Wickedstepmother/article/details/130993164

pyttsx3包主要就是调用电脑自带的声音tts工具，windows是调用微软自带的声音库；linux就是调用安装的espeak

安装

pip install pyttsx3 -i https://pypi.tuna.tsinghua.edu.cn/simple --user

使用

import pyttsx3

engine = pyttsx3.init()

engine.setProperty('voice','zh')  ##可以替换zhy 粤语；zh+f3中文女声
engine.say("你好，晚上吃什么？")
engine.runAndWait()

                        
原文链接：https://blog.youkuaiyun.com/weixin_42357472/article/details/134340397