Android SoundPool

一个学渣的自我救赎

已于 2024-05-15 08:55:31 修改

阅读量667

点赞数 15

分类专栏： Android Framework 文章标签： android

于 2024-05-06 08:49:01 首次发布

本文链接：https://blog.csdn.net/liuning1985622/article/details/138481832

版权

Android Framework 专栏收录该内容

343 篇文章 1 订阅

订阅专栏

一、SoundPool介绍

SoundPool 一般用来播放密集，急促而又短暂的音效。

SoundPool 对象可以看作是一个可以从 APK 中导入资源或者从文件系统中载入文件的样本集合。

它利用 MediaPlayer 服务为音频解码为一个原始16位 PCM流。这个特性使得应用程序可以进行流压缩，而无须忍受在播放音频时解压所带来的 CPU 负载和延时。SoundPool 使用音效池的概念来管理多个播放流，如果超过流的最大数目， SoundPool 会基于优先级自动停止先前播放的流，另外，SoundPool 还支持自行设置声音的品质、音量、播放比率等参数。

在我们车机能想到的使用场景就是播放beep音和警告音等用户提示音。

二、SoundPool相关类

JAVA类

SoundPool

SoundPool是Android平台上的一个类，用于播放短时、急促、密集的提示音。它适用于需要快速播放音频的场景，比如支付收款的到账通知等。SoundPool支持异步初始化，可以导入本地音频文件并解码。但需要注意的是，如果资源未加载完全时播放，则可能会出现不完整的情况。SoundPool由java类和C++类。

JAVA类

SoundPool代码位于：

frameworks/base/media/java/android/media/SoundPool.java

SoundPool的定义：

public class SoundPool extends PlayerBase {
    private final class EventHandler extends Handler {}
    public static class Builder {}
}

PlayerBase

PlayerBase是一个Android平台的抽象类，是MediaPlayer和AudioTrack等音视频播放器的基类，提供了部分公共的属性和方法。PlayerBase中封装了播放器的状态以及状态转换的方法，同时提供了一个Handler用于处理内部的消息，保证了内部交互的同步性和可靠性。另外，PlayerBase中也提供了一些常用的变量。子类可以继承并实现PlayerBase的抽象方法，然后使用该类提供的状态的变化来实现相应的逻辑。

PlayerBase代码位于：

frameworks/base/media/java/android/media/PlayerBase.java

PlayerBase的定义：

public abstract class PlayerBase {}

AudioAttributes

AudioAttributes用于描述音频属性，如下为几种常见属性设置：

setAllowedCapturePolicy：设置允许捕获音频的策略

public static final int ALLOW_CAPTURE_BY_ALL = 1：Indicates that the audio may be captured by any app.
public static final int ALLOW_CAPTURE_BY_SYSTEM = 2：Indicates that the audio may only be captured by system apps.
public static final int ALLOW_CAPTURE_BY_NONE = 3：Indicates that the audio is not to be recorded by any app, even if it is a system app.

setContentType：设置内容类型

public final static int CONTENT_TYPE_UNKNOWN = 0：Content type value to use when the content type is unknown, or other than the ones defined.
public final static int CONTENT_TYPE_SPEECH = 1：Content type value to use when the content type is speech.
public final static int CONTENT_TYPE_MUSIC = 2：Content type value to use when the content type is music.
public final static int CONTENT_TYPE_MOVIE = 3：Content type value to use when the content type is a soundtrack, typically accompanying a movie or TV program Content type value to use when the content type is a soundtrack, typically accompanying a movie or TV program.
public final static int CONTENT_TYPE_SONIFICATION = 4：Content type value to use when the content type is a sound used to accompany a user action, such as a beep or sound effect expressing a key click, or event, such as the type of a sound for a bonus being received in a game. These sounds are mostly synthesized or short Foley sounds.
public static final int CONTENT_TYPE_ULTRASOUND = 1997：Content type value to use when the content type is ultrasound.

setFlags：

public final static int FLAG_AUDIBILITY_ENFORCED = 0x1 << 0：Flag defining a behavior where the audibility of the sound will be ensured by the system. To ensure sound audibility, the system only uses built-in speakers or wired headphones and specifically excludes wireless audio devices.
public final static int FLAG_SECURE = 0x1 << 1：Flag defining a behavior where the playback of the sound is ensured without degradation only when going to a secure sink.
public final static int FLAG_SCO = 0x1 << 2：Flag to enable when the stream is associated with SCO usage. Internal use only for dealing with legacy STREAM_BLUETOOTH_SCO
public final static int FLAG_BEACON = 0x1 << 3：Flag defining a behavior where the system ensures that the playback of the sound will be compatible with its use as a broadcast for surrounding people and/or devices. Ensures audibility with no or minimal post-processing applied.
public final static int FLAG_HW_AV_SYNC = 0x1 << 4：Flag requesting the use of an output stream supporting hardware A/V synchronization.
public final static int FLAG_HW_HOTWORD = 0x1 << 5：Flag requesting capture from the source used for hardware hotword detection. To be used with capture preset MediaRecorder.AudioSource.HOTWORD or MediaRecorder.AudioSource.VOICE_RECOGNITION.
public final static int FLAG_BYPASS_INTERRUPTION_POLICY = 0x1 << 6：Flag requesting audible playback even under limited interruptions.
public final static int FLAG_BYPASS_MUTE = 0x1 << 7：Flag requesting audible playback even when the underlying stream is muted.
public final static int FLAG_LOW_LATENCY = 0x1 << 8：Flag requesting a low latency path when creating an AudioTrack. When using this flag, the sample rate must match the native sample rate of the device. Effects processing is also unavailable.
public final static int FLAG_DEEP_BUFFER = 0x1 << 9：Flag requesting a deep buffer path when creating an {@code AudioTrack}. A deep buffer path, if available, may consume less power and is suitable for media playback where latency is not a concern.
public static final int FLAG_NO_MEDIA_PROJECTION = 0x1 << 10：Flag specifying that the audio shall not be captured by third-party apps with a MediaProjection.
public static final int FLAG_MUTE_HAPTIC = 0x1 << 11：Flag indicating force muting haptic channels. public static final int FLAG_NO_SYSTEM_CAPTURE = 0x1 << 12：Flag specifying that the audio shall not be captured by any apps, not even system apps.
public static final int FLAG_CAPTURE_PRIVATE = 0x1 << 13：Flag requesting private audio capture. When set in audio attributes passed to an AudioRecord, this prevents a privileged Assistant from capturing audio while this AudioRecord is active.
public static final int FLAG_CONTENT_SPATIALIZED = 0x1 << 14：Flag indicating the audio content has been processed to provide a virtual multichannel audio experience
public static final int FLAG_NEVER_SPATIALIZE = 0x1 << 15：Flag indicating the audio content is to never be spatialized
public static final int FLAG_CALL_REDIRECTION = 0x1 << 16：Flag indicating the audio is part of a call redirection. Valid for playback and capture.

setUsage：设置使用场景：

private static final int USAGE_INVALID = -1： Invalid value, only ever used for an uninitialized usage value
public final static int USAGE_UNKNOWN = 0：Usage value to use when the usage is unknown.
public final static int USAGE_MEDIA = 1：Usage value to use when the usage is media, such as music, or movie soundtracks.
public final static int USAGE_VOICE_COMMUNICATION = 2：Usage value to use when the usage is voice communications, such as telephony or VoIP.
public final static int USAGE_VOICE_COMMUNICATION_SIGNALLING = 3：Usage value to use when the usage is in-call signalling, such as with a "busy" beep, or DTMF tones.
public final static int USAGE_ALARM = 4：Usage value to use when the usage is an alarm (e.g. wake-up alarm).
public final static int USAGE_NOTIFICATION = 5：Usage value to use when the usage is notification. See other notification usages for more specialized uses.
public final static int USAGE_NOTIFICATION_RINGTONE = 6：Usage value to use when the usage is telephony ringtone.
public final static int USAGE_NOTIFICATION_COMMUNICATION_REQUEST = 7：Usage value to use when the usage is a request to enter/end a communication, such as a VoIP communication or video-conference.
public final static int USAGE_NOTIFICATION_COMMUNICATION_INSTANT = 8：Usage value to use when the usage is notification for an "instant" communication such as a chat, or SMS.
public final static int USAGE_NOTIFICATION_COMMUNICATION_DELAYED = 9：Usage value to use when the usage is notification for a non-immediate type of communication such as e-mail.
public final static int USAGE_NOTIFICATION_EVENT = 10：Usage value to use when the usage is to attract the user's attention, such as a reminder or low battery warning.
public final static int USAGE_ASSISTANCE_ACCESSIBILITY = 11：Usage value to use when the usage is for accessibility, such as with a screen reader.
public final static int USAGE_ASSISTANCE_NAVIGATION_GUIDANCE = 12：Usage value to use when the usage is driving or navigation directions.
public final static int USAGE_ASSISTANCE_SONIFICATION = 13：Usage value to use when the usage is sonification, such as with user interface sounds.
public final static int USAGE_GAME = 14：Usage value to use when the usage is for game audio.
public final static int USAGE_VIRTUAL_SOURCE = 15：Usage value to use when feeding audio to the platform and replacing "traditional" audio source, such as audio capture devices.
public final static int USAGE_ASSISTANT = 16：Usage value to use for audio responses to user queries, audio instructions or help utterances.
public static final int USAGE_CALL_ASSISTANT = 17：Usage value to use for assistant voice interaction with remote caller on Cell and VoIP calls.

AudioAttributes代码位于：

frameworks/base/media/java/android/media/AudioAttributes.java

AudioAttributes的定义：

public final class AudioAttributes implements Parcelable {
    public static class Builder {}
}

SoundEffectsHelper

一个帮助程序类，用于管理 AudioService 使用的声音效果加载/卸载。当在 AudioService 的消息处理程序线程上调用其方法时，实际工作将卸载到专用线程。

这有助于保持 AudioService 的响应速度。

SoundEffectsHelper代码位于：

frameworks/base/service/java/com/android/server/audio/SoundEffectsHelper.java

SoundEffectsHelper的定义：

class SoundEffectsHelper {
    private static final class Resource {}
    private class SfxWorker extends Thread {}
    private class SfxHandler extends Handler {}
    private class SoundPoolLoader implements
            android.media.SoundPool.OnLoadCompleteListener {}
}

C++类

SoundPool

SoundPool代码位于：

frameworks/base/media/jni/SoundPool/SoundPool.h

frameworks/base/media/jni/SoundPool/SoundPool.cpp

SoundPool的定义：

class SoundPool {}

StreamManager

StreamManager 用于管理流（由 Java 中的 StreamID 访问）。

StreamManager代码位于：

frameworks/base/media/jni/SoundPool/StreamManager.h

frameworks/base/media/jni/SoundPool/StreamManager.cpp

StreamManager的定义：

class JavaThread {} //JavaThread is used like std::thread but for threads that may call the JVM.
class ThreadPool {} //The ThreadPool manages thread lifetimes of SoundPool worker threads.
class PerfectHash {} //A Perfect HashTable for IDs (key) to pointers (value).
class StreamMap {} //StreamMap contains the all the valid streams available to SoundPool.
class StreamManager : public StreamMap {} //StreamManager is used to manage the streams (accessed by StreamID from Java).

SoundManager

此类管理 SoundPool 的声音。

SoundManager代码位于：

frameworks/base/media/jni/SoundPool/SoundManager.h

frameworks/base/media/jni/SoundPool/SoundManager.cpp

SoundManager的定义：

class SoundPoolEvent {}
class SoundManager {}
class CallbackHandler {}

Sound

Sound是SoundPool使用的资源，由soundID引用。

Sound 代码位于：

frameworks/base/media/jni/SoundPool/Sound.h

frameworks/base/media/jni/SoundPool/Sound.cpp

Sound 的定义：

class Sound {}

Sound 方法：

int32_t getSoundID() const { return mSoundID; }
int32_t getChannelCount() const { return mChannelCount; }
uint32_t getSampleRate() const { return mSampleRate; }
audio_format_t getFormat() const { return mFormat; }
audio_channel_mask_t getChannelMask() const { return mChannelMask; }
size_t getSizeInBytes() const { return mSizeInBytes; }
sound_state getState() const { return mState; }
uint8_t* getData() const { return static_cast<uint8_t*>(mData->unsecurePointer()); }
sp<IMemory> getIMemory() const { return mData; }

SoundDecoder

SoundDecoder处理后台解码任务。

SoundDecoder代码位于：

frameworks/base/media/jni/SoundPool/SoundDecoder.h

frameworks/base/media/jni/SoundPool/SoundDecoder.cpp

SoundDecoder的定义：

class SoundDecoder {}

SoundDecoder方法：

void loadSound(int32_t soundID) NO_THREAD_SAFETY_ANALYSIS; // uses unique_lock
void quit();

Stream

soundpool Stream相关处理。

Stream代码位于：

frameworks/base/media/jni/SoundPool/Stream.h

frameworks/base/media/jni/SoundPool/Stream.cpp

Stream的定义：

class Stream : public CSubsystemObject {}

Stream 方法：

bool requestStop(int32_t streamID);
        void stop();
        void clearAudioTrack();
        void pause(int32_t streamID);
        void autoPause();
        void resume(int32_t streamID);
        void autoResume();
        void mute(bool muting);
        Stream* playPairStream(std::vector<std::any>& garbage);
        void setVolume(int32_t streamID, float leftVolume, float rightVolume);
        void setRate(int32_t streamID, float rate);
        void setPriority(int32_t streamID, int priority);
        void setLoop(int32_t streamID, int loop);
        void setPlay(int32_t streamID, const std::shared_ptr<Sound> &sound, int32_t soundID float leftVolume, float rightVolume, int32_t priority, int32_t loop, float rate);
        void setStopTimeNs(int64_t stopTimeNs); 
        int32_t getPriority();
        int32_t getPairPriority();
        int64_t getStopTimeNs();
        int32_t getStreamID();
        int32_t getSoundID();
        bool hasSound();
        Stream* getPairStream() const;
        int getCorrespondingStreamID();

三、SoundPool API

SoundPool(int maxStreams, int streamType, int srcQuality)：构造函数

int load(Context context, int resId, int priority)：加载声音资源，返回一个声音的ID，后面我们可以通过这个ID来播放指定的声音

int load(String path, int priority)：加载声音资源，返回一个声音的ID，后面我们可以通过这个ID来播放指定的声音

int load(FileDescriptor fd, long offset, long length, int priority)：加载声音资源，返回一个声音的ID，后面我们可以通过这个ID来播放指定的声音

int load(AssetFileDescriptor afd, int priority)：加载声音资源，返回一个声音的ID，后面我们可以通过这个ID来播放指定的声音

int play(int soundID, float leftVolume, float rightVolume, int priority, int loop, float rate)：播放

void setVolume(int streamID, float volume)：设置音量

void setPriority(int streamID, int priority)：设置优先级

void pause(int streamID)：暂停

void resume(int streamID)：重播

void stop(int streamID)：停止

void release() ：释放资源

参数介绍：

maxStreams 同时播放流的最大数量，当播放的流的数目大于此值，则会选择性停止优先级较低的流

streamType 流类型，比如 STREAM_MUSIC

srcQuality 采样率转换器质量,目前没有什么作用,默认填充0

context：上下文

resId：资源id

priority：没什么用的一个参数，建议设置为1，保持和未来的兼容性

path：文件路径

FileDescriptor：貌似是流吧，这个我也不知道

AssetFileDescriptor：从asset目录读取某个资源文件，用法： AssetFileDescriptor descriptor = - assetManager.openFd(“biaobiao.mp3”)；

soundID：Load()返回的声音ID号

leftVolume：左声道音量设置

rightVolume：右声道音量设置

priority：指定播放声音的优先级，数值越高，优先级越大。

loop：指定是否循环：-1表示无限循环，0表示不循环，其他值表示要重复播放的次数

rate：指定播放速率：1.0的播放率可以使声音按照其原始频率，而2.0的播放速率，可以使声音按照其原始频率的两倍播放。如果为0.5的播放率，则播放速率是原始频率的一半。播放速率的取值范围是0.5至2.0。

四、SoundPool流程分析

关注

15
点赞
踩
29

收藏

觉得还不错? 一键收藏
0
评论
Android SoundPool

SoundPool 一般用来播放密集，急促而又短暂的音效。SoundPool 对象可以看作是一个可以从 APK 中导入资源或者从文件系统中载入文件的样本集合。它利用 MediaPlayer 服务为音频解码为一个原始16位 PCM流。这个特性使得应用程序可以进行流压缩，而无须忍受在播放音频时解压所带来的 CPU 负载和延时。
复制链接

扫一扫