Python soundcard库实战：手把手教你用电脑声卡搭建简易音频分析仪（附完整代码）

Python soundcard库实战：手把手教你用电脑声卡搭建简易音频分析仪（附完整代码）

在电子工程和音频处理领域，专业音频分析设备往往价格昂贵，让许多爱好者和学生望而却步。但你可能不知道，通过Python的soundcard库，配合普通的电脑声卡，就能构建一个功能强大的音频分析工具。本文将带你从零开始，实现一个完整的音频分析系统，包括信号采集、处理和可视化全流程。

1. 环境准备与基础配置

1.1 硬件需求清单

构建这个音频分析系统，你只需要以下基础设备：

• 一台普通电脑（Windows/macOS/Linux均可）
• 3.5mm音频线（用于连接信号源）
• 可选：信号发生器（如DG1062）或音频测试设备

关键点：电脑内置声卡或外置USB声卡均可，但建议使用独立声卡以获得更好的信噪比。以下是常见声卡参数对比：

1.2 Python环境搭建

首先确保已安装Python 3.6+，然后通过pip安装必要库：

pip install soundcard numpy matplotlib scipy

注意：在Linux系统上可能需要额外安装libportaudio2等依赖库

验证安装是否成功：

import soundcard as sc print(sc.default_speaker()) print(sc.default_microphone())

1.3 声卡基础操作

soundcard库提供了简洁的API来操作声卡设备。以下是一些基本功能示例：

import soundcard as sc import numpy as np # 获取所有音频设备 mics = sc.all_microphones() speakers = sc.all_speakers() # 选择默认设备 default_mic = sc.default_microphone() default_speaker = sc.default_speaker() # 录制5秒音频 sample_rate = 48000 recording = default_mic.record(samplerate=sample_rate, numframes=5*sample_rate) # 播放录制内容 default_speaker.play(recording, samplerate=sample_rate)

2. 音频信号采集与分析

2.1 实时音频采集系统

构建一个实时音频采集系统是分析的基础。以下代码实现了带缓冲的实时采集：

import soundcard as sc import numpy as np import time def real_time_capture(duration=5, sample_rate=48000): mic = sc.default_microphone() frames = [] with mic.recorder(samplerate=sample_rate) as recorder: start_time = time.time() while time.time() - start_time < duration: # 每次采集100ms数据 data = recorder.record(numframes=int(0.1 * sample_rate)) frames.append(data) return np.concatenate(frames) # 采集5秒音频 audio_data = real_time_capture()

2.2 频谱分析技术

对采集到的音频进行频谱分析是音频分析仪的核心功能。我们使用FFT实现：

import numpy as np from scipy.fft import fft import matplotlib.pyplot as plt def analyze_spectrum(signal, sample_rate): n = len(signal) yf = fft(signal) xf = np.linspace(0, sample_rate/2, n//2) # 计算幅度谱 magnitude = 2/n * np.abs(yf[:n//2]) # 绘制频谱图 plt.figure(figsize=(10,4)) plt.plot(xf, 20*np.log10(magnitude)) # 转换为dB单位 plt.xlabel('Frequency (Hz)') plt.ylabel('Magnitude (dB)') plt.grid(True) plt.show() return xf, magnitude # 示例：分析1kHz正弦波 sample_rate = 48000 t = np.linspace(0, 1, sample_rate) test_signal = 0.5 * np.sin(2 * np.pi * 1000 * t) analyze_spectrum(test_signal, sample_rate)

2.3 自动化扫频测量

结合信号发生器，我们可以实现自动化扫频测量系统：

import soundcard as sc import numpy as np import matplotlib.pyplot as plt import time def frequency_sweep_analysis(start_freq, end_freq, steps, duration=0.5): mic = sc.default_microphone() sample_rate = 48000 frequencies = np.linspace(start_freq, end_freq, steps) magnitudes = [] with mic.recorder(samplerate=sample_rate) as recorder: for freq in frequencies: # 这里需要连接信号发生器并设置频率 # 实际应用中可以通过visa等库控制信号源 print(f"Testing frequency: {freq:.1f}Hz") # 等待信号稳定 time.sleep(0.1) # 采集数据 data = recorder.record(numframes=int(duration * sample_rate)) # 计算幅度 magnitude = np.max(data) - np.min(data) magnitudes.append(magnitude) # 绘制幅频特性曲线 plt.figure(figsize=(10,5)) plt.semilogx(frequencies, 20*np.log10(magnitudes/np.max(magnitudes))) plt.xlabel('Frequency (Hz)') plt.ylabel('Normalized Magnitude (dB)') plt.title('Frequency Response') plt.grid(True) plt.show() return frequencies, magnitudes # 示例：20Hz-20kHz扫频 freq, mag = frequency_sweep_analysis(20, 20000, 50)

3. 高级功能实现

3.1 实时频谱显示

使用Matplotlib的动画功能，可以实现实时频谱显示：

import soundcard as sc import numpy as np from scipy.fft import fft import matplotlib.pyplot as plt from matplotlib.animation import FuncAnimation def real_time_spectrum_analyzer(): mic = sc.default_microphone() sample_rate = 48000 frame_size = 4096 fig, ax = plt.subplots(figsize=(10,5)) line, = ax.plot([], []) ax.set_xlim(20, 20000) ax.set_ylim(-80, 0) ax.set_xscale('log') ax.set_xlabel('Frequency (Hz)') ax.set_ylabel('Magnitude (dB)') ax.grid(True) def update(frame): with mic.recorder(samplerate=sample_rate) as recorder: data = recorder.record(numframes=frame_size) n = len(data) yf = fft(data[:,0]) # 只分析左声道 xf = np.linspace(0, sample_rate/2, n//2) magnitude = 20*np.log10(2/n * np.abs(yf[:n//2])) line.set_data(xf[1:], magnitude[1:]) # 跳过DC分量 return line, ani = FuncAnimation(fig, update, interval=100, blit=True) plt.show() real_time_spectrum_analyzer()

3.2 多通道同步分析

专业声卡通常支持多通道输入，我们可以扩展系统支持多通道分析：

def multi_channel_analysis(channels=2): mic = sc.default_microphone() sample_rate = 48000 duration = 2 with mic.recorder(samplerate=sample_rate) as recorder: data = recorder.record(numframes=duration*sample_rate) plt.figure(figsize=(12,6)) for ch in range(min(channels, data.shape[1])): # 计算每个通道的频谱 n = len(data) yf = fft(data[:,ch]) xf = np.linspace(0, sample_rate/2, n//2) magnitude = 20*np.log10(2/n * np.abs(yf[:n//2])) plt.plot(xf, magnitude, label=f'Channel {ch+1}') plt.xscale('log') plt.xlabel('Frequency (Hz)') plt.ylabel('Magnitude (dB)') plt.legend() plt.grid(True) plt.show()

3.3 声卡性能测试

通过标准测试信号，我们可以评估声卡的关键性能指标：

def test_soundcard_performance(): mic = sc.default_microphone() sample_rate = 48000 # 测试频率响应 freqs = np.logspace(np.log10(20), np.log10(20000), 50) responses = [] with mic.recorder(samplerate=sample_rate) as recorder: for freq in freqs: # 实际应用中应连接信号发生器产生纯净测试信号 time.sleep(0.1) data = recorder.record(numframes=int(0.5*sample_rate)) response = np.max(data) - np.min(data) responses.append(response) # 测试本底噪声 noise_data = recorder.record(numframes=sample_rate) noise_level = 20*np.log10(np.std(noise_data)) print(f"Soundcard noise floor: {noise_level:.1f} dB") # 绘制频率响应 plt.figure(figsize=(12,5)) plt.semilogx(freqs, 20*np.log10(responses/np.max(responses))) plt.title('Frequency Response') plt.xlabel('Frequency (Hz)') plt.ylabel('Normalized Response (dB)') plt.grid(True) plt.show()

4. 实战应用案例

4.1 音频设备频响测试

利用这套系统，我们可以测试耳机、麦克风等音频设备的频率响应：

def test_device_response(device_output, test_frequencies): """测试音频设备的频率响应 参数： device_output: 待测设备的输出信号 test_frequencies: 要测试的频率列表 """ mic = sc.default_microphone() sample_rate = 48000 responses = [] with mic.recorder(samplerate=sample_rate) as recorder: for freq in test_frequencies: # 播放测试频率 device_output.play_test_tone(freq) time.sleep(0.5) # 采集响应 data = recorder.record(numframes=int(0.5*sample_rate)) response = np.max(data) - np.min(data) responses.append(response) # 归一化处理 responses = responses / np.max(responses) # 绘制结果 plt.figure(figsize=(10,5)) plt.semilogx(test_frequencies, 20*np.log10(responses)) plt.title('Device Frequency Response') plt.xlabel('Frequency (Hz)') plt.ylabel('Response (dB)') plt.grid(True) plt.show()

4.2 房间声学分析

通过脉冲响应测量，可以分析房间的声学特性：

def measure_room_acoustics(): mic = sc.default_microphone() speaker = sc.default_speaker() sample_rate = 48000 # 生成测试信号（对数扫频） duration = 5 t = np.linspace(0, duration, duration*sample_rate) start_freq, end_freq = 20, 20000 chirp = np.sin(2*np.pi*start_freq*(end_freq/start_freq)**(t/duration)*duration/np.log(end_freq/start_freq)) # 播放并录制 with mic.recorder(samplerate=sample_rate) as recorder, \ speaker.player(samplerate=sample_rate) as player: player.play(chirp) time.sleep(0.1) # 等待播放开始 recording = recorder.record(numframes=int((duration+1)*sample_rate)) # 计算脉冲响应（简化版） impulse_response = recording[len(chirp):len(chirp)+sample_rate//2] # 绘制结果 plt.figure(figsize=(12,5)) plt.plot(np.linspace(0, 0.5, len(impulse_response)), impulse_response) plt.title('Room Impulse Response') plt.xlabel('Time (s)') plt.ylabel('Amplitude') plt.grid(True) plt.show()

4.3 音频电路测试

对于音频电路设计者，这套系统可以用于测试滤波器、放大器等电路：

def test_audio_circuit(input_signal_generator, circuit): """测试音频电路 参数： input_signal_generator: 输入信号发生器 circuit: 待测电路 """ mic = sc.default_microphone() sample_rate = 48000 test_frequencies = np.logspace(np.log10(20), np.log10(20000), 50) input_levels = [] output_levels = [] with mic.recorder(samplerate=sample_rate) as recorder: for freq in test_frequencies: # 生成输入信号 input_signal = input_signal_generator.generate(freq) # 通过待测电路 output_signal = circuit.process(input_signal) # 播放输出信号并录制 time.sleep(0.1) recorded = recorder.record(numframes=int(0.5*sample_rate)) # 计算输入输出电平 input_level = np.max(input_signal) - np.min(input_signal) output_level = np.max(recorded) - np.min(recorded) input_levels.append(input_level) output_levels.append(output_level) # 计算增益 gain = 20*np.log10(np.array(output_levels)/np.array(input_levels)) # 绘制频率响应 plt.figure(figsize=(12,5)) plt.semilogx(test_frequencies, gain) plt.title('Circuit Frequency Response') plt.xlabel('Frequency (Hz)') plt.ylabel('Gain (dB)') plt.grid(True) plt.show()