Java使用Jacob調用SAPI合成語音

基於語音測試須要，須要將文字數據轉換成語音數據，文章未經原做者容許轉載，只用於記錄。

1.Jar包

• 版本號：1.18
• maven地址：https://mvnrepository.com/artifact/com.hynnet/jacob

2.環境設置

• JDK版本：1.8（原來採用1.7設置與1.8一致，一直報錯，還不知道是否是版本的緣由，原做者使用1.7沒問題）
• 具體設置：解壓後，將x86或x64的dll文件放入c:\windows\system32或syswow64目錄下；同時也放入jdk和jre的bin、lib目錄裏；

3.MSTTSSpeech代碼

package com.thinkgem.jeesite.common.utils;

import com.jacob.activeX.ActiveXComponent;
import com.jacob.com.ComThread;
import com.jacob.com.Dispatch;
import com.jacob.com.Variant;

public class MSTTSSpeech {
	private int volume=100;// 聲音：1到100  
    private int rate=0;// 頻率：-10到10  
    private int voice=0;// 語音庫序號  
    private int audio=0;// 輸出設備序號  
    private ActiveXComponent ax=null;  
    private Dispatch spVoice=null;// 聲音對象  
    private Dispatch spFileStream=null;// 音頻文件輸出流對象，在讀取或保存音頻文件時使用  
    private Dispatch spAudioFormat=null;// 音頻格式對象  
    private Dispatch spMMAudioOut=null;// 音頻輸出對象  
    private int formatType=22;// 音頻的輸出格式，默認爲：SAFT22kHz16BitMono  
    public MSTTSSpeech()   
    {  
        ComThread.InitSTA();  
        if(ax==null)  
        {  
            ax=new ActiveXComponent("Sapi.SpVoice");  
            spVoice=ax.getObject();  
        }  
    }  
    /** 
     * 改變語音庫 
     * @param voice 語音庫序號 
     */  
    public void changeVoice(int voice)  
    {  
        if(this.voice != voice)  
        {  
            this.voice=voice;  
        }  
        try  
        {  
            Dispatch voiceItems=Dispatch.call(spVoice,"GetVoices").toDispatch();  
            int count=Integer.valueOf(Dispatch.call(voiceItems,"Count").toString());  
            if(count>0)  
            {  
                Dispatch voiceItem=Dispatch.call(voiceItems,"Item",new Variant(this.voice)).toDispatch();  
                Dispatch.put(spVoice,"Voice",voiceItem);  
            }  
        }  
        catch(Exception e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
    }  
    /** 
     * 改變音頻輸出設備 
     * @param audio 音頻設備序號 
     */  
    public void changeAudioOutput(int audio)  
    {  
        if(this.audio != audio)  
        {  
            this.audio=audio;  
        }  
        try  
        {  
            Dispatch audioOutputs=Dispatch.call(spVoice,"GetAudioOutputs").toDispatch();  
            int count=Integer.valueOf(Dispatch.call(audioOutputs,"Count").toString());  
            if(count > 0)  
            {  
                Dispatch audioOutput=Dispatch.call(audioOutputs,"Item",new Variant(this.audio)).toDispatch();  
                Dispatch.put(spVoice,"AudioOutput",audioOutput);  
            }  
        }  
        catch(Exception e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
    }  
    /** 
     * 播放語音  
     * @param text 要轉換成語音的文本 
     */  
    public void speak(String text)  
    {  
        this.speak(text,0);  
    }  
    /** 
     * 中止播放語音 
     */  
    public void stop()  
    {  
        // this.speak("", 1);  
        Dispatch.call(spVoice,"Pause");  
    }  
    /** 
     * 播放語音 
     * @param text 要轉換成語音的文本 
     * @param type 類型0:播放，1：中止 
     */  
    private void speak(String text,int type)  
    {  
        switch(type)   
        {  
            case 0:  
                try  
                {  
                    // 調整音量和讀的速度  
                    Dispatch.put(spVoice,"Volume",new Variant(this.volume));// 設置音量  
                    Dispatch.put(spVoice,"Rate",new Variant(this.rate));// 設置速率  
                    // 設置音頻格式類型  
                    if(spAudioFormat==null)  
                    {  
                        ax=new ActiveXComponent("Sapi.SpAudioFormat");  
                        spAudioFormat=ax.getObject();  
                        ax=new ActiveXComponent("Sapi.SpMMAudioOut");  
                        spMMAudioOut=ax.getObject();  
                    }  
                    Dispatch.put(spAudioFormat,"Type",new Variant(this.formatType));  
                    Dispatch.putRef(spMMAudioOut,"Format",spAudioFormat);  
                    Dispatch.put(spVoice,"AllowAudioOutputFormatChangesOnNextSet",new Variant(false));  
                    Dispatch.putRef(spVoice,"AudioOutputStream",spMMAudioOut);  
                    // 開始朗讀  
                    Dispatch.call(spVoice,"Speak",new Variant(text));  
                }  
                catch(Exception e)  
                {  
                    System.out.println(e.getMessage());  
                    e.printStackTrace();  
                }  
                break;  
            case 1:  
                try  
                {  
                    Dispatch.call(spVoice,"Speak",new Variant(text),new Variant(2));  
                }  
                catch(Exception e)  
                {  
                    System.out.println(e.getMessage());  
                    e.printStackTrace();  
                }  
                break;  
            default:  
                break;  
        }  
    }  
    /** 
     * 獲取系統中全部的語音庫名稱數組  
     * @return String[] 
     */  
    public String[] getVoices()  
    {  
        String[] voices=null;  
        try  
        {  
            Dispatch voiceItems=Dispatch.call(spVoice,"GetVoices").toDispatch();  
            int count=Integer.valueOf(Dispatch.call(voiceItems,"Count").toString());  
            if(count > 0)  
            {  
                voices=new String[count];  
                for(int i=0;i<count;i++)  
                {  
                    Dispatch voiceItem=Dispatch.call(voiceItems,"Item",new Variant(i)).toDispatch();  
                    String voice=Dispatch.call(voiceItem,"GetDescription").toString();  
                    voices[i]=voice;  
                }  
            }  
        }  
        catch(Exception e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
        return voices;  
    }  
    /** 
     * 獲取音頻輸出設備名稱數組 
     * @return String[] 
     */  
    public String[] getAudioOutputs()  
    {  
        String[] result=null;  
        try  
        {  
            Dispatch audioOutputs=Dispatch.call(spVoice,"GetAudioOutputs").toDispatch();  
            int count=Integer.valueOf(Dispatch.call(audioOutputs,"Count").toString());  
            if(count > 0)  
            {  
                result=new String[count];  
                for(int i=0;i<count;i++)  
                {  
                    Dispatch voiceItem=Dispatch.call(audioOutputs,"Item",new Variant(i)).toDispatch();  
                    String voice=Dispatch.call(voiceItem,"GetDescription").toString();  
                    result[i]=voice;  
                }  
            }  
        }  
        catch(Exception e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
        return result;  
    }  
    /** 
     * 將文字轉換成音頻信號，而後輸出到.WAV文件 
     * @param text 文本字符串 
     * @param filePath 輸出文件路徑 
     */  
    public void saveToWav(String text,String filePath)  
    {  
        // 建立輸出文件流對象  
        ax=new ActiveXComponent("Sapi.SpFileStream");  
        spFileStream=ax.getObject();  
        // 建立音頻流格式對象  
        if(spAudioFormat==null)  
        {  
            ax=new ActiveXComponent("Sapi.SpAudioFormat");  
            spAudioFormat=ax.getObject();  
        }  
        // 設置音頻流格式類型  
        Dispatch.put(spAudioFormat,"Type",new Variant(this.formatType));  
        // 設置文件輸出流的格式  
        Dispatch.putRef(spFileStream,"Format",spAudioFormat);  
        // 調用輸出文件流對象的打開方法，建立一個.wav文件  
        Dispatch.call(spFileStream,"Open",new Variant(filePath),new Variant(3),new Variant(true));  
        // 設置聲音對象的音頻輸出流爲輸出文件流對象  
        Dispatch.putRef(spVoice,"AudioOutputStream",spFileStream);  
        // 調整音量和讀的速度  
        Dispatch.put(spVoice,"Volume",new Variant(this.volume));// 設置音量  
        Dispatch.put(spVoice,"Rate",new Variant(this.rate));// 設置速率  
        // 開始朗讀  
        Dispatch.call(spVoice,"Speak",new Variant(text));  
        // 關閉輸出文件流對象，釋放資源  
        Dispatch.call(spFileStream,"Close");  
        Dispatch.putRef(spVoice,"AudioOutputStream",null);  
    }  
    /** 
     * @return the volume 
     */  
    public int getVolume()  
    {  
        return volume;  
    }  
    /** 
     * @param volume 
     * the volume to set 
     */  
    public void setVolume(int volume)  
    {  
        this.volume = volume;  
    }  
    /** 
     * @return the rate 
     */  
    public int getRate()  
    {  
        return rate;  
    }  
    /** 
     * @param rate 
     * the rate to set 
     */  
    public void setRate(int rate)  
    {  
        this.rate = rate;  
    }  
    /** 
     * @return the voice 
     */  
    public int getVoice()  
    {  
        return voice;  
    }  
    /** 
     * @param voice 
     * the voice to set 
     */  
    public void setVoice(int voice)  
    {  
        this.voice = voice;  
    }  
    /** 
     * @return the audio 
     */  
    public int getAudio()  
    {  
        return audio;  
    }  
    /** 
     * @param audio 
     * the audio to set 
     */  
    public void setAudio(int audio)  
    {  
        this.audio=audio;  
    }  
    /** 
     * @return the ax 
     */  
    public ActiveXComponent getAx()  
    {  
        return ax;  
    }  
    /** 
     * @param ax 
     * the ax to set 
     */  
    public void setAx(ActiveXComponent ax)  
    {  
        this.ax=ax;  
    }  
    /** 
     * @return the formatType 
     */  
    public int getFormatType()  
    {  
        return formatType;  
    }  
    /** 
     * 設置音頻輸出格式類型<br> 
     * SAFTDefault = -1<br> 
     * SAFTNoAssignedFormat = 0<br> 
     * SAFTText = 1<br> 
     * SAFTNonStandardFormat = 2<br> 
     * SAFTExtendedAudioFormat = 3<br> 
     * // Standard PCM wave formats<br> 
     * SAFT8kHz8BitMono = 4<br> 
     * SAFT8kHz8BitStereo = 5<br> 
     * SAFT8kHz16BitMono = 6<br> 
     * SAFT8kHz16BitStereo = 7<br> 
     * SAFT11kHz8BitMono = 8<br> 
     * SAFT11kHz8BitStereo = 9<br> 
     * SAFT11kHz16BitMono = 10<br> 
     * SAFT11kHz16BitStereo = 11<br> 
     * SAFT12kHz8BitMono = 12<br> 
     * SAFT12kHz8BitStereo = 13<br> 
     * SAFT12kHz16BitMono = 14<br> 
     * SAFT12kHz16BitStereo = 15<br> 
     * SAFT16kHz8BitMono = 16<br> 
     * SAFT16kHz8BitStereo = 17<br> 
     * SAFT16kHz16BitMono = 18<br> 
     * SAFT16kHz16BitStereo = 19<br> 
     * SAFT22kHz8BitMono = 20<br> 
     * SAFT22kHz8BitStereo = 21<br> 
     * SAFT22kHz16BitMono = 22<br> 
     * SAFT22kHz16BitStereo = 23<br> 
     * SAFT24kHz8BitMono = 24<br> 
     * SAFT24kHz8BitStereo = 25<br> 
     * SAFT24kHz16BitMono = 26<br> 
     * SAFT24kHz16BitStereo = 27<br> 
     * SAFT32kHz8BitMono = 28<br> 
     * SAFT32kHz8BitStereo = 29<br> 
     * SAFT32kHz16BitMono = 30<br> 
     * SAFT32kHz16BitStereo = 31<br> 
     * SAFT44kHz8BitMono = 32<br> 
     * SAFT44kHz8BitStereo = 33<br> 
     * SAFT44kHz16BitMono = 34<br> 
     * SAFT44kHz16BitStereo = 35<br> 
     * SAFT48kHz8BitMono = 36<br> 
     * SAFT48kHz8BitStereo = 37<br> 
     * SAFT48kHz16BitMono = 38<br> 
     * SAFT48kHz16BitStereo = 39<br> 
     * <br> 
     * // TrueSpeech format<br> 
     * SAFTTrueSpeech_8kHz1BitMono = 40<br> 
     * // A-Law formats<br> 
     * SAFTCCITT_ALaw_8kHzMono = 41<br> 
     * SAFTCCITT_ALaw_8kHzStereo = 42<br> 
     * SAFTCCITT_ALaw_11kHzMono = 43<br> 
     * SAFTCCITT_ALaw_11kHzStereo = 4<br> 
     * SAFTCCITT_ALaw_22kHzMono = 44<br> 
     * SAFTCCITT_ALaw_22kHzStereo = 45<br> 
     * SAFTCCITT_ALaw_44kHzMono = 46<br> 
     * SAFTCCITT_ALaw_44kHzStereo = 47<br> 
     * <br> 
     * // u-Law formats<br> 
     * SAFTCCITT_uLaw_8kHzMono = 48<br> 
     * SAFTCCITT_uLaw_8kHzStereo = 49<br> 
     * SAFTCCITT_uLaw_11kHzMono = 50<br> 
     * SAFTCCITT_uLaw_11kHzStereo = 51<br> 
     * SAFTCCITT_uLaw_22kHzMono = 52<br> 
     * SAFTCCITT_uLaw_22kHzStereo = 53<br> 
     * SAFTCCITT_uLaw_44kHzMono = 54<br> 
     * SAFTCCITT_uLaw_44kHzStereo = 55<br> 
     * SAFTADPCM_8kHzMono = 56<br> 
     * SAFTADPCM_8kHzStereo = 57<br> 
     * SAFTADPCM_11kHzMono = 58<br> 
     * SAFTADPCM_11kHzStereo = 59<br> 
     * SAFTADPCM_22kHzMono = 60<br> 
     * SAFTADPCM_22kHzStereo = 61<br> 
     * SAFTADPCM_44kHzMono = 62<br> 
     * SAFTADPCM_44kHzStereo = 63<br> 
     * <br> 
     * // GSM 6.10 formats<br> 
     * SAFTGSM610_8kHzMono = 64<br> 
     * SAFTGSM610_11kHzMono = 65<br> 
     * SAFTGSM610_22kHzMono = 66<br> 
     * SAFTGSM610_44kHzMono = 67<br> 
     * // Other formats<br> 
     * SAFTNUM_FORMATS = 68<br> 
     *  
     * @param formatType 
     *            音頻輸出格式類型 
     */  
    public void setFormatType(int formatType)  
    {  
        this.formatType=formatType;  
    }  
    public static void main(String[] args)  
    {  
        MSTTSSpeech speech=new MSTTSSpeech();  
        String text="這是個人測試,物理內存至少須要512MB，建議2GB以上,虛擬內存是主機物理內存的兩倍，不要設到系統盤,硬盤空閒空間大於4.77GB.";  
        speech.setFormatType(6);  
        // speech.setRate(-1);  
        speech.saveToWav(text,"E://test.wav");  
        //speech.speak(text);  
        SoundUtil.wavToPcm("E://test.wav");  
    }
}

4.SoundUtil代碼

把wav語音文件轉換爲pcm，其中用的Jar包爲sound.jar，該jar包自帶於jdk裏，不用處處找了。

package com.thinkgem.jeesite.common.utils;

import java.io.File;
import java.io.IOException;

import javax.sound.sampled.AudioFileFormat;
import javax.sound.sampled.AudioFormat;
import javax.sound.sampled.AudioInputStream;
import javax.sound.sampled.AudioSystem;
import javax.sound.sampled.UnsupportedAudioFileException;

/** 
 * 與聲音有關的工具類 
 * @author suyunlong 
 * 
 */  
public class SoundUtil  
{  
    /** 
     * 將WAV格式的音頻文件轉換爲PCM格式的文件 
     * @param wavFilePath WAV文件路徑 
     * @throws UnsupportedAudioFileException 
     * @throws IOException 
     */  
    public static void wavToPcm(String wavFilePath)  
    {  
        try  
        {  
            File wavFile=new File(wavFilePath);  
            AudioInputStream sourceAudioInputStream=AudioSystem.getAudioInputStream(wavFile);  
            // 此處的轉換必需是16bit的音頻文件  
            AudioInputStream targetAudioInputStream=AudioSystem.getAudioInputStream(AudioFormat.Encoding.ULAW,  
                    sourceAudioInputStream);  
            String newFilePath=wavFilePath.substring(0,wavFilePath.lastIndexOf(".")+1)+"pcm";  
            AudioSystem.write(targetAudioInputStream,AudioFileFormat.Type.WAVE,new File(newFilePath));  
        }  
        catch(UnsupportedAudioFileException e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
        catch(IOException e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
    }  
    /** 
     * 獲取音頻文件的編碼格式  
     * @param wavFilePath 音頻文件格式 
     * @return String 
     */  
    public static String getWavFormat(String wavFilePath)  
    {  
        File wavFile=new File(wavFilePath);  
        AudioInputStream ais;  
        String result="";  
        try  
        {  
            ais=AudioSystem.getAudioInputStream(wavFile);  
            AudioFormat af=ais.getFormat();  
            result=af.toString();  
        }  
        catch(UnsupportedAudioFileException e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
        catch(IOException e)  
        {  
            System.out.println(e.getMessage());  
            e.printStackTrace();  
        }  
        return result;  
    }  
    public static void main(String[] args)  
    {  
        System.out.println(SoundUtil.getWavFormat("E://syl//C//test.wav"));  
        SoundUtil.wavToPcm("E://syl//C//test.wav");  
    }  
}

5.測試報錯問題

• 問題：java.lang.UnsatisfiedLinkError: no jacob in java.library.path
• 解決：x86或x64的dll文件放入c:\windows\system32或syswow64目錄下；同時也放入jdk和jre的bin、lib目錄裏；
• 問題：Could not initialize class com.jacob.com.ComThread
• 解決：更換JDK版本，我在本機上測試1.7會報錯，1.8經過，請自行嘗試；

6.參考

• http://blog.csdn.net/asuyunlong/article/details/50083421
• http://www.iteye.com/topic/338223