Add support for WaveShaperNode.oversample

Source/core/platform/audio/DownSampler.cpp

+/*
+ * Copyright (C) 2013 Google Inc. All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1.  Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ * 2.  Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ * 3.  Neither the name of Apple Computer, Inc. ("Apple") nor the names of
+ *     its contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY APPLE AND ITS CONTRIBUTORS "AS IS" AND ANY
+ * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL APPLE OR ITS CONTRIBUTORS BE LIABLE FOR ANY
+ * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+ * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "config.h"
+
+#if ENABLE(WEB_AUDIO)
+
+#include "core/platform/audio/DownSampler.h"
+
+#include "wtf/MathExtras.h"
+
+namespace WebCore {
+
+DownSampler::DownSampler(size_t inputBlockSize)
+    : m_inputBlockSize(inputBlockSize)
+    , m_reducedKernel(DefaultKernelSize / 2)
+    , m_tempBuffer(inputBlockSize / 2)
+    , m_inputBuffer(inputBlockSize * 2)
+    , m_convolver(inputBlockSize / 2) // runs at 1/2 source sample-rate
+{
+    initializeKernel();
+}
+
+void DownSampler::initializeKernel()
+{
+    // Blackman window parameters.
+    double alpha = 0.16;
+    double a0 = 0.5 * (1.0 - alpha);
+    double a1 = 0.5;
+    double a2 = 0.5 * alpha;
+
+    int n = DefaultKernelSize;
+    int halfSize = n / 2;
+
+    // Half-band filter.
+    double sincScaleFactor = 0.5;
+
+    // Compute only the odd terms because the even ones are zero, except
+    // right in the middle at halfSize, which is 0.5 and we'll handle specially during processing
+    // after doing the main convolution using m_reducedKernel.
+    for (int i = 1; i < n; i += 2) {
+        // Compute the sinc() with offset.
+        double s = sincScaleFactor * piDouble * (i - halfSize);
+        double sinc = !s ? 1.0 : sin(s) / s;
+        sinc *= sincScaleFactor;
+
+        // Compute Blackman window, matching the offset of the sinc().
+        double x = static_cast<double>(i) / n;
+        double window = a0 - a1 * cos(2.0 * piDouble * x) + a2 * cos(4.0 * piDouble * x);
+
+        // Window the sinc() function.
+        // Then store only the odd terms in the kernel.
+        // In a sense, this is shifting forward in time by one sample-frame at the destination sample-rate.
+        m_reducedKernel[(i - 1) / 2] = sinc * window;
+    }
+}
+
+void DownSampler::process(const float* sourceP, float* destP, size_t sourceFramesToProcess)
+{
+    ASSERT(sourceFramesToProcess == m_inputBlockSize);
+    if (sourceFramesToProcess != m_inputBlockSize)
+        return;
+
+    size_t destFramesToProcess = sourceFramesToProcess / 2;
+
+    ASSERT(destFramesToProcess == m_tempBuffer.size());
+    if (destFramesToProcess != m_tempBuffer.size())
+        return;
+
+    // Copy source samples to 2nd half of input buffer.
+    float* inputP = m_inputBuffer.data() + m_inputBlockSize;
+    memcpy(inputP, sourceP, sizeof(float) * sourceFramesToProcess);
+
+    // Copy the odd sample-frames from sourceP, delayed by one sample-frame (destination sample-rate)
+    // to match shifting forward in time in m_reducedKernel.
+    float* oddSamplesP = m_tempBuffer.data();
+    for (unsigned i = 0; i < destFramesToProcess; ++i)
+        oddSamplesP[i] = inputP[i * 2 - 1];
+
+    // Actually process oddSamplesP with m_reducedKernel for efficiency.
+    // The theoretical kernel is double this size with 0 values for even terms (except center).
+    m_convolver.process(&m_reducedKernel, oddSamplesP, destP, destFramesToProcess);

[Ken Russell (switch to Gerrit), 2013/05/23 02:29:25]

Checks are needed around the size of the reduced kernel. Currently it's based on
DefaultKernelSize, but the rest of the buffers are sized around the
inputBlockSize. Assertions and run-time checks are needed to make sure these
match up.

[Chris Rogers, 2013/05/24 20:09:04]

Added ASSERTS to check all (I think) necessary conditions.  The input buffer
size is now checked versus halfSize (half of kernel size), which is the
important check for the kernel size since the actual convolver can accept any
kernel size unrelated to the input block size, etc.

+
+    // Now, account for the 0.5 term right in the middle of the kernel.
+    // This amounts to a delay-line of length halfSize (at the source sample-rate),
+    // scaled by 0.5.
+    int halfSize = DefaultKernelSize / 2;
+
+    // Sum into the destination.
+    for (unsigned i = 0; i < destFramesToProcess; ++i)
+        destP[i] += 0.5 * inputP[i * 2 - halfSize];
+
+    // Copy 2nd half of input buffer to 1st half.
+    memcpy(m_inputBuffer.data(), inputP, sizeof(float) * sourceFramesToProcess);

[Ken Russell (switch to Gerrit), 2013/05/23 02:29:25]

I don't understand why m_inputBuffer is needed. It doesn't seem to be exposed
outside this class, and all it's used for is temporary storage for the source
data. Could you please explain this to me?

[Chris Rogers, 2013/05/24 20:09:04]

Added comment in header file for this.

+}
+
+void DownSampler::reset()
+{
+    m_convolver.reset();
+    m_inputBuffer.zero();
+}
+
+size_t DownSampler::latencyFrames() const
+{
+    // Divide by two since this is a linear phase kernel and the delay is at the center of the kernel.
+    return m_reducedKernel.size() / 2;
+}
+
+} // namespace WebCore
+
+#endif // ENABLE(WEB_AUDIO)