Clarify API/docs by introducing the audio "frame", and add helpers.

The frame is the interleaved set of one sample for each channel.  Add
some docs and some methods that make working with samples a bit more
convenient.

Adjusted QAudioBuffer to use these helper functions and terminology.

Change-Id: I96db48e659561972d6de2aa19893d29f9a828cd3
Reviewed-by: Dmytro Poplavskiy <dmytro.poplavskiy@nokia.com>

1 files changed, 65 insertions, 53 deletions

diff --git a/src/multimedia/audio/qaudiobuffer.cpp b/src/multimedia/audio/qaudiobuffer.cpp
index bde04501f..299a610fa 100644
--- a/src/multimedia/audio/qaudiobuffer.cpp
+++ b/src/multimedia/audio/qaudiobuffer.cpp
@@ -106,18 +106,18 @@ public:
 // Private class to go in .cpp file
 class QMemoryAudioBufferProvider : public QAbstractAudioBuffer {
 public:
-    QMemoryAudioBufferProvider(const void *data, int sampleCount, const QAudioFormat &format, qint64 startTime)
+    QMemoryAudioBufferProvider(const void *data, int frameCount, const QAudioFormat &format, qint64 startTime)
         : mStartTime(startTime)
-        , mSampleCount(sampleCount)
+        , mFrameCount(frameCount)
         , mFormat(format)
     {
-        int numBytes = (sampleCount * format.sampleSize()) / 8;
+        int numBytes = format.bytesForFrames(frameCount);
         if (numBytes > 0) {
             mBuffer = malloc(numBytes);
             if (!mBuffer) {
                 // OOM, if that's likely
                 mStartTime = -1;
-                mSampleCount = 0;
+                mFrameCount = 0;
                 mFormat = QAudioFormat();
             } else {
                 // Allocated, see if we have data to copy
@@ -149,19 +149,19 @@ public:
     void release() {delete this;}
     QAudioFormat format() const {return mFormat;}
     qint64 startTime() const {return mStartTime;}
-    int sampleCount() const {return mSampleCount;}
+    int frameCount() const {return mFrameCount;}
 
     void *constData() const {return mBuffer;}
 
     void *writableData() {return mBuffer;}
     QAbstractAudioBuffer *clone() const
     {
-        return new QMemoryAudioBufferProvider(mBuffer, mSampleCount, mFormat, mStartTime);
+        return new QMemoryAudioBufferProvider(mBuffer, mFrameCount, mFormat, mStartTime);
     }
 
     void *mBuffer;
     qint64 mStartTime;
-    int mSampleCount;
+    int mFrameCount;
     QAudioFormat mFormat;
 };
 
@@ -176,7 +176,7 @@ QAudioBufferPrivate *QAudioBufferPrivate::clone()
         QAbstractAudioBuffer *abuf = mProvider->clone();
 
         if (!abuf) {
-            abuf = new QMemoryAudioBufferProvider(mProvider->constData(), mProvider->sampleCount(), mProvider->format(), mProvider->startTime());
+            abuf = new QMemoryAudioBufferProvider(mProvider->constData(), mProvider->frameCount(), mProvider->format(), mProvider->startTime());
         }
 
         if (abuf) {
@@ -236,7 +236,7 @@ QAudioBuffer::QAudioBuffer(const QAudioBuffer &other)
     and sizes of the samples are interpreted from the \a data.
 
     If the supplied \a data is not an integer multiple of the
-    calculated sample size, the excess data will not be used.
+    calculated frame size, the excess data will not be used.
 
     This audio buffer will copy the contents of \a data.
 
@@ -247,25 +247,25 @@ QAudioBuffer::QAudioBuffer(const QAudioBuffer &other)
 QAudioBuffer::QAudioBuffer(const QByteArray &data, const QAudioFormat &format, qint64 startTime)
 {
     if (format.isValid()) {
-        int sampleCount = (data.size() * 8) / format.sampleSize(); // truncate
-        d = new QAudioBufferPrivate(new QMemoryAudioBufferProvider(data.constData(), sampleCount, format, startTime));
+        int frameCount = format.framesForBytes(data.size());
+        d = new QAudioBufferPrivate(new QMemoryAudioBufferProvider(data.constData(), frameCount, format, startTime));
     } else
         d = 0;
 }
 
 /*!
-    Creates a new audio buffer with space for \a numSamples samples of
-    the given \a format.  The samples will be initialized to the default
-    for the format.
+    Creates a new audio buffer with space for \a numFrames frames of
+    the given \a format.  The individual samples will be initialized
+    to the default for the format.
 
     \a startTime (in microseconds) indicates when this buffer
     starts in the stream.
     If this buffer is not part of a stream, set it to -1.
  */
-QAudioBuffer::QAudioBuffer(int numSamples, const QAudioFormat &format, qint64 startTime)
+QAudioBuffer::QAudioBuffer(int numFrames, const QAudioFormat &format, qint64 startTime)
 {
     if (format.isValid())
-        d = new QAudioBufferPrivate(new QMemoryAudioBufferProvider(0, numSamples, format, startTime));
+        d = new QAudioBufferPrivate(new QMemoryAudioBufferProvider(0, numFrames, format, startTime));
     else
         d = 0;
 }
@@ -292,13 +292,13 @@ QAudioBuffer::~QAudioBuffer()
 
 /*!
     Returns true if this is a valid buffer.  A valid buffer
-    has more than zero samples in it and a valid format.
+    has more than zero frames in it and a valid format.
  */
 bool QAudioBuffer::isValid() const
 {
     if (!d || !d->mProvider)
         return false;
-    return d->mProvider->format().isValid() && (d->mProvider->sampleCount() > 0);
+    return d->mProvider->format().isValid() && (d->mProvider->frameCount() > 0);
 }
 
 /*!
@@ -306,7 +306,7 @@ bool QAudioBuffer::isValid() const
 
     Several properties of this format influence how
     the \l duration() or \l byteCount() are calculated
-    from the \l sampleCount().
+    from the \l frameCount().
  */
 QAudioFormat QAudioBuffer::format() const
 {
@@ -316,6 +316,19 @@ QAudioFormat QAudioBuffer::format() const
 }
 
 /*!
+    Returns the number of complete audio frames in this buffer.
+
+    An audio frame is an interleaved set of one sample per channel
+    for the same instant in time.
+*/
+int QAudioBuffer::frameCount() const
+{
+    if (!isValid())
+        return 0;
+    return d->mProvider->frameCount();
+}
+
+/*!
     Returns the number of samples in this buffer.
 
     If the format of this buffer has multiple channels,
@@ -323,12 +336,15 @@ QAudioFormat QAudioBuffer::format() const
     that a stereo buffer with 1000 samples in total will
     have 500 left samples and 500 right samples (interleaved),
     and this function will return 1000.
- */
+
+    \sa frameCount()
+*/
 int QAudioBuffer::sampleCount() const
 {
     if (!isValid())
         return 0;
-    return d->mProvider->sampleCount();
+
+    return frameCount() * format().channelCount();
 }
 
 /*!
@@ -337,21 +353,17 @@ int QAudioBuffer::sampleCount() const
 int QAudioBuffer::byteCount() const
 {
     const QAudioFormat f(format());
-    return (f.sampleSize() * sampleCount()) / 8; // sampleSize is in bits
+    return format().bytesForFrames(frameCount());
 }
 
 /*!
     Returns the duration of audio in this buffer, in microseconds.
 
-    This depends on the /l format(), and the \l sampleCount().
+    This depends on the /l format(), and the \l frameCount().
 */
 qint64 QAudioBuffer::duration() const
 {
-    int divisor = format().sampleRate() * format().channelCount();
-    if (divisor > 0)
-        return (sampleCount() * 1000000LL) / divisor;
-    else
-        return 0;
+    return format().durationForFrames(frameCount());
 }
 
 /*!
@@ -471,7 +483,7 @@ void *QAudioBuffer::data()
     }
 
     // Wasn't writable, so turn it into a memory provider
-    QAbstractAudioBuffer *memBuffer = new QMemoryAudioBufferProvider(constData(), sampleCount(), format(), startTime());
+    QAbstractAudioBuffer *memBuffer = new QMemoryAudioBufferProvider(constData(), frameCount(), format(), startTime());
 
     if (memBuffer) {
         d->mProvider->release();
@@ -487,35 +499,35 @@ void *QAudioBuffer::data()
 // Template helper classes worth documenting
 
 /*!
-    \class QAudioBuffer::StereoSampleDefault
+    \class QAudioBuffer::StereoFrameDefault
     \internal
 
     Just a trait class for the default value.
 */
 
 /*!
-    \class QAudioBuffer::StereoSample
-    \brief The StereoSample class provides a simple wrapper for a stereo audio sample.
+    \class QAudioBuffer::StereoFrame
+    \brief The StereoFrame class provides a simple wrapper for a stereo audio frame.
     \inmodule QtMultimedia
     \ingroup multimedia
     \ingroup multimedia_audio
 
     This templatized structure lets you treat a block of individual samples as an
-    interleaved stereo stream.  This is most useful when used with the templatized
+    interleaved stereo stream frame.  This is most useful when used with the templatized
     \l {QAudioBuffer::data()}{data()} functions of QAudioBuffer.  Generally the data
     is accessed as a pointer, so no copying should occur.
 
     There are some predefined instantiations of this template for working with common
     stereo sample depths in a convenient way.
 
-    This structure has \e left and \e right members for accessing individual channel data.
+    This frame structure has \e left and \e right members for accessing individual channel data.
 
     For example:
     \code
     // Assuming 'buffer' is an unsigned 16 bit stereo buffer..
-    QAudioBuffer::S16U *sample = buffer->data<QAudioBuffer::S16U>();
-    for (int i=0; i < buffer->sampleCount() / 2; i++) {
-        qSwap(sample[i].left, sample[i].right);
+    QAudioBuffer::S16U *frames = buffer->data<QAudioBuffer::S16U>();
+    for (int i=0; i < buffer->frameCount(); i++) {
+        qSwap(frames[i].left, frames[i].right);
     }
     \endcode
 
@@ -523,77 +535,77 @@ void *QAudioBuffer::data()
 */
 
 /*!
-    \fn QAudioBuffer::StereoSample::StereoSample()
+    \fn QAudioBuffer::StereoFrame::StereoFrame()
 
-    Constructs a new sample with the "silent" value for this
+    Constructs a new frame with the "silent" value for this
     sample format (0 for signed formats and floats, 0x8* for unsigned formats).
 */
 
 /*!
-    \fn QAudioBuffer::StereoSample::StereoSample(T leftSample, T rightSample)
+    \fn QAudioBuffer::StereoFrame::StereoFrame(T leftSample, T rightSample)
 
-    Constructs a new sample with the supplied \a leftSample and \a rightSample values.
+    Constructs a new frame with the supplied \a leftSample and \a rightSample values.
 */
 
 /*!
-    \fn QAudioBuffer::StereoSample::operator=(const StereoSample &other)
+    \fn QAudioBuffer::StereoFrame::operator=(const StereoFrame &other)
 
-    Assigns \a other to this sample.
+    Assigns \a other to this frame.
  */
 
 
 /*!
-    \fn QAudioBuffer::StereoSample::average() const
+    \fn QAudioBuffer::StereoFrame::average() const
 
     Returns the arithmetic average of the left and right samples.
  */
 
-/*! \fn QAudioBuffer::StereoSample::clear()
+/*! \fn QAudioBuffer::StereoFrame::clear()
 
-    Sets the values of this sample to the "silent" value.
+    Sets the values of this frame to the "silent" value.
 */
 
 /*!
-    \variable QAudioBuffer::StereoSample::left
+    \variable QAudioBuffer::StereoFrame::left
     \brief the left sample
 */
 
 /*!
-    \variable QAudioBuffer::StereoSample::right
+    \variable QAudioBuffer::StereoFrame::right
     \brief the right sample
 */
 
 /*!
     \typedef QAudioBuffer::S8U
-    \relates QAudioBuffer::StereoSample
+    \relates QAudioBuffer::StereoFrame
 
     This is a predefined specialization for an unsigned stereo 8 bit sample.  Each
     channel is an \e {unsigned char}.
 */
 /*!
     \typedef QAudioBuffer::S8S
-    \relates QAudioBuffer::StereoSample
+    \relates QAudioBuffer::StereoFrame
 
     This is a predefined specialization for a signed stereo 8 bit sample.  Each
     channel is a \e {signed char}.
 */
 /*!
     \typedef QAudioBuffer::S16U
-    \relates QAudioBuffer::StereoSample
+    \relates QAudioBuffer::StereoFrame
 
     This is a predefined specialization for an unsigned stereo 16 bit sample.  Each
     channel is an \e {unsigned short}.
 */
 /*!
     \typedef QAudioBuffer::S16S
-    \relates QAudioBuffer::StereoSample
+    \relates QAudioBuffer::StereoFrame
 
     This is a predefined specialization for a signed stereo 16 bit sample.  Each
     channel is a \e {signed short}.
 */
 /*!
     \typedef QAudioBuffer::S32F
-    \relates QAudioBuffer::StereoSample
+    \relates QAudioBuffer::StereoFrame
 
     This is a predefined specialization for an 32 bit float sample.  Each
     channel is a \e float.