Why no voice?

My english is poor,so I beg your pardon.
My codes are about playing voice,and it doesn't return error,
but it has no voice,why?
Look:

Code:

#define OUT_BUFFER_SIZE 6000
void main()  
{

	HWAVEOUT     hWaveOut ;
	PWAVEHDR     pWaveHdr ;
	WAVEFORMATEX waveformat ;
	PBYTE        pBuffer;
	
	waveformat.wFormatTag      = WAVE_FORMAT_PCM ;
	waveformat.nChannels       = 2 ;
	waveformat.nSamplesPerSec  = 44100; ;
	waveformat.nAvgBytesPerSec = 44100*16*2/8 ;
	waveformat.nBlockAlign     = 16*2/8 ;
	waveformat.wBitsPerSample  = 16 ;
	waveformat.cbSize          = 0 ;
	
	
	
	// open   
	if (waveOutOpen (&hWaveOut, WAVE_MAPPER, &waveformat,0, 0, 0)!= MMSYSERR_NOERROR)
	{
		printf("waveOutOpen error \n");
		return ;
	}
	
    
	pWaveHdr =(PWAVEHDR)malloc (sizeof (WAVEHDR)) ;
        pBuffer  = (PBYTE)malloc (OUT_BUFFER_SIZE) ;
	
	// prepare buffer   
        pWaveHdr->lpData          = (char*)pBuffer ;
	pWaveHdr->dwBufferLength  = OUT_BUFFER_SIZE ;
	pWaveHdr->dwBytesRecorded = 0 ;
	pWaveHdr->dwUser          = 0 ;
	pWaveHdr->dwFlags         = 0 ;
	pWaveHdr->dwLoops         = 1 ;
	pWaveHdr->lpNext          = NULL ;
	pWaveHdr->reserved        = 0 ;
	
       if(waveOutPrepareHeader(hWaveOut, pWaveHdr,sizeof (WAVEHDR))!=MMSYSERR_NOERROR) 
	{
		printf("waveOutPrepareHeader error \n");
		return ;
	}
	// play   
	memset(pBuffer,70,OUT_BUFFER_SIZE);
	if(waveOutWrite (hWaveOut, pWaveHdr, sizeof (WAVEHDR)) !=MMSYSERR_NOERROR) 
	{
		printf("waveOutWrite error \n");
		return ;
	}
	
	// clean   
       if(waveOutReset(hWaveOut) !=MMSYSERR_NOERROR)
	{
		printf("waveOutReset error \n");
		return ;
	}

      if(waveOutUnprepareHeader (hWaveOut, pWaveHdr, sizeof (WAVEHDR)) !=MMSYSERR_NOERROR)
	{
		printf("waveOutUnprepareHeader error \n");
		return ;
	}

       if(waveOutClose(hWaveOut)!=MMSYSERR_NOERROR)
	{
		printf("waveOutClose error \n");
		return ;
	}
}

Pulse-code modulation - Wikipedia, the free encyclopedia
You get noise, by having changes in the data.

> memset(pBuffer,70,OUT_BUFFER_SIZE);
6K of constant values is boring.
Try filling it with random values, or maybe a chunk of a suitably encoded wav file.

Is your "voice" data recorded in a wave file of some kind?

If so, just use the PlaySound() API.

If you are trying to generate voice programattically by wave output, you are in for one BIG job... Not only will you need to assemble phonetic samples in several variations, you will need to play those samples in response to text... that is you will need to parse "This is a test" (for example) into it's 8 phonetic components and play them one after the other. Major software houses have worked on this technology for years and are only now getting it so the voice sounds natural enough to not irritate people...

Microsoft Speech Technologies

Originally Posted by Salem

Pulse-code modulation - Wikipedia, the free encyclopedia
You get noise, by having changes in the data.

> memset(pBuffer,70,OUT_BUFFER_SIZE);
6K of constant values is boring.
Try filling it with random values, or maybe a chunk of a suitably encoded wav file.

6000 values at 44100 values per second will give him about 0.13 seconds of sound... the very best he'll get from that is a click from the speaker.

60,000 random values could produce a buzzing sound for about 1.3 seconds.

Quite - in fact half that time, since the header specifies 16-bit samples.
And half again, since the header specifies stereo output.

but I define OUT_BUFFER_SIZE,look:
#define OUT_BUFFER_SIZE (44100*16*2/8*10)
it has no voice,too.
Why?

Originally Posted by leetow2003

but I define OUT_BUFFER_SIZE,look:
#define OUT_BUFFER_SIZE (44100*16*2/8*10)
it has no voice,too.
Why?

Tell me something... when you opened the wave output... exactly what did you expect to happen?

Originally Posted by CommonTater

Tell me something... when you opened the wave output... exactly what did you expect to happen?

I want to listen to voice,and when I
#define OUT_BUFFER_SIZE (44100*16*2/8*10)
I want to listen for 10 seconds,but I don't listen,
why?

Like I said - did you write in any valid PCM data or did you just fill it with a constant again?

And I think it should be 44100 * 2 * 2 * 10

That's nSamplesPerSec * nChannels * wBitsPerSample/8 * seconds.

Originally Posted by leetow2003

I want to listen to voice,and when I
#define OUT_BUFFER_SIZE (44100*16*2/8*10)
I want to listen for 10 seconds,but I don't listen,
why?

Ok let me ask my question one more time...

EXACTLY what did you expect to happen when you opened the waveout?

Did you expect it to just start talking?