关于「次级通路时延:相位误差 2πfτ 与 10 dB 降噪的带宽上限」,《耳内降噪、堵耳效应与听力努力》给出的定量结论是什么?

结论:第二章推导时延 τ 下的降噪量 −20log10|1−e^{−jφ}|,10 dB 降噪要求 fτ 不超过 0.05054。

数据口径:书中的耳道尺寸、噪声谱、环路增益与认知参数均为示意;全部计算由书中程序生成;本书不是医疗建议,不宣称任何产品或聆听方式对听力、耳鸣或健康有任何益处。 第一章在合成噪声上运行最小均方与滤波 x 最小均方算法,说明教科书步长界 2/(L·P) 对高斯输入偏松:16 抽头时精确界是 0.1111,而次级通路模型的相位误差越过 90 度时算法发散。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.05054, 10, 20

phase allowed6 dB 29.03 deg, 10 dB 18.19 deg, 20 dB 5.73 deg
f*tau at 10 dB0.05054; at 0 dB (phi = 60 deg) = 0.16667
tau (us) f(20 dB) f(10 dB) f(6 dB) f(0 dB) Hz
53184 10108 16125 33333
101592 5054 8063 16667
25637 2022 3225 6667
50318 1011 1613 3333
100159 505 806 1667
20080 253 403 833
50032 101 161 333
tau50 us, reduction by frequency (negative = boost):
100 Hz30.06 dB
250 Hz22.10 dB
500 Hz16.09 dB
1000 Hz10.09 dB
2000 Hz4.18 dB
3333 Hz0.00 dB
5000 Hz-3.01 dB
10000 Hz-6.02 dB

核心公式

\[R(f)=-20\log_{10}|1-e^{-j\varphi}|=-20\log_{10}\bigl(2\sin(\pi f\tau)\bigr).\]
查看完整验算程序
# Chapter 2. p21: How much bandwidth a secondary-path delay leaves.
# Model: the anti-noise is a perfect inverted copy of an unpredictable
# broadband disturbance but arrives tau seconds late. At frequency f the
# residual is 1 - exp(-j phi) with phi = 2 pi f tau, so the reduction is
#   R(f) = -20 log10 |1 - exp(-j phi)| = -20 log10 (2 sin(phi / 2)).
# Solving R = A dB gives phi_A = 2 asin(10^(-A/20) / 2) and the upper
# band edge f_A = phi_A / (2 pi tau). Exact; the delays are examples.
import math


def phi_for(a_db):
    return 2 * math.asin(10 ** (-a_db / 20) / 2)


def f_edge(a_db, tau):
    return phi_for(a_db) / (2 * math.pi * tau)


if __name__ == "__main__":
    print("phase allowed: 6 dB {:.2f} deg, 10 dB {:.2f} deg,"
          " 20 dB {:.2f} deg".format(*(math.degrees(phi_for(a))
                                       for a in (6, 10, 20))))
    print("f*tau at 10 dB = {:.5f}; at 0 dB (phi = 60 deg) = {:.5f}".format(
        phi_for(10) / (2 * math.pi), 1 / 6))
    print("   tau (us)   f(20 dB)   f(10 dB)    f(6 dB)  f(0 dB)  Hz")
    for us in (5, 10, 25, 50, 100, 200, 500):
        tau = us * 1e-6
        print("{:>10}  {:>9.0f}  {:>9.0f}  {:>9.0f}  {:>7.0f}".format(
            us, f_edge(20, tau), f_edge(10, tau), f_edge(6, tau),
            1 / (6 * tau)))
    tau = 50e-6
    print("tau = 50 us, reduction by frequency (negative = boost):")
    for f in (100, 250, 500, 1000, 2000, 3333, 5000, 10000):
        phi = 2 * math.pi * f * tau
        r = -20 * math.log10(2 * math.sin(phi / 2))
        print("   {:>6} Hz  {:>7.2f} dB".format(f, r))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《耳内主动降噪最小均方自适应滤波、堵耳效应低频骨导增益与信噪比认知恢复》第 2 章「次级通路时延:相位误差 2πfτ 与 10 dB 降噪的带宽上限」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-164 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026164,
  author    = {Fan, Yafu},
  title     = {LMS Adaptive Filtering in In-Ear Active Noise Cancellation, Occlusion-Effect Bone-Conduction Gain and SNR-Driven Cognitive Recovery},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论