Audiospan documentation
This commit is contained in:
parent
93a363fef6
commit
d737a7a3ae
1 changed files with 197 additions and 3 deletions
|
|
@ -34,6 +34,58 @@
|
||||||
|
|
||||||
namespace sfz
|
namespace sfz
|
||||||
{
|
{
|
||||||
|
/**
|
||||||
|
* @brief Extension of the concept of spans to multiple channels.
|
||||||
|
*
|
||||||
|
* A span (and by extension an audiospan) is at its core a structure
|
||||||
|
* containing a pointer and a size to a buffer that is owned by another
|
||||||
|
* object. A span is thus a view into a buffer that is cheap to copy and
|
||||||
|
* pass around, and safe as long as the underlying buffer is allocated.
|
||||||
|
* The goal of the class is to reduce interfaces and usage annoyance for
|
||||||
|
* codebases. Obviously, this requires that most functions use AudioSpans.
|
||||||
|
* It also protects against overreading a buffer. Users can still indicate
|
||||||
|
* that they
|
||||||
|
* @code{.cpp}
|
||||||
|
* constexpr int bufferSize { 1024 };
|
||||||
|
* void gain(AudioSpan<float, 2> arrayView, float gain)
|
||||||
|
* {
|
||||||
|
* for (auto& f: arrayView)
|
||||||
|
* f *= gain;
|
||||||
|
* }
|
||||||
|
*
|
||||||
|
* int main(char argc, char** argv)
|
||||||
|
* {
|
||||||
|
* float leftChannel [bufferSize];
|
||||||
|
* float rightChannel [bufferSize];
|
||||||
|
*
|
||||||
|
* for (int i = 0; i < bufferSize; ++i)
|
||||||
|
* {
|
||||||
|
* leftChannel[i] = 1.0f;
|
||||||
|
* rightChannel[i] = 1.0f;
|
||||||
|
* }
|
||||||
|
*
|
||||||
|
* // Type is inferred
|
||||||
|
* AudioSpan explicitView { { leftChannel, rightChannel }, bufferSize };
|
||||||
|
* // Size will be taken as the minimum size of all the spans given
|
||||||
|
* // for all types that can be automatically cast to absl::Span or std::span
|
||||||
|
* AudioSpan explicitView2 { { leftChannel, rightChannel } };
|
||||||
|
*
|
||||||
|
* gain(explicitView, 0.5f); // the array elements are now equal to 0.5f
|
||||||
|
* gain(explicitView2, 0.5f); // the array elements are now equal to 0.25f
|
||||||
|
*
|
||||||
|
* // You can also build spans implicitely
|
||||||
|
* gain({ { leftChannel, rightChannel }, bufferSize }, 0.5f); // elements equal to 0.125f
|
||||||
|
* }
|
||||||
|
* @endcode
|
||||||
|
* You can build AudioSpans from AudioBuffers directly, the AudioBufferT.cpp file in the test
|
||||||
|
* folder show some example.
|
||||||
|
* As with many things templated in C++ the `Type` can be`const` or `volatile`, and `const float`
|
||||||
|
* is not the same type as `float`. You thus cannot build an `AudioSpan<float>` from
|
||||||
|
* a `const float *` buffer for example.
|
||||||
|
*
|
||||||
|
* @tparam Type the underlying buffer type
|
||||||
|
* @tparam MaxChannels the maximum number of channels. Defaults to sfz::config::numChannels
|
||||||
|
*/
|
||||||
template <class Type, unsigned int MaxChannels = sfz::config::numChannels>
|
template <class Type, unsigned int MaxChannels = sfz::config::numChannels>
|
||||||
class AudioSpan {
|
class AudioSpan {
|
||||||
public:
|
public:
|
||||||
|
|
@ -42,8 +94,16 @@ public:
|
||||||
{
|
{
|
||||||
}
|
}
|
||||||
|
|
||||||
AudioSpan(const std::array<Type*, MaxChannels>& spans, int numChannels, size_type offset, size_type size)
|
/**
|
||||||
: numFrames(size)
|
* @brief Construct a new Audio Span object
|
||||||
|
*
|
||||||
|
* @param spans an array of MaxChannels pointers to buffers.
|
||||||
|
* @param numChannels the number of spans to take in from the array
|
||||||
|
* @param offset starting offset for the AudioSpan
|
||||||
|
* @param numFrames size of the AudioSpan
|
||||||
|
*/
|
||||||
|
AudioSpan(const std::array<Type*, MaxChannels>& spans, int numChannels, size_type offset, size_type numFrames)
|
||||||
|
: numFrames(numFrames)
|
||||||
, numChannels(numChannels)
|
, numChannels(numChannels)
|
||||||
{
|
{
|
||||||
ASSERT(static_cast<unsigned int>(numChannels) <= MaxChannels);
|
ASSERT(static_cast<unsigned int>(numChannels) <= MaxChannels);
|
||||||
|
|
@ -51,6 +111,12 @@ public:
|
||||||
this->spans[i] = spans[i] + offset;
|
this->spans[i] = spans[i] + offset;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Construct a new Audio Span object from initializer lists
|
||||||
|
*
|
||||||
|
* @param spans the list of span
|
||||||
|
* @param numFrames the size of the audio span
|
||||||
|
*/
|
||||||
AudioSpan(std::initializer_list<Type*> spans, size_type numFrames)
|
AudioSpan(std::initializer_list<Type*> spans, size_type numFrames)
|
||||||
: numFrames(numFrames)
|
: numFrames(numFrames)
|
||||||
, numChannels(spans.size())
|
, numChannels(spans.size())
|
||||||
|
|
@ -65,6 +131,15 @@ public:
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Construct a new Audio Span object from a list of absl::Span<Type>
|
||||||
|
*
|
||||||
|
* This constructor can be implicitely called for any source that can be cast transparently to
|
||||||
|
* an absl::Span<Type>. The size of the AudioSpan is inferred from the size of the smallest
|
||||||
|
* absl::Span<Type>.
|
||||||
|
*
|
||||||
|
* @param spans a list of objects compatible with absl::Span<Type>
|
||||||
|
*/
|
||||||
AudioSpan(std::initializer_list<absl::Span<Type>> spans)
|
AudioSpan(std::initializer_list<absl::Span<Type>> spans)
|
||||||
: numChannels(spans.size())
|
: numChannels(spans.size())
|
||||||
{
|
{
|
||||||
|
|
@ -78,6 +153,17 @@ public:
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Construct a new Audio Span object from an AudioBuffer with a const Type.
|
||||||
|
*
|
||||||
|
* This constructor can be implicitely called for any source that can be cast transparently to
|
||||||
|
* an AudioBuffer<const Type>.
|
||||||
|
*
|
||||||
|
* @tparam U the underlying type compatible with the template Type of the AudioSpan
|
||||||
|
* @tparam N the number of channels in the AudioBuffer
|
||||||
|
* @tparam Alignment the alignment block size for the platform
|
||||||
|
* @param audioBuffer the source AudioBuffer.
|
||||||
|
*/
|
||||||
template <class U, unsigned int N, unsigned int Alignment, typename = std::enable_if<N <= MaxChannels>, typename = std::enable_if_t<std::is_const<U>::value, int>>
|
template <class U, unsigned int N, unsigned int Alignment, typename = std::enable_if<N <= MaxChannels>, typename = std::enable_if_t<std::is_const<U>::value, int>>
|
||||||
AudioSpan(AudioBuffer<U, N, Alignment>& audioBuffer)
|
AudioSpan(AudioBuffer<U, N, Alignment>& audioBuffer)
|
||||||
: numFrames(audioBuffer.getNumFrames())
|
: numFrames(audioBuffer.getNumFrames())
|
||||||
|
|
@ -87,6 +173,18 @@ public:
|
||||||
this->spans[i] = audioBuffer.channelReader(i);
|
this->spans[i] = audioBuffer.channelReader(i);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Construct a new Audio Span object from an AudioBuffer with a non-const Type.
|
||||||
|
*
|
||||||
|
* This constructor can be implicitely called for any source that can be cast transparently to
|
||||||
|
* an AudioBuffer<Type>.
|
||||||
|
*
|
||||||
|
* @tparam U the underlying type compatible with the template Type of the AudioSpan
|
||||||
|
* @tparam N the number of channels in the AudioBuffer
|
||||||
|
* @tparam Alignment the alignment block size for the platform
|
||||||
|
* @param audioBuffer the source AudioBuffer.
|
||||||
|
*/
|
||||||
template <class U, unsigned int N, unsigned int Alignment, typename = std::enable_if<N <= MaxChannels>>
|
template <class U, unsigned int N, unsigned int Alignment, typename = std::enable_if<N <= MaxChannels>>
|
||||||
AudioSpan(AudioBuffer<U, N, Alignment>& audioBuffer)
|
AudioSpan(AudioBuffer<U, N, Alignment>& audioBuffer)
|
||||||
: numFrames(audioBuffer.getNumFrames())
|
: numFrames(audioBuffer.getNumFrames())
|
||||||
|
|
@ -96,6 +194,12 @@ public:
|
||||||
this->spans[i] = audioBuffer.channelWriter(i);
|
this->spans[i] = audioBuffer.channelWriter(i);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief AudioSpan copy constructor
|
||||||
|
*
|
||||||
|
* @param other the other AudioSpan
|
||||||
|
*/
|
||||||
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
||||||
AudioSpan(const AudioSpan<U, N>& other)
|
AudioSpan(const AudioSpan<U, N>& other)
|
||||||
: numFrames(other.getNumFrames())
|
: numFrames(other.getNumFrames())
|
||||||
|
|
@ -106,6 +210,12 @@ public:
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get a raw pointer to a specific channel from the AudioSpan
|
||||||
|
*
|
||||||
|
* @param channelIndex the channel
|
||||||
|
* @return Type* the raw pointer to the channel
|
||||||
|
*/
|
||||||
Type* getChannel(int channelIndex)
|
Type* getChannel(int channelIndex)
|
||||||
{
|
{
|
||||||
ASSERT(channelIndex < numChannels);
|
ASSERT(channelIndex < numChannels);
|
||||||
|
|
@ -115,6 +225,12 @@ public:
|
||||||
return {};
|
return {};
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get a Span<Type> corresponding to a specific channel
|
||||||
|
*
|
||||||
|
* @param channelIndex the channel
|
||||||
|
* @return absl::Span<Type>
|
||||||
|
*/
|
||||||
absl::Span<Type> getSpan(int channelIndex)
|
absl::Span<Type> getSpan(int channelIndex)
|
||||||
{
|
{
|
||||||
ASSERT(channelIndex < numChannels);
|
ASSERT(channelIndex < numChannels);
|
||||||
|
|
@ -124,6 +240,12 @@ public:
|
||||||
return {};
|
return {};
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get a Span<const Type> corresponding to a specific channel
|
||||||
|
*
|
||||||
|
* @param channelIndex the channel
|
||||||
|
* @return absl::Span<const Type>
|
||||||
|
*/
|
||||||
absl::Span<const Type> getConstSpan(int channelIndex)
|
absl::Span<const Type> getConstSpan(int channelIndex)
|
||||||
{
|
{
|
||||||
ASSERT(channelIndex < numChannels);
|
ASSERT(channelIndex < numChannels);
|
||||||
|
|
@ -133,6 +255,11 @@ public:
|
||||||
return {};
|
return {};
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get the mean of the squared values of the AudioSpan elements on all channels.
|
||||||
|
*
|
||||||
|
* @return Type
|
||||||
|
*/
|
||||||
Type meanSquared() noexcept
|
Type meanSquared() noexcept
|
||||||
{
|
{
|
||||||
if (numChannels == 0)
|
if (numChannels == 0)
|
||||||
|
|
@ -143,27 +270,52 @@ public:
|
||||||
return result / numChannels;
|
return result / numChannels;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Fills all the elements of the AudioSpan with the same value
|
||||||
|
*
|
||||||
|
* @param value the filling value
|
||||||
|
*/
|
||||||
void fill(Type value) noexcept
|
void fill(Type value) noexcept
|
||||||
{
|
{
|
||||||
|
static_assert(!std::is_const_v<Type>, "Can't allow mutating operations on const AudioSpans");
|
||||||
for (int i = 0; i < numChannels; ++i)
|
for (int i = 0; i < numChannels; ++i)
|
||||||
sfz::fill<Type>(getSpan(i), value);
|
sfz::fill<Type>(getSpan(i), value);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Apply a gain span elementwise to all channels in the AudioSpan.
|
||||||
|
*
|
||||||
|
* @param gain the gain to apply
|
||||||
|
*/
|
||||||
void applyGain(absl::Span<const Type> gain) noexcept
|
void applyGain(absl::Span<const Type> gain) noexcept
|
||||||
{
|
{
|
||||||
|
static_assert(!std::is_const_v<Type>, "Can't allow mutating operations on const AudioSpans");
|
||||||
for (int i = 0; i < numChannels; ++i)
|
for (int i = 0; i < numChannels; ++i)
|
||||||
sfz::applyGain<Type>(gain, getSpan(i));
|
sfz::applyGain<Type>(gain, getSpan(i));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Apply a gain to all channels in the AudioSpan.
|
||||||
|
*
|
||||||
|
* @param gain the gain to apply
|
||||||
|
*/
|
||||||
void applyGain(Type gain) noexcept
|
void applyGain(Type gain) noexcept
|
||||||
{
|
{
|
||||||
|
static_assert(!std::is_const_v<Type>, "Can't allow mutating operations on const AudioSpans");
|
||||||
for (int i = 0; i < numChannels; ++i)
|
for (int i = 0; i < numChannels; ++i)
|
||||||
sfz::applyGain<Type>(gain, getSpan(i));
|
sfz::applyGain<Type>(gain, getSpan(i));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Add another AudioSpan with a compatible number of channels to the current
|
||||||
|
* AudioSpan.
|
||||||
|
*
|
||||||
|
* @param other the other AudioSpan
|
||||||
|
*/
|
||||||
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
||||||
void add(AudioSpan<U, N>& other)
|
void add(AudioSpan<U, N>& other)
|
||||||
{
|
{
|
||||||
|
static_assert(!std::is_const_v<Type>, "Can't allow mutating operations on const AudioSpans");
|
||||||
ASSERT(other.getNumChannels() == numChannels);
|
ASSERT(other.getNumChannels() == numChannels);
|
||||||
if (other.getNumChannels() == numChannels) {
|
if (other.getNumChannels() == numChannels) {
|
||||||
for (int i = 0; i < numChannels; ++i)
|
for (int i = 0; i < numChannels; ++i)
|
||||||
|
|
@ -171,9 +323,16 @@ public:
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Copy the elements of another AudioSpan with a compatible number of channels
|
||||||
|
* to the current AudioSpan.
|
||||||
|
*
|
||||||
|
* @param other the other AudioSpan
|
||||||
|
*/
|
||||||
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
template <class U, unsigned int N, typename = std::enable_if<N <= MaxChannels>>
|
||||||
void copy(AudioSpan<U, N>& other)
|
void copy(AudioSpan<U, N>& other)
|
||||||
{
|
{
|
||||||
|
static_assert(!std::is_const_v<Type>, "Can't allow mutating operations on const AudioSpans");
|
||||||
ASSERT(other.getNumChannels() == numChannels);
|
ASSERT(other.getNumChannels() == numChannels);
|
||||||
if (other.getNumChannels() == numChannels) {
|
if (other.getNumChannels() == numChannels) {
|
||||||
for (int i = 0; i < numChannels; ++i)
|
for (int i = 0; i < numChannels; ++i)
|
||||||
|
|
@ -181,34 +340,69 @@ public:
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get the size of this AudioSpan.
|
||||||
|
*
|
||||||
|
* @returns size_type the number of frames in the AudioSpan
|
||||||
|
*/
|
||||||
size_type getNumFrames()
|
size_type getNumFrames()
|
||||||
{
|
{
|
||||||
return numFrames;
|
return numFrames;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Get the number of channels of this AudioSpan.
|
||||||
|
*
|
||||||
|
* @returns size_type the number of channels in the AudioSpan
|
||||||
|
*/
|
||||||
int getNumChannels()
|
int getNumChannels()
|
||||||
{
|
{
|
||||||
return numChannels;
|
return numChannels;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Creates a new AudioSpan but with only the `length` first elements of each channel.
|
||||||
|
*
|
||||||
|
* @param length the number of elements to take on each channel
|
||||||
|
*/
|
||||||
AudioSpan<Type> first(size_type length)
|
AudioSpan<Type> first(size_type length)
|
||||||
{
|
{
|
||||||
ASSERT(length <= numFrames);
|
ASSERT(length <= numFrames);
|
||||||
return { spans, numChannels, 0, length };
|
return { spans, numChannels, 0, length };
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Creates a new AudioSpan but with only the `length` last elements of each channel.
|
||||||
|
*
|
||||||
|
* @param length the number of elements to take on each channel
|
||||||
|
*/
|
||||||
AudioSpan<Type> last(size_type length)
|
AudioSpan<Type> last(size_type length)
|
||||||
{
|
{
|
||||||
ASSERT(length <= numFrames);
|
ASSERT(length <= numFrames);
|
||||||
return { spans, numChannels, numFrames - length, length };
|
return { spans, numChannels, numFrames - length, length };
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Creates a new AudioSpan starting at an offset `offset` on each channel and
|
||||||
|
* taking `length` elements. The new AudioSpan will have `length` elements. This basically
|
||||||
|
* removes the first `offset` elements and the last `numFrames - length - offset` elements
|
||||||
|
* from the AudioSpan.
|
||||||
|
*
|
||||||
|
* @param length the number of elements to take on each channel
|
||||||
|
*/
|
||||||
AudioSpan<Type> subspan(size_type offset, size_type length)
|
AudioSpan<Type> subspan(size_type offset, size_type length)
|
||||||
{
|
{
|
||||||
ASSERT(length + offset <= numFrames);
|
ASSERT(length + offset <= numFrames);
|
||||||
return { spans, numChannels, offset, length };
|
return { spans, numChannels, offset, length };
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* @brief Creates a new AudioSpan starting at an offset `offset` on each channel and
|
||||||
|
* taking all the remaining elements. This function basically removes the first `offset`
|
||||||
|
* elements from the AudioSpan. The new Audiospan will have size `numFrames - offset`.
|
||||||
|
*
|
||||||
|
* @param length the number of elements to take on each channel
|
||||||
|
*/
|
||||||
AudioSpan<Type> subspan(size_type offset)
|
AudioSpan<Type> subspan(size_type offset)
|
||||||
{
|
{
|
||||||
ASSERT(offset <= numFrames);
|
ASSERT(offset <= numFrames);
|
||||||
|
|
|
||||||
Loading…
Add table
Reference in a new issue