diff --git a/dom/media/webaudio/AnalyserNode.cpp b/dom/media/webaudio/AnalyserNode.cpp
index c73d002f9c48..d110cc1c6172 100644
--- a/dom/media/webaudio/AnalyserNode.cpp
+++ b/dom/media/webaudio/AnalyserNode.cpp
@@ -329,7 +329,7 @@ AnalyserNode::GetTimeDomainData(float* aData, size_t aLength)
 
   for (size_t writeIndex = 0; writeIndex < aLength; ) {
     const AudioChunk& chunk = mChunks[readChunk & (CHUNK_COUNT - 1)];
-    const size_t channelCount = chunk.mChannelData.Length();
+    const size_t channelCount = chunk.ChannelCount();
     size_t copyLength =
       std::min<size_t>(aLength - writeIndex, WEBAUDIO_BLOCK_SIZE);
     float* dataOut = &aData[writeIndex];
diff --git a/dom/media/webaudio/AudioDestinationNode.cpp b/dom/media/webaudio/AudioDestinationNode.cpp
index fdb49766085e..2b0651001b0b 100644
--- a/dom/media/webaudio/AudioDestinationNode.cpp
+++ b/dom/media/webaudio/AudioDestinationNode.cpp
@@ -79,7 +79,7 @@ public:
     // Record our input buffer
     MOZ_ASSERT(mWriteIndex < mLength, "How did this happen?");
     const uint32_t duration = std::min(WEBAUDIO_BLOCK_SIZE, mLength - mWriteIndex);
-    const uint32_t inputChannelCount = aInput.mChannelData.Length();
+    const uint32_t inputChannelCount = aInput.ChannelCount();
     for (uint32_t i = 0; i < outputChannelCount; ++i) {
       float* outputData = mBuffer->GetDataForWrite(i) + mWriteIndex;
       if (aInput.IsNull() || i >= inputChannelCount) {
diff --git a/dom/media/webaudio/AudioNodeEngine.cpp b/dom/media/webaudio/AudioNodeEngine.cpp
index 47952bf28884..46c1a3d915c8 100644
--- a/dom/media/webaudio/AudioNodeEngine.cpp
+++ b/dom/media/webaudio/AudioNodeEngine.cpp
@@ -40,7 +40,7 @@ WriteZeroesToAudioBlock(AudioChunk* aChunk, uint32_t aStart, uint32_t aLength)
   if (aLength == 0)
     return;
 
-  for (uint32_t i = 0; i < aChunk->mChannelData.Length(); ++i) {
+  for (uint32_t i = 0; i < aChunk->ChannelCount(); ++i) {
     PodZero(aChunk->ChannelFloatsForWrite(i) + aStart, aLength);
   }
 }
diff --git a/dom/media/webaudio/AudioNodeStream.cpp b/dom/media/webaudio/AudioNodeStream.cpp
index 4758d35d22e2..e6eaf721be81 100644
--- a/dom/media/webaudio/AudioNodeStream.cpp
+++ b/dom/media/webaudio/AudioNodeStream.cpp
@@ -393,20 +393,20 @@ AudioNodeStream::ObtainInputBlock(AudioChunk& aTmpChunk, uint32_t aPortIndex)
 
     inputChunks.AppendElement(chunk);
     outputChannelCount =
-      GetAudioChannelsSuperset(outputChannelCount, chunk->mChannelData.Length());
+      GetAudioChannelsSuperset(outputChannelCount, chunk->ChannelCount());
   }
 
   outputChannelCount = ComputedNumberOfChannels(outputChannelCount);
 
   uint32_t inputChunkCount = inputChunks.Length();
   if (inputChunkCount == 0 ||
-      (inputChunkCount == 1 && inputChunks[0]->mChannelData.Length() == 0)) {
+      (inputChunkCount == 1 && inputChunks[0]->ChannelCount() == 0)) {
     aTmpChunk.SetNull(WEBAUDIO_BLOCK_SIZE);
     return;
   }
 
   if (inputChunkCount == 1 &&
-      inputChunks[0]->mChannelData.Length() == outputChannelCount) {
+      inputChunks[0]->ChannelCount() == outputChannelCount) {
     aTmpChunk = *inputChunks[0];
     return;
   }
@@ -431,7 +431,7 @@ AudioNodeStream::AccumulateInputChunk(uint32_t aInputIndex, const AudioChunk& aC
                                       nsTArray<float>* aDownmixBuffer)
 {
   nsAutoTArray<const float*,GUESS_AUDIO_CHANNELS> channels;
-  UpMixDownMixChunk(&aChunk, aBlock->mChannelData.Length(), channels, *aDownmixBuffer);
+  UpMixDownMixChunk(&aChunk, aBlock->ChannelCount(), channels, *aDownmixBuffer);
 
   for (uint32_t c = 0; c < channels.Length(); ++c) {
     const float* inputData = static_cast<const float*>(channels[c]);
@@ -458,7 +458,7 @@ AudioNodeStream::UpMixDownMixChunk(const AudioChunk* aChunk,
 {
   static const float silenceChannel[WEBAUDIO_BLOCK_SIZE] = {0.f};
 
-  for (uint32_t i = 0; i < aChunk->mChannelData.Length(); i++) {
+  for (uint32_t i = 0; i < aChunk->ChannelCount(); i++) {
     aOutputChannels.AppendElement(static_cast<const float*>(aChunk->mChannelData[i]));
   }
   if (aOutputChannels.Length() < aOutputChannelCount) {
diff --git a/dom/media/webaudio/BiquadFilterNode.cpp b/dom/media/webaudio/BiquadFilterNode.cpp
index e837c017d088..3305eae56456 100644
--- a/dom/media/webaudio/BiquadFilterNode.cpp
+++ b/dom/media/webaudio/BiquadFilterNode.cpp
@@ -168,7 +168,7 @@ public:
 
       PodArrayZero(inputBuffer);
 
-    } else if(mBiquads.Length() != aInput.mChannelData.Length()){
+    } else if(mBiquads.Length() != aInput.ChannelCount()){
       if (mBiquads.IsEmpty()) {
         nsRefPtr<PlayingRefChangeHandler> refchanged =
           new PlayingRefChangeHandler(aStream, PlayingRefChangeHandler::ADDREF);
@@ -179,7 +179,7 @@ public:
       }
 
       // Adjust the number of biquads based on the number of channels
-      mBiquads.SetLength(aInput.mChannelData.Length());
+      mBiquads.SetLength(aInput.ChannelCount());
     }
 
     uint32_t numberOfChannels = mBiquads.Length();
diff --git a/dom/media/webaudio/ChannelMergerNode.cpp b/dom/media/webaudio/ChannelMergerNode.cpp
index 0de9087523c2..a7e42850670c 100644
--- a/dom/media/webaudio/ChannelMergerNode.cpp
+++ b/dom/media/webaudio/ChannelMergerNode.cpp
@@ -33,7 +33,7 @@ public:
     // Get the number of output channels, and allocate it
     size_t channelCount = 0;
     for (uint16_t i = 0; i < InputCount(); ++i) {
-      channelCount += aInput[i].mChannelData.Length();
+      channelCount += aInput[i].ChannelCount();
     }
     if (channelCount == 0) {
       aOutput[0].SetNull(WEBAUDIO_BLOCK_SIZE);
@@ -46,7 +46,7 @@ public:
     size_t channelIndex = 0;
     for (uint16_t i = 0; true; ++i) {
       MOZ_ASSERT(i < InputCount());
-      for (size_t j = 0; j < aInput[i].mChannelData.Length(); ++j) {
+      for (size_t j = 0; j < aInput[i].ChannelCount(); ++j) {
         AudioBlockCopyChannelWithScale(
             static_cast<const float*>(aInput[i].mChannelData[j]),
             aInput[i].mVolume,
diff --git a/dom/media/webaudio/ChannelSplitterNode.cpp b/dom/media/webaudio/ChannelSplitterNode.cpp
index 6c8db5e81fc0..4004a9943e58 100644
--- a/dom/media/webaudio/ChannelSplitterNode.cpp
+++ b/dom/media/webaudio/ChannelSplitterNode.cpp
@@ -32,7 +32,7 @@ public:
 
     aOutput.SetLength(OutputCount());
     for (uint16_t i = 0; i < OutputCount(); ++i) {
-      if (i < aInput[0].mChannelData.Length()) {
+      if (i < aInput[0].ChannelCount()) {
         // Split out existing channels
         AllocateAudioBlock(1, &aOutput[i]);
         AudioBlockCopyChannelWithScale(
diff --git a/dom/media/webaudio/ConvolverNode.cpp b/dom/media/webaudio/ConvolverNode.cpp
index 0c12fe5f3d76..58d11e2c9e4f 100644
--- a/dom/media/webaudio/ConvolverNode.cpp
+++ b/dom/media/webaudio/ConvolverNode.cpp
@@ -131,7 +131,7 @@ public:
     } else {
       if (aInput.mVolume != 1.0f) {
         // Pre-multiply the input's volume
-        uint32_t numChannels = aInput.mChannelData.Length();
+        uint32_t numChannels = aInput.ChannelCount();
         AllocateAudioBlock(numChannels, &input);
         for (uint32_t i = 0; i < numChannels; ++i) {
           const float* src = static_cast<const float*>(aInput.mChannelData[i]);
diff --git a/dom/media/webaudio/DelayBuffer.cpp b/dom/media/webaudio/DelayBuffer.cpp
index d21e36335407..8a1c16aec4f0 100644
--- a/dom/media/webaudio/DelayBuffer.cpp
+++ b/dom/media/webaudio/DelayBuffer.cpp
@@ -29,7 +29,7 @@ void
 DelayBuffer::Write(const AudioChunk& aInputChunk)
 {
   // We must have a reference to the buffer if there are channels
-  MOZ_ASSERT(aInputChunk.IsNull() == !aInputChunk.mChannelData.Length());
+  MOZ_ASSERT(aInputChunk.IsNull() == !aInputChunk.ChannelCount());
 #ifdef DEBUG
   MOZ_ASSERT(!mHaveWrittenBlock);
   mHaveWrittenBlock = true;
@@ -116,7 +116,7 @@ DelayBuffer::ReadChannels(const double aPerFrameDelays[WEBAUDIO_BLOCK_SIZE],
                           uint32_t aFirstChannel, uint32_t aNumChannelsToRead,
                           ChannelInterpretation aChannelInterpretation)
 {
-  uint32_t totalChannelCount = aOutputChunk->mChannelData.Length();
+  uint32_t totalChannelCount = aOutputChunk->ChannelCount();
   uint32_t readChannelsEnd = aFirstChannel + aNumChannelsToRead;
   MOZ_ASSERT(readChannelsEnd <= totalChannelCount);
 
diff --git a/dom/media/webaudio/DynamicsCompressorNode.cpp b/dom/media/webaudio/DynamicsCompressorNode.cpp
index c7f6aecbdffa..2a0a5ad02528 100644
--- a/dom/media/webaudio/DynamicsCompressorNode.cpp
+++ b/dom/media/webaudio/DynamicsCompressorNode.cpp
@@ -103,11 +103,11 @@ public:
       return;
     }
 
-    const uint32_t channelCount = aInput.mChannelData.Length();
+    const uint32_t channelCount = aInput.ChannelCount();
     if (mCompressor->numberOfChannels() != channelCount) {
       // Create a new compressor object with a new channel count
       mCompressor = new WebCore::DynamicsCompressor(aStream->SampleRate(),
-                                                    aInput.mChannelData.Length());
+                                                    aInput.ChannelCount());
     }
 
     StreamTime pos = aStream->GetCurrentPosition();
diff --git a/dom/media/webaudio/GainNode.cpp b/dom/media/webaudio/GainNode.cpp
index ce62d50f325d..ca6f6cd7bdaa 100644
--- a/dom/media/webaudio/GainNode.cpp
+++ b/dom/media/webaudio/GainNode.cpp
@@ -81,7 +81,7 @@ public:
       // First, compute a vector of gains for each track tick based on the
       // timeline at hand, and then for each channel, multiply the values
       // in the buffer with the gain vector.
-      AllocateAudioBlock(aInput.mChannelData.Length(), aOutput);
+      AllocateAudioBlock(aInput.ChannelCount(), aOutput);
 
       // Compute the gain values for the duration of the input AudioChunk
       StreamTime tick = aStream->GetCurrentPosition();
@@ -93,7 +93,7 @@ public:
       }
 
       // Apply the gain to the output buffer
-      for (size_t channel = 0; channel < aOutput->mChannelData.Length(); ++channel) {
+      for (size_t channel = 0; channel < aOutput->ChannelCount(); ++channel) {
         const float* inputBuffer = static_cast<const float*> (aInput.mChannelData[channel]);
         float* buffer = aOutput->ChannelFloatsForWrite(channel);
         AudioBlockCopyChannelWithScale(inputBuffer, computedGain, buffer);
diff --git a/dom/media/webaudio/PannerNode.cpp b/dom/media/webaudio/PannerNode.cpp
index 8dca61ce9105..76b5bf107049 100644
--- a/dom/media/webaudio/PannerNode.cpp
+++ b/dom/media/webaudio/PannerNode.cpp
@@ -323,7 +323,7 @@ PannerNodeEngine::EqualPowerPanningFunction(const AudioChunk& aInput,
                                             AudioChunk* aOutput)
 {
   float azimuth, elevation, gainL, gainR, normalizedAzimuth, distanceGain, coneGain;
-  int inputChannels = aInput.mChannelData.Length();
+  int inputChannels = aInput.ChannelCount();
 
   // If both the listener are in the same spot, and no cone gain is specified,
   // this node is noop.
diff --git a/dom/media/webaudio/PanningUtils.h b/dom/media/webaudio/PanningUtils.h
index dfe7a84afece..4895ec0a1b0c 100644
--- a/dom/media/webaudio/PanningUtils.h
+++ b/dom/media/webaudio/PanningUtils.h
@@ -52,7 +52,7 @@ template<typename T, typename U>
 void ApplyStereoPanning(const AudioChunk& aInput, AudioChunk* aOutput,
                         T aGainL, T aGainR, U aOnLeft)
 {
-  if (aInput.mChannelData.Length() == 1) {
+  if (aInput.ChannelCount() == 1) {
     GainMonoToStereo(aInput, aOutput, aGainL, aGainR);
   } else {
     GainStereoToStereo(aInput, aOutput, aGainL, aGainR, aOnLeft);
diff --git a/dom/media/webaudio/ScriptProcessorNode.cpp b/dom/media/webaudio/ScriptProcessorNode.cpp
index 9749d2c0dfae..c74761ac56f7 100644
--- a/dom/media/webaudio/ScriptProcessorNode.cpp
+++ b/dom/media/webaudio/ScriptProcessorNode.cpp
@@ -315,7 +315,7 @@ public:
         PodZero(writeData, aInput.GetDuration());
       } else {
         MOZ_ASSERT(aInput.GetDuration() == WEBAUDIO_BLOCK_SIZE, "sanity check");
-        MOZ_ASSERT(aInput.mChannelData.Length() == inputChannelCount);
+        MOZ_ASSERT(aInput.ChannelCount() == inputChannelCount);
         AudioBlockCopyChannelWithScale(static_cast<const float*>(aInput.mChannelData[i]),
                                        aInput.mVolume, writeData);
       }
diff --git a/dom/media/webaudio/WaveShaperNode.cpp b/dom/media/webaudio/WaveShaperNode.cpp
index 56465cf053af..d9943191985a 100644
--- a/dom/media/webaudio/WaveShaperNode.cpp
+++ b/dom/media/webaudio/WaveShaperNode.cpp
@@ -219,7 +219,7 @@ public:
                             AudioChunk* aOutput,
                             bool* aFinished) override
   {
-    uint32_t channelCount = aInput.mChannelData.Length();
+    uint32_t channelCount = aInput.ChannelCount();
     if (!mCurve.Length() || !channelCount) {
       // Optimize the case where we don't have a curve buffer,
       // or the input is null.
diff --git a/dom/media/webaudio/blink/DynamicsCompressor.cpp b/dom/media/webaudio/blink/DynamicsCompressor.cpp
index 746356d1d018..2b6eb85e1c8b 100644
--- a/dom/media/webaudio/blink/DynamicsCompressor.cpp
+++ b/dom/media/webaudio/blink/DynamicsCompressor.cpp
@@ -154,8 +154,8 @@ void DynamicsCompressor::process(const AudioChunk* sourceChunk, AudioChunk* dest
     // It's because we internally match sourceChannels's size to destinationBus by channel up/down mix. Thus we need numberOfChannels
     // to do the loop work for both m_sourceChannels and m_destinationChannels.
 
-    unsigned numberOfChannels = destinationChunk->mChannelData.Length();
-    unsigned numberOfSourceChannels = sourceChunk->mChannelData.Length();
+    unsigned numberOfChannels = destinationChunk->ChannelCount();
+    unsigned numberOfSourceChannels = sourceChunk->ChannelCount();
 
     MOZ_ASSERT(numberOfChannels == m_numberOfChannels && numberOfSourceChannels);
 
diff --git a/dom/media/webaudio/blink/HRTFPanner.cpp b/dom/media/webaudio/blink/HRTFPanner.cpp
index 54f08e21bf54..4bccab22a8c8 100644
--- a/dom/media/webaudio/blink/HRTFPanner.cpp
+++ b/dom/media/webaudio/blink/HRTFPanner.cpp
@@ -132,13 +132,13 @@ void HRTFPanner::pan(double desiredAzimuth, double elevation, const AudioChunk*
 {
 #ifdef DEBUG
     unsigned numInputChannels =
-        inputBus->IsNull() ? 0 : inputBus->mChannelData.Length();
+        inputBus->IsNull() ? 0 : inputBus->ChannelCount();
 
     MOZ_ASSERT(numInputChannels <= 2);
     MOZ_ASSERT(inputBus->mDuration == WEBAUDIO_BLOCK_SIZE);
 #endif
 
-    bool isOutputGood = outputBus && outputBus->mChannelData.Length() == 2 && outputBus->mDuration == WEBAUDIO_BLOCK_SIZE;
+    bool isOutputGood = outputBus && outputBus->ChannelCount() == 2 && outputBus->mDuration == WEBAUDIO_BLOCK_SIZE;
     MOZ_ASSERT(isOutputGood);
 
     if (!isOutputGood) {
diff --git a/dom/media/webaudio/blink/Reverb.cpp b/dom/media/webaudio/blink/Reverb.cpp
index a4f380e14d54..326511c5837e 100644
--- a/dom/media/webaudio/blink/Reverb.cpp
+++ b/dom/media/webaudio/blink/Reverb.cpp
@@ -154,7 +154,7 @@ void Reverb::process(const AudioChunk* sourceBus, AudioChunk* destinationBus, si
 {
     // Do a fairly comprehensive sanity check.
     // If these conditions are satisfied, all of the source and destination pointers will be valid for the various matrixing cases.
-    bool isSafeToProcess = sourceBus && destinationBus && sourceBus->mChannelData.Length() > 0 && destinationBus->mChannelData.Length() > 0
+    bool isSafeToProcess = sourceBus && destinationBus && sourceBus->ChannelCount() > 0 && destinationBus->mChannelData.Length() > 0
         && framesToProcess <= MaxFrameSize && framesToProcess <= size_t(sourceBus->mDuration) && framesToProcess <= size_t(destinationBus->mDuration);
 
     MOZ_ASSERT(isSafeToProcess);
@@ -162,14 +162,14 @@ void Reverb::process(const AudioChunk* sourceBus, AudioChunk* destinationBus, si
         return;
 
     // For now only handle mono or stereo output
-    MOZ_ASSERT(destinationBus->mChannelData.Length() <= 2);
+    MOZ_ASSERT(destinationBus->ChannelCount() <= 2);
 
     float* destinationChannelL = static_cast<float*>(const_cast<void*>(destinationBus->mChannelData[0]));
     const float* sourceBusL = static_cast<const float*>(sourceBus->mChannelData[0]);
 
     // Handle input -> output matrixing...
-    size_t numInputChannels = sourceBus->mChannelData.Length();
-    size_t numOutputChannels = destinationBus->mChannelData.Length();
+    size_t numInputChannels = sourceBus->ChannelCount();
+    size_t numOutputChannels = destinationBus->ChannelCount();
     size_t numReverbChannels = m_convolvers.Length();
 
     if (numInputChannels == 2 && numReverbChannels == 2 && numOutputChannels == 2) {