介绍
从IOS5开始，IOS系统已经在siri上集成了语音合成的功能，但是是私有API。但是在IOS7，新增了一个简单的API----AVSpeechSynthesizer来做这件事情。
嗓音
IOS7包含了一组可以用来合成的声音的嗓音，你可以自定义多种嗓音来合成。 AVSpeechSynthesisVoice:speechVoices返回当前可用的嗓音数组：
2013-07-12 10:49:26.929 GreetingSpeaker[31267:70b] (
"[AVSpeechSynthesisVoice 0x978a0b0] Language: th-TH",
"[AVSpeechSynthesisVoice 0x977a450] Language: pt-BR",
"[AVSpeechSynthesisVoice 0x977a480] Language: sk-SK",
"[AVSpeechSynthesisVoice 0x978ad50] Language: fr-CA",
"[AVSpeechSynthesisVoice 0x978ada0] Language: ro-RO",
"[AVSpeechSynthesisVoice 0x97823f0] Language: no-NO",
"[AVSpeechSynthesisVoice 0x978e7b0] Language: fi-FI",
"[AVSpeechSynthesisVoice 0x978af50] Language: pl-PL",
"[AVSpeechSynthesisVoice 0x978afa0] Language: de-DE",
"[AVSpeechSynthesisVoice 0x978e390] Language: nl-NL",
"[AVSpeechSynthesisVoice 0x978b030] Language: id-ID",
"[AVSpeechSynthesisVoice 0x978b080] Language: tr-TR",
"[AVSpeechSynthesisVoice 0x978b0d0] Language: it-IT",
"[AVSpeechSynthesisVoice 0x978b120] Language: pt-PT",
"[AVSpeechSynthesisVoice 0x978b170] Language: fr-FR",
"[AVSpeechSynthesisVoice 0x978b1c0] Language: ru-RU",
"[AVSpeechSynthesisVoice 0x978b210]Language: es-MX",
"[AVSpeechSynthesisVoice 0x978b2d0] Language: zh-HK",
"[AVSpeechSynthesisVoice 0x978b320] Language: sv-SE",
"[AVSpeechSynthesisVoice 0x978b010] Language: hu-HU",
"[AVSpeechSynthesisVoice 0x978b440] Language: zh-TW",
"[AVSpeechSynthesisVoice 0x978b490] Language: es-ES",
"[AVSpeechSynthesisVoice 0x978b4e0] Language: zh-CN",
"[AVSpeechSynthesisVoice 0x978b530] Language: nl-BE",
"[AVSpeechSynthesisVoice 0x978b580] Language: en-GB",
"[AVSpeechSynthesisVoice 0x978b5d0] Language: ar-SA",
"[AVSpeechSynthesisVoice 0x978b620] Language: ko-KR",
"[AVSpeechSynthesisVoice 0x978b670] Language: cs-CZ",
"[AVSpeechSynthesisVoice 0x978b6c0] Language: en-ZA",
"[AVSpeechSynthesisVoice 0x978aed0] Language: en-AU",
"[AVSpeechSynthesisVoice 0x978af20] Language: da-DK",
"[AVSpeechSynthesisVoice 0x978b810] Language: en-US",
"[AVSpeechSynthesisVoice 0x978b860] Language: en-IE",
"[AVSpeechSynthesisVoice 0x978b8b0] Language: hi-IN",
"[AVSpeechSynthesisVoice 0x978b900] Language: el-GR",
"[AVSpeechSynthesisVoice 0x978b950] Language: ja-JP" )
用下面的方法来创建一个特殊的嗓音：
AVSpeechSynthesisVoice *voice = [AVSpeechSynthesisVoice voiceWithLanguage:@"en-US"];
如果嗓音不能被识别，返回值为nil。
话语
话语代表一段讲话，可以传递到语音合成器来生成一段语音流。话语可以用待发音的字符串来创建：
AVSpeechUtterance *utterance = [AVSpeechUtterance speechUtteranceWithString:@"Hello world!"];
我们可以设置它的嗓音属性voice：
utterence.voice = voice;
还有其他各种属性，包括：rate，volume和pitchMultiplier。例如，稍微放慢语速：
utterance.rate *= 0.7;
当话语创建完成，可以把它放到语音合成器中。语音合成器会生成音频：
AVSpeechSynthesizer *speechSynthesizer = [[AVSpeechSynthesizer alloc] init];
[speechSynthesizer speakUtterance:utterance];
话语会被合成器放到队列中，所以不用等待上一个结束，可以连续传送。但是，如果你尝试传送一个已经在队列中的话语，系统会抛出异常。
实现
本文的例子是一个关于多语言问候语的应用。这充分体现了语音合成的通用性。
注意：定义话语的字符串都是罗马字符，例如中文的'Ni hao'。示例项目定义了一个类，可以生成一组嗓音对应的话语。
项目有一个选择器供用户选择语言，然后按按钮可以听到对应的生成的语音。
总结
在IOS7，语音合成器非常简单，而且包含很多语言。合理的使用，可以提高应用的访问性，而不用时刻用你的手和眼来操作应用。