首先到訊飛上注冊一個帳號(感覺訊飛官網和阿里云官網的差不多)

隨后創建一個新應用(Windows平臺下)

最后我們要的是Appid如下圖所示

下面是訊飛SDK(含說明文檔)下載地址：

鏈接：http://pan.baidu.com/s/1i45xEo1 密碼：m8il

因為此程序涉及WAV語音知識，未了解WAV的朋友請百度補充，或者在下面這個鏈接初略的學習下

http://blog.csdn.net/qq78442761/article/details/53385561

下面先來認識下幾個API(從文檔提取出來)

1.int MSPAPI MSPLogin (const char *usr, const char *pwd, const char *params) //初始化msc，用戶登錄。

usr[in] 此參數保留，傳入NULL即可。
pwd[in] 此參數保留，傳入NULL即可。

官方例子

const char* usr = NULL; const char* pwd = NULL; const char* lgi_param = "appid = ********"; int ret = MSPLogin(usr, pwd, lgi_param); if( MSP_SUCCESS != ret ) {printf( "MSPLogin failed, error code is: %d", ret ); }

2.const char *MSPAPI QTTSSessionBegin (const char *params, int *errorCode) //開始一次語音合成，分配語音合成資源。

(信息有點多,參數具體見文檔qtts.h 文件參考)

下面是官方例子

const char * ssb_param = "voice_name = xiaoyan, aue = speex-wb;7, sample_rate = 16000, speed = 50, volume = 50, pitch = 50, rdn = 2"; int ret = -1; const char * sessionID = QTTSSessionBegin( ssb_param, &ret ); if( MSP_SUCCESS != ret ) {printf( “QTTSSessionBegin failed, error code is: %d”, ret ); }

3.const void *MSPAPI QTTSAudioGet (const char *sessionID, unsigned int *audioLen, int *synthStatus, int *errorCode) //獲取合成音頻
(信息有點多,參數具體見文檔qtts.h 文件參考)

例子

FILE* fp = fopen("tts.pcm", "wb"); while (1) {const void * data = QTTSAudioGet(sessionID, &audio_len, &synth_status, &ret);if (NULL != data){fwrite(data, audio_len, 1, fp);}if (MSP_TTS_FLAG_DATA_END == synth_status || MSP_SUCCESS != ret){break;} } fclose(fp);

4.int MSPAPI QTTSSessionEnd (const char *sessionID, const char *hints) //結束本次語音合成。

int ret = QTTSSessionEnd ( sessionID, “normal end” ); if( MSP_SUCCESS != ret ) {printf( “QTTSSessionEnd failed, error code is: %d”, ret ); }


下面是完整項目的代碼:

幾個個文件1.stdafx.h。2.targetver.h。3.QTTSDemo.cpp。4.stdafx.cpp。

除了QTTSDemo.cpp外，其他文件都是vs2013創建控制臺程序自帶的，在此只給出QTTSDemo.cpp代碼

但注意在stdafx.h中加入#define _CRT_SECURE_NO_WARNINGS

QTTSDemo.cpp

// QTTSDemo.cpp : 定義控制臺應用程序的入口點。 //#include "stdafx.h" #include <msp_cmn.h> #include <msp_errors.h> #include <qtts.h> #include <string.h> #include <windows.h>#pragma comment(lib,"WinMM.lib")#ifdef _WIN64 #pragma comment(lib,"msc_x64.lib") #else #pragma comment(lib,"msc.lib") #endif//wav音頻頭部格式 typedef struct _wave_pcm_hdr {char riff[4]; //="RIFF"int size_8; //=FileSize=8char wave[4]; //="WAVE"char fmt[4]; //="fmt"int fmt_size; //=下一個結構體的大小:16short int format_tag; //=PCM:1short int channels; //=通道數:1int samples_per_sec; //=采樣率：8000|6000|16000int avg_bytes_per_sec; //=每秒字節數：samples_per_sec*bit_per_sampleshort int block_align; //=每采樣點字節數：wBitsPerSample/8short int bits_per_sample; //=量化比特數:8|16char data[4]; //="data";int data_size; //=純數據長度：FileSize-44 }wave_pcm_hdr;/*默認wav音頻頭部數據*/ wave_pcm_hdr default_wav_hdr = {{ 'R', 'I', 'F', 'F' },0,{ 'W', 'A', 'V', 'E' },{ 'f', 'm', 't', ' ' },16,1,1,16000,32000,2,16,{ 'd', 'a', 't', 'a' },0 };int _tmain(int argc, _TCHAR* argv[]) {//登錄const char* usr = NULL;const char* pwd = NULL;const char* lgi_param = "appid = 583aea17";int ret = MSPLogin(usr, pwd, lgi_param);if (MSP_SUCCESS != ret){printf("MSPLogin failed, error code is: %d", ret);}//開始合成const char * ssb_param = "voice_name = xiaorong, aue = speex-wb;7, sample_rate = 16000, speed = 50, volume = 80, pitch = 50, rdn = 2";ret = -1;const char * sessionID = QTTSSessionBegin(ssb_param, &ret);if (MSP_SUCCESS != ret){printf("QTTSSessionBegin failed, error code is : %d", ret);}//設置待合成文本char src_text[1024];printf("請輸入一段文字(中文注意逗號,句話)：\n");gets(src_text);unsigned int text_len = strlen(src_text); //textLen參數為合成文本所占字節數ret = QTTSTextPut(sessionID, src_text, text_len, NULL);if (MSP_SUCCESS != ret){printf("QTTSTextPut failed, error code is : %d", ret);}//獲取合成的音頻wavFILE* fp = fopen("Demo.wav", "wb"); //一定是二進制模式fwrite(&default_wav_hdr, sizeof(default_wav_hdr), 1, fp);unsigned int audio_len = 0;int synth_status = 0;while (1){const void * data = QTTSAudioGet(sessionID, &audio_len, &synth_status, &ret);if (NULL != data){fwrite(data, audio_len, 1, fp);default_wav_hdr.data_size += audio_len;}if (MSP_TTS_FLAG_DATA_END == synth_status || MSP_SUCCESS != ret){break;}}default_wav_hdr.size_8 += default_wav_hdr.data_size + (sizeof(default_wav_hdr)-8);fseek(fp, 4, 0);fwrite(&default_wav_hdr.size_8, sizeof(default_wav_hdr.size_8), 1, fp); //寫入size_8的值fseek(fp, 40, 0); //將文件指針偏移到存儲data_size值的位置fwrite(&default_wav_hdr.data_size, sizeof(default_wav_hdr.data_size), 1, fp);//寫入data_size的值fclose(fp);PlaySoundA("Demo.wav", NULL, SND_ASYNC);ret = QTTSSessionEnd(sessionID, "normal end");if (MSP_SUCCESS != ret){printf("QTTSSessionEnd failed, error code is : %d", ret);}//退出ret = MSPLogout();if (MSP_SUCCESS != ret){printf("MSPLogout failed, error code is: %d", ret);}system("pause");return 0; }

總結

以上是生活随笔為你收集整理的讯飞C/C++语音合成基础篇的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。