devices/devicemocap/speech/kinectaudiostream.cxx

/***************************************************************************************
 Autodesk(R) Open Reality(R) Samples
 
 (C) 2013 Autodesk, Inc. and/or its licensors
 All rights reserved.
 
 AUTODESK SOFTWARE LICENSE AGREEMENT
 Autodesk, Inc. licenses this Software to you only upon the condition that 
 you accept all of the terms contained in the Software License Agreement ("Agreement") 
 that is embedded in or that is delivered with this Software. By selecting 
 the "I ACCEPT" button at the end of the Agreement or by copying, installing, 
 uploading, accessing or using all or any portion of the Software you agree 
 to enter into the Agreement. A contract is then formed between Autodesk and 
 either you personally, if you acquire the Software for yourself, or the company 
 or other legal entity for which you are acquiring the software.
 
 AUTODESK, INC., MAKES NO WARRANTY, EITHER EXPRESS OR IMPLIED, INCLUDING BUT 
 NOT LIMITED TO ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR 
 PURPOSE REGARDING THESE MATERIALS, AND MAKES SUCH MATERIALS AVAILABLE SOLELY ON AN 
 "AS-IS" BASIS.
 
 IN NO EVENT SHALL AUTODESK, INC., BE LIABLE TO ANYONE FOR SPECIAL, COLLATERAL, 
 INCIDENTAL, OR CONSEQUENTIAL DAMAGES IN CONNECTION WITH OR ARISING OUT OF PURCHASE 
 OR USE OF THESE MATERIALS. THE SOLE AND EXCLUSIVE LIABILITY TO AUTODESK, INC., 
 REGARDLESS OF THE FORM OF ACTION, SHALL NOT EXCEED THE PURCHASE PRICE OF THE 
 MATERIALS DESCRIBED HEREIN.
 
 Autodesk, Inc., reserves the right to revise and improve its products as it sees fit.
 
 Autodesk and Open Reality are registered trademarks or trademarks of Autodesk, Inc., 
 in the U.S.A. and/or other countries. All other brand names, product names, or 
 trademarks belong to their respective holders. 
 
 GOVERNMENT USE
 Use, duplication, or disclosure by the U.S. Government is subject to restrictions as 
 set forth in FAR 12.212 (Commercial Computer Software-Restricted Rights) and 
 DFAR 227.7202 (Rights in Technical Data and Computer Software), as applicable. 
 Manufacturer is Autodesk, Inc., 10 Duke Street, Montreal, Quebec, Canada, H3C 2L7.
***************************************************************************************/

#include "stdafx.h"
#include "kinectaudiostream.h"

#include <stdio.h>

KinectAudioStream::KinectAudioStream(IMediaObject *pKinectDmo) :
      m_cRef(1),
      m_CurrentWriteBuffer(NULL),
      m_CurrentReadBuffer(NULL),
      m_CurrentReadBufferIndex(0),
      m_BytesRead(0),
      m_hStopEvent(NULL),
      m_hDataReady(NULL),
      m_hCaptureThread(NULL)
{
    pKinectDmo->AddRef();
    m_pKinectDmo = pKinectDmo;
    InitializeCriticalSection(&m_Lock);
}

KinectAudioStream::~KinectAudioStream()
{
    SafeRelease(m_pKinectDmo);
    DeleteCriticalSection(&m_Lock);
}

HRESULT KinectAudioStream::StartCapture()
{
    HRESULT hr = S_OK;

    m_hStopEvent = CreateEvent( NULL, TRUE, FALSE, NULL );
    m_hDataReady = CreateEvent( NULL, FALSE, FALSE, NULL );
    m_BytesRead = 0;

    for (UINT i = 0; i < NumBuffers; i++)
    {
        CStaticMediaBuffer *pBuf = new CStaticMediaBuffer();
        m_BufferPool.push(pBuf);
    }

    m_CurrentWriteBuffer = NULL;

    m_hCaptureThread = CreateThread(NULL, 0, CaptureThread, this, 0, NULL);

    return hr;
}

HRESULT KinectAudioStream::StopCapture()
{
    HRESULT hr = S_OK;
    if ( NULL != m_hStopEvent )
    {
        // Signal the thread
        SetEvent(m_hStopEvent);

        // Wait for thread to stop
        if ( NULL != m_hCaptureThread )
        {
            WaitForSingleObject( m_hCaptureThread, INFINITE );
            CloseHandle( m_hCaptureThread );
            m_hCaptureThread = NULL;
        }
        CloseHandle( m_hStopEvent );
        m_hStopEvent = NULL;
    }

    if (NULL != m_hDataReady)
    {
        SetEvent(m_hDataReady);
        CloseHandle(m_hDataReady);
        m_hDataReady = NULL;
    }

    // Release all buffers to buffer pool and then free all buffers in pool
    ReleaseAllBuffers();
    while (!m_BufferPool.empty())
    {
        CStaticMediaBuffer* mediaBuffer = m_BufferPool.top();
        delete mediaBuffer;
        m_BufferPool.pop();
    }

    return hr;
}

// IStream methods
STDMETHODIMP KinectAudioStream::Read(void *pBuffer, ULONG cbBuffer, ULONG *pcbRead)
{
    HRESULT hr = S_OK;

   if (pcbRead == NULL)
   {
       return E_INVALIDARG;
   }

    ULONG bytesPendingToRead = cbBuffer;
    while (bytesPendingToRead > 0 && IsCapturing())
    {
        ReadOneBuffer((BYTE**)&pBuffer, &bytesPendingToRead);

        if (NULL == m_CurrentReadBuffer) //no data, wait ...
        {
            WaitForSingleObject(m_hDataReady, INFINITE);
        }
    }
    ULONG bytesRead = cbBuffer - bytesPendingToRead;
    m_BytesRead += bytesRead;

    *pcbRead = bytesRead;

    return hr;
}

STDMETHODIMP KinectAudioStream::Write(const void *,ULONG,ULONG *)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::Seek(LARGE_INTEGER dlibMove,DWORD dwOrigin, ULARGE_INTEGER *plibNewPosition )   
{
    dwOrigin = dwOrigin;
    if (plibNewPosition != NULL)
    {
        plibNewPosition->QuadPart = m_BytesRead + dlibMove.QuadPart;
    }
    return S_OK;
}

STDMETHODIMP KinectAudioStream::SetSize(ULARGE_INTEGER)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::CopyTo(IStream *,ULARGE_INTEGER,ULARGE_INTEGER *,ULARGE_INTEGER *)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::Commit(DWORD)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::Revert()
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::LockRegion(ULARGE_INTEGER,ULARGE_INTEGER,DWORD)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::UnlockRegion(ULARGE_INTEGER,ULARGE_INTEGER,DWORD)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::Stat(STATSTG *,DWORD)
{
    return E_NOTIMPL;
}

STDMETHODIMP KinectAudioStream::Clone(IStream **)
{
    return E_NOTIMPL;
}

// Private KinectAudioStream methods

CStaticMediaBuffer *KinectAudioStream::GetWriteBuffer()
{
    CStaticMediaBuffer *pBuf = NULL;

    EnterCriticalSection(&m_Lock);

    //Get a free buffer if available. Otherwise, get the oldest buffer
    //from the read queue. This is a way of overwriting the oldest data
    if (m_BufferPool.size() > 0)
    {   
        pBuf = m_BufferPool.top();
        m_BufferPool.pop();
        pBuf->SetLength(0);
    }                           
    else if (m_ReadBufferQueue.size() > 0)
    {
        pBuf = m_ReadBufferQueue.front();
        m_ReadBufferQueue.pop();
        pBuf->SetLength(0);
    }

    LeaveCriticalSection(&m_Lock);

    return pBuf;
}


void KinectAudioStream::ReleaseBuffer(CStaticMediaBuffer* pBuffer)
{
    if (pBuffer != NULL)
    {
        EnterCriticalSection(&m_Lock);
        pBuffer->SetLength(0);
        m_BufferPool.push(pBuffer);
        LeaveCriticalSection(&m_Lock);
    }
}

void KinectAudioStream::ReleaseAllBuffers()
{
    EnterCriticalSection(&m_Lock);
    while (m_ReadBufferQueue.size() > 0)
    {
        CStaticMediaBuffer *pBuf = m_ReadBufferQueue.front();
        m_ReadBufferQueue.pop();
        ReleaseBuffer(pBuf);
    }
    if (m_CurrentReadBuffer != NULL)
    {
        ReleaseBuffer(m_CurrentReadBuffer);
    }

    m_CurrentReadBufferIndex = 0;
    m_CurrentReadBuffer = NULL;
    LeaveCriticalSection(&m_Lock);
}

void KinectAudioStream::QueueCapturedData(BYTE *pData, UINT cbData)
{
    BYTE *pWriteData = NULL;
    DWORD cbWriteData = 0;
    DWORD cbMaxLength = 0;

    if (cbData <= 0)
    {
        return;
    }

    if (NULL == m_CurrentWriteBuffer)
    {
        m_CurrentWriteBuffer = GetWriteBuffer();
    }

    m_CurrentWriteBuffer->GetBufferAndLength(&pWriteData, &cbWriteData);
    m_CurrentWriteBuffer->GetMaxLength(&cbMaxLength);

    if (cbWriteData + cbData < cbMaxLength)
    {
        memcpy(pWriteData + cbWriteData, pData, cbData);
        m_CurrentWriteBuffer->SetLength(cbWriteData + cbData);
    }
    else
    {
        QueueCapturedBuffer(m_CurrentWriteBuffer);

        m_CurrentWriteBuffer = GetWriteBuffer();
        m_CurrentWriteBuffer->GetBufferAndLength(&pWriteData, &cbWriteData);

        memcpy(pWriteData, pData, cbData);
        m_CurrentWriteBuffer->SetLength(cbData);
    }
}

void KinectAudioStream::QueueCapturedBuffer(CStaticMediaBuffer *pBuffer)
{
    EnterCriticalSection(&m_Lock);
    m_ReadBufferQueue.push(pBuffer);
    SetEvent(m_hDataReady);

    LeaveCriticalSection(&m_Lock);
}

void KinectAudioStream::ReadOneBuffer(BYTE **ppbData, ULONG* pcbData)
{
    EnterCriticalSection(&m_Lock);

    //Do we already have a buffer we are reading from? Otherwise grab one from the queue
    if (m_CurrentReadBuffer == NULL) 
    {
        if(m_ReadBufferQueue.size() != 0)
        {
            m_CurrentReadBuffer = m_ReadBufferQueue.front();
            m_ReadBufferQueue.pop();
        }
    }

    if (m_CurrentReadBuffer != NULL) 
    {
        //Copy as much data as we can or need
        BYTE *pData = NULL;
        DWORD dwDataLength = 0;
        m_CurrentReadBuffer->GetBufferAndLength(&pData, &dwDataLength);

        ULONG cbToCopy = min(dwDataLength - m_CurrentReadBufferIndex, *pcbData);
        memcpy(*ppbData, pData + m_CurrentReadBufferIndex, cbToCopy);
        *ppbData = (*ppbData)+cbToCopy;
        *pcbData = (*pcbData)-cbToCopy;
        m_CurrentReadBufferIndex += cbToCopy;

        //If we are done with this buffer put it back in the queue
        if (m_CurrentReadBufferIndex >= dwDataLength)
        {
            ReleaseBuffer(m_CurrentReadBuffer);
            m_CurrentReadBuffer = NULL;
            m_CurrentReadBufferIndex = 0;

            if(m_ReadBufferQueue.size() != 0)
            {
                m_CurrentReadBuffer = m_ReadBufferQueue.front();
                m_ReadBufferQueue.pop();
            }
        }
    }

    LeaveCriticalSection(&m_Lock);
}

DWORD WINAPI KinectAudioStream::CaptureThread(LPVOID pParam)
{
    KinectAudioStream *pthis = (KinectAudioStream *) pParam;
    return pthis->CaptureThread();
}

DWORD WINAPI KinectAudioStream::CaptureThread()
{
    HANDLE mmHandle = NULL;
    DWORD mmTaskIndex = 0;
    HRESULT hr = S_OK;
    bool bContinue = true;
    BYTE *pbOutputBuffer = NULL;
    CStaticMediaBuffer outputBuffer;
    DMO_OUTPUT_DATA_BUFFER OutputBufferStruct = {0};
    OutputBufferStruct.pBuffer = &outputBuffer;
    DWORD dwStatus = 0;
    ULONG cbProduced = 0;

    // Set high priority to avoid getting preempted while capturing sound
    mmHandle = AvSetMmThreadCharacteristics(L"Audio", &mmTaskIndex);

    while (bContinue)
    {
        if (WaitForSingleObject(m_hStopEvent, 0) == WAIT_OBJECT_0)
        {
            bContinue = false;
            continue;
        }

        do
        {
            outputBuffer.Init(0);
            OutputBufferStruct.dwStatus = 0;
            hr = m_pKinectDmo->ProcessOutput(0, 1, &OutputBufferStruct, &dwStatus);
            if (FAILED(hr))
            {
                bContinue = false;
                break;
            }

            if (hr == S_FALSE)
            {
                cbProduced = 0;
            }
            else
            {
                outputBuffer.GetBufferAndLength(&pbOutputBuffer, &cbProduced);
            }

            // Queue audio data to be read by IStream client
            if (cbProduced > 0)
            {
                QueueCapturedData(pbOutputBuffer, cbProduced);
            }
        } while (OutputBufferStruct.dwStatus & DMO_OUTPUT_DATA_BUFFERF_INCOMPLETE);

        Sleep(10); //sleep 10ms
    }

    SetEvent(m_hDataReady);
    AvRevertMmThreadCharacteristics(mmHandle);

    if (FAILED(hr))
    {
        return 0;
    }

    return 1;
}