liwei1dao
/
deepvoice_client
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Browse Source

修复耳机麦克风启动异常

weicu
fdp 1 year ago
parent
7e53b0f88d
commit
6f56362fc3
1 changed files with 97 additions and 33 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 130
      local_plugins/azure_speech/ios/azure_speech/Sources/tools/SimpleAudioReceiver.swift

130
local_plugins/azure_speech/ios/azure_speech/Sources/tools/SimpleAudioReceiver.swift
View File

@ -195,11 +195,9 @@ public class SimpleAudioReceiver: NSObject {
/**
* 开始采集音频数据
* @param bufferHandler 音频缓冲区处理回调
* @throws 音频配置或启动错误
* @throws 音频引擎启动失败时抛出错误
*/
func startCapture(bufferHandler: @escaping (AVAudioPCMBuffer) -> Void) throws {
guard let audioEngine = audioEngine else {
throw NSError(domain: "SimpleAudioReceiver", code: -1, userInfo: [NSLocalizedDescriptionKey: "Audio engine not initialized"])
}
@ -216,59 +214,92 @@ public class SimpleAudioReceiver: NSObject {
// 重置音频引擎（在获取格式之前）
audioEngine.reset()
try setupAudioSession()
try setupAudioSession()
// 重新获取输入节点和格式（reset后需要重新获取）
let inputNode = audioEngine.inputNode
let inputFormat = inputNode.outputFormat(forBus: 0)
print("硬件输入格式: 采样率=\(inputFormat.sampleRate)Hz, 声道数=\(inputFormat.channelCount), 格式=\(inputFormat.commonFormat.rawValue)")
// 验证格式是否有效
if inputFormat.sampleRate <= 0 || inputFormat.channelCount <= 0 {
// 如果格式无效，使用默认格式
let defaultFormat = AVAudioFormat(standardFormatWithSampleRate: 48000, channels: 1)!
print("使用默认音频格式: 采样率=\(defaultFormat.sampleRate)Hz, 声道数=\(defaultFormat.channelCount)")
// 创建目标格式 - 统一使用16000Hz单声道
guard let targetFormat = AVAudioFormat(commonFormat: .pcmFormatFloat32,
sampleRate: 16000,
channels: 1,
interleaved: false) else {
throw NSError(domain: "SimpleAudioReceiver", code: -2, userInfo: [NSLocalizedDescriptionKey: "无法创建目标音频格式"])
}
print("使用目标音频格式: 采样率=\(targetFormat.sampleRate)Hz, 声道数=\(targetFormat.channelCount)")
// 检查硬件格式是否有效
let isHardwareFormatValid = inputFormat.sampleRate > 0 && inputFormat.channelCount > 0
inputNode.installTap(onBus: 0,
if isHardwareFormatValid {
print("使用硬件原生格式安装tap")
// 使用硬件原生格式安装tap，然后在回调中进行格式转换
inputNode.installTap(onBus: 0,
bufferSize: 1024,
format: defaultFormat) { (buffer, time) in
bufferHandler(buffer)
}
format: inputFormat) { (buffer, time) in
// 如果硬件格式与目标格式不匹配，进行转换
if inputFormat.sampleRate != targetFormat.sampleRate ||
inputFormat.channelCount != targetFormat.channelCount {
// 执行格式转换
if let convertedBuffer = self.convertAudioBuffer(buffer, to: targetFormat) {
bufferHandler(convertedBuffer)
} else {
print("音频格式转换失败，跳过此缓冲区")
}
} else {
bufferHandler(buffer)
}
}
} else {
// 使用硬件原生格式
inputNode.installTap(onBus: 0,
print("硬件音频格式无效，使用目标格式直接安装tap")
// 硬件格式无效时，直接使用目标格式安装tap
inputNode.installTap(onBus: 0,
bufferSize: 1024,
format: inputFormat) { (buffer, time) in
bufferHandler(buffer)
}
format: targetFormat) { (buffer, time) in
// 直接使用目标格式的缓冲区
bufferHandler(buffer)
}
}
if #available(iOS 13.0, *) {
try inputNode.setVoiceProcessingEnabled(true)
do {
try inputNode.setVoiceProcessingEnabled(true)
} catch {
print("启用语音处理失败: \(error.localizedDescription)")
}
}
audioEngine.prepare()
try audioEngine.start()
print("音频引擎启动成功")
}
/**
* 安全地移除音频tap
* 避免在移除不存在的tap时出错
* 避免在移除tap时出现崩溃
*/
/**
* 安全地移除音频tap
* 避免在移除tap时出现崩溃
*/
private func removeTapSafely() {
guard let audioEngine = audioEngine else { return }
let inputNode = audioEngine.inputNode
// 尝试移除tap，如果没有tap也不会出错
do {
inputNode.removeTap(onBus: 0)
print("成功移除音频tap")
} catch {
// 如果没有tap可移除，这是正常的
print("移除tap时出现错误（可能没有tap存在）: \(error)")
// 检查是否有tap需要移除
if inputNode.numberOfInputs > 0 {
do {
inputNode.removeTap(onBus: 0)
print("成功移除音频tap")
} catch {
print("移除音频tap时出错: \(error.localizedDescription)")
}
}
}
private func runMicrophoneCapture() {
@ -429,6 +460,40 @@ private func audioBufferToData(_ buffer: AVAudioPCMBuffer) -> Data {
}
}
/**
* 转换音频缓冲区格式
* @param buffer 原始音频缓冲区
* @param targetFormat 目标音频格式
* @return 转换后的音频缓冲区，转换失败时返回nil
*/
private func convertAudioBuffer(_ buffer: AVAudioPCMBuffer, to targetFormat: AVAudioFormat) -> AVAudioPCMBuffer? {
guard let converter = AVAudioConverter(from: buffer.format, to: targetFormat) else {
print("无法创建音频转换器: 从\(buffer.format.sampleRate)Hz到\(targetFormat.sampleRate)Hz")
return nil
}
// 计算转换后的帧数
let capacity = AVAudioFrameCount(Double(buffer.frameLength) * targetFormat.sampleRate / buffer.format.sampleRate)
guard let convertedBuffer = AVAudioPCMBuffer(pcmFormat: targetFormat, frameCapacity: capacity) else {
print("无法创建转换后的音频缓冲区")
return nil
}
var error: NSError?
let status = converter.convert(to: convertedBuffer, error: &error) { _, outStatus in
outStatus.pointee = .haveData
return buffer
}
if status == .error {
print("音频转换失败: \(error?.localizedDescription ?? "未知错误")")
return nil
}
return convertedBuffer
}
/**
* 配置音频会话
* 设置录音和播放所需的音频会话参数
@ -438,10 +503,10 @@ private func setupAudioSession() throws {
try audioSession.setCategory(.playAndRecord, mode: .default, options: [.defaultToSpeaker, .allowBluetooth])
try audioSession.setActive(true)
// 设置首选的音频参数
try audioSession.setPreferredSampleRate(48000)
// 设置输入声道数
try audioSession.setPreferredInputNumberOfChannels(1)
// 设置首选的音频参数 - 修改为16000Hz以匹配Azure Speech要求
try audioSession.setPreferredSampleRate(16000)
// 设置输入声道数
try audioSession.setPreferredInputNumberOfChannels(1)
print("音频会话配置成功: 采样率=\(audioSession.sampleRate)Hz, 输入声道=\(audioSession.inputNumberOfChannels)")
}
@ -574,4 +639,3 @@ private class LinkedBlockingQueue<T> {
}
}
}

Write Preview
Loading…
Cancel
Save