如何检测wav文件中蜂鸣声的频率和持续时间？

Question

我有一些

test.wav

文件，我需要从该文件中获取声音的频率及其长度。为了测试我的解决方案，我用这些“嘟嘟声”创建了

test.wav

声音：

1500 Hz 持续 100 ms、1900 Hz 持续 300 ms、1200 Hz 持续 10 ms、1900 Hz 持续 300 ms

我希望我的解决方案能够打印“嘟嘟声”，指示频率和长度。但我的解决方案忽略了 10 毫秒的短蜂鸣声（经过一些实验，我发现我的解决方案忽略了所有短 100 毫秒的蜂鸣声（并且它还增加了一些频率）

我的无效解决方案：

import numpy as np
from scipy.io import wavfile

def extract_frequencies_and_durations(wav_path, frame_duration=0.02, freq_tolerance=5):
    rate, audio_data = wavfile.read(wav_path)
    if audio_data.ndim > 1:
        audio_data = np.mean(audio_data, axis=1)
    frame_length = int(rate * frame_duration)
    frequencies = []
    durations = []
    current_freq = None
    current_duration = 0
    for i in range(0, len(audio_data), frame_length):
        frame = audio_data[i:i + frame_length]
        fft_spectrum = np.fft.fft(frame)
        freqs = np.fft.fftfreq(len(fft_spectrum), 1 / rate)
        magnitude = np.abs(fft_spectrum)
        dominant_freq = abs(freqs[np.argmax(magnitude)])
        if current_freq is None or abs(dominant_freq - current_freq) > freq_tolerance:
            if current_freq is not None:
                frequencies.append(current_freq)
                durations.append(current_duration)
            current_freq = dominant_freq
            current_duration = frame_duration
        else:
            current_duration += frame_duration
    if current_freq is not None:
        frequencies.append(current_freq)
        durations.append(current_duration)
    
    return frequencies, durations

# output
wav_path = 'test.wav'
frequencies, durations = extract_frequencies_and_durations(wav_path)
for freq, dur in zip(frequencies, durations):
    print(f"{freq:.2f}Hz - {dur * 1000:.2f}ms")

我该如何解决此解决方案或还有其他解决方案吗？

Answer 1

帧持续时间猜测为 0.02 秒，这会影响频率分析。 最好的猜测似乎是最短突发的持续时间，它可以提供更好的频率/持续时间结果。

print("Test audio: 1517:0.05; 1921:0.3; 1245:0.01; 1977:0.3\n")
wav_path = 'test2.wav'
frame_duration = 0.01
print(f"frame_duration={frame_duration}")
frequencies, durations = extract_frequencies_and_durations(wav_path, frame_duration=frame_duration)
for freq, dur in zip(frequencies, durations):
    print(f"{freq:.2f}Hz - {dur * 1000:.2f}ms")


frame_duration = 0.02
print(f"\nframe_duration={frame_duration}")
frequencies, durations = extract_frequencies_and_durations(wav_path, frame_duration=frame_duration)
for freq, dur in zip(frequencies, durations):
    print(f"{freq:.2f}Hz - {dur * 1000:.2f}ms")

生成的音频：

ffmpeg -f lavfi -i "sine=f=1517:d=0.05[0];sine=f=1921:d=0.3[1];sine=f=1245:d=0.01[2];sine=f=1977:d=0.3[3]; [0][1][2][3]concat=n=4:v=0:a=1" test2.wav

结果：

Test audio: 1517:0.05; 1921:0.3; 1245:0.01; 1977:0.3

frame_duration=0.01
rate: 44100, frame_duration: 0.01, frame_length: 441
1500.00Hz - 50.00ms
1900.00Hz - 300.00ms
1200.00Hz - 10.00ms
2000.00Hz - 300.00ms

frame_duration=0.02
rate: 44100, frame_duration: 0.02, frame_length: 882
1500.00Hz - 40.00ms
1900.00Hz - 300.00ms
1250.00Hz - 20.00ms
2000.00Hz - 300.00ms