1
0
Fork 0
mirror of https://github.com/Ryujinx/Ryujinx.git synced 2024-10-01 12:30:00 +02:00
Ryujinx/Ryujinx.Audio/Renderers/SoundIo/SoundIoAudioOut.cs

330 lines
10 KiB
C#
Raw Normal View History

Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
using Ryujinx.Audio.SoundIo;
using SoundIOSharp;
using System;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
using System.Collections.Generic;
namespace Ryujinx.Audio
{
/// <summary>
/// An audio renderer that uses libsoundio as the audio backend
/// </summary>
public class SoundIoAudioOut : IAalOutput
{
/// <summary>
/// The maximum amount of tracks we can issue simultaneously
/// </summary>
private const int MaximumTracks = 256;
/// <summary>
/// The volume of audio renderer
/// </summary>
private float _volume = 1.0f;
/// <summary>
/// True if the volume of audio renderer have changed
/// </summary>
private bool _volumeChanged;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// The <see cref="SoundIO"/> audio context
/// </summary>
private SoundIO _audioContext;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// The <see cref="SoundIODevice"/> audio device
/// </summary>
private SoundIODevice _audioDevice;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// An object pool containing <see cref="SoundIoAudioTrack"/> objects
/// </summary>
private SoundIoAudioTrackPool _trackPool;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// True if SoundIO is supported on the device
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// </summary>
public static bool IsSupported
{
get
{
return IsSupportedInternal();
}
}
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// Constructs a new instance of a <see cref="SoundIoAudioOut"/>
/// </summary>
public SoundIoAudioOut()
{
_audioContext = new SoundIO();
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
_audioContext.Connect();
_audioContext.FlushEvents();
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
_audioDevice = FindNonRawDefaultAudioDevice(_audioContext, true);
_trackPool = new SoundIoAudioTrackPool(_audioContext, _audioDevice, MaximumTracks);
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
public bool SupportsChannelCount(int channels)
{
return _audioDevice.SupportsChannelCount(channels);
}
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// Creates a new audio track with the specified parameters
/// </summary>
/// <param name="sampleRate">The requested sample rate</param>
/// <param name="hardwareChannels">The requested hardware channels</param>
/// <param name="virtualChannels">The requested virtual channels</param>
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <param name="callback">A <see cref="ReleaseCallback" /> that represents the delegate to invoke when a buffer has been released by the audio track</param>
/// <returns>The created track's Track ID</returns>
public int OpenHardwareTrack(int sampleRate, int hardwareChannels, int virtualChannels, ReleaseCallback callback)
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (!_trackPool.TryGet(out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
return -1;
}
// Open the output. We currently only support 16-bit signed LE
track.Open(sampleRate, hardwareChannels, virtualChannels, callback, SoundIOFormat.S16LE);
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
return track.TrackID;
}
/// <summary>
/// Stops playback and closes the track specified by <paramref name="trackId"/>
/// </summary>
/// <param name="trackId">The ID of the track to close</param>
public void CloseTrack(int trackId)
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
// Close and dispose of the track
track.Close();
// Recycle the track back into the pool
_trackPool.Put(track);
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
}
/// <summary>
/// Returns a value indicating whether the specified buffer is currently reserved by the specified track
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// </summary>
/// <param name="trackId">The track to check</param>
/// <param name="bufferTag">The buffer tag to check</param>
public bool ContainsBuffer(int trackId, long bufferTag)
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
return track.ContainsBuffer(bufferTag);
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
return false;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
/// <summary>
/// Gets a list of buffer tags the specified track is no longer reserving
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// </summary>
/// <param name="trackId">The track to retrieve buffer tags from</param>
/// <param name="maxCount">The maximum amount of buffer tags to retrieve</param>
/// <returns>Buffers released by the specified track</returns>
public long[] GetReleasedBuffers(int trackId, int maxCount)
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
List<long> bufferTags = new List<long>();
while(maxCount-- > 0 && track.ReleasedBuffers.TryDequeue(out long tag))
{
bufferTags.Add(tag);
}
return bufferTags.ToArray();
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
return new long[0];
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
/// <summary>
/// Appends an audio buffer to the specified track
/// </summary>
/// <typeparam name="T">The sample type of the buffer</typeparam>
/// <param name="trackId">The track to append the buffer to</param>
/// <param name="bufferTag">The internal tag of the buffer</param>
/// <param name="buffer">The buffer to append to the track</param>
public void AppendBuffer<T>(int trackId, long bufferTag, T[] buffer) where T : struct
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_volumeChanged)
{
track.AudioStream.SetVolume(_volume);
_volumeChanged = false;
}
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
track.AppendBuffer(bufferTag, buffer);
}
}
/// <summary>
/// Starts playback
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// </summary>
/// <param name="trackId">The ID of the track to start playback on</param>
public void Start(int trackId)
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
track.Start();
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
}
/// <summary>
/// Stops playback
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// </summary>
/// <param name="trackId">The ID of the track to stop playback on</param>
public void Stop(int trackId)
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
{
track.Stop();
}
}
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// Get playback volume
/// </summary>
public float GetVolume() => _volume;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
/// <summary>
/// Set playback volume
/// </summary>
/// <param name="volume">The volume of the playback</param>
public void SetVolume(float volume)
{
if (!_volumeChanged)
{
_volume = volume;
_volumeChanged = true;
}
}
/// <summary>
/// Gets the current playback state of the specified track
/// </summary>
/// <param name="trackId">The track to retrieve the playback state for</param>
public PlaybackState GetState(int trackId)
{
if (_trackPool.TryGet(trackId, out SoundIoAudioTrack track))
{
return track.State;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
return PlaybackState.Stopped;
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
/// <summary>
/// Releases the unmanaged resources used by the <see cref="SoundIoAudioOut" />
/// </summary>
public void Dispose()
{
_trackPool.Dispose();
_audioContext.Disconnect();
_audioContext.Dispose();
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
/// <summary>
/// Searches for a shared version of the default audio device
/// </summary>
/// <param name="audioContext">The <see cref="SoundIO"/> audio context</param>
/// <param name="fallback">Whether to fallback to the raw default audio device if a non-raw device cannot be found</param>
private static SoundIODevice FindNonRawDefaultAudioDevice(SoundIO audioContext, bool fallback = false)
{
SoundIODevice defaultAudioDevice = audioContext.GetOutputDevice(audioContext.DefaultOutputDeviceIndex);
if (!defaultAudioDevice.IsRaw)
{
return defaultAudioDevice;
}
for (int i = 0; i < audioContext.BackendCount; i++)
{
SoundIODevice audioDevice = audioContext.GetOutputDevice(i);
if (audioDevice.Id == defaultAudioDevice.Id && !audioDevice.IsRaw)
{
return audioDevice;
}
}
return fallback ? defaultAudioDevice : null;
}
/// <summary>
/// Determines if SoundIO can connect to a supported backend
/// </summary>
/// <returns></returns>
private static bool IsSupportedInternal()
{
SoundIO context = null;
SoundIODevice device = null;
SoundIOOutStream stream = null;
bool backendDisconnected = false;
try
{
context = new SoundIO();
context.OnBackendDisconnect = (i) => {
backendDisconnected = true;
};
context.Connect();
context.FlushEvents();
if (backendDisconnected)
{
return false;
}
if (context.OutputDeviceCount == 0)
{
return false;
}
device = FindNonRawDefaultAudioDevice(context);
if (device == null || backendDisconnected)
{
return false;
}
stream = device.CreateOutStream();
if (stream == null || backendDisconnected)
{
return false;
}
return true;
}
catch
{
return false;
}
finally
{
if (stream != null)
{
stream.Dispose();
}
if (context != null)
{
context.Dispose();
}
}
}
Implement libsoundio as an alternative audio backend (#406) * Audio: Implement libsoundio as an alternative audio backend libsoundio will be preferred over OpenAL if it is available on the machine. If neither are available, it will fallback to a dummy audio renderer that outputs no sound. * Audio: Fix SoundIoRingBuffer documentation * Audio: Unroll and optimize the audio write callback Copying one sample at a time is slow, this unrolls the most common audio channel layouts and manually copies the bytes between source and destination. This is over 2x faster than calling CopyBlockUnaligned every sample. * Audio: Optimize the write callback further This dramatically reduces the audio buffer copy time. When the sample size is one of handled sample sizes the buffer copy operation is almost 10x faster than CopyBlockAligned. This works by copying full samples at a time, rather than the individual bytes that make up the sample. This allows for 2x or 4x faster copy operations depending on sample size. * Audio: Fix typo in Stereo write callback * Audio: Fix Surround (5.1) audio write callback * Audio: Update Documentation * Audio: Use built-in Unsafe.SizeOf<T>() Built-in `SizeOf<T>()` is 10x faster than our `TypeSize<T>` helper. This also helps reduce code surface area. * Audio: Keep fixed buffer style consistent * Audio: Address styling nits * Audio: More style nits * Audio: Add additional documentation * Audio: Move libsoundio bindings internal As per discussion, moving the libsoundio native bindings into Ryujinx.Audio * Audio: Bump Target Framework back up to .NET Core 2.1 * Audio: Remove voice mixing optimizations. Leaves Saturation optimizations in place.
2018-11-15 03:22:50 +01:00
}
}