Skip to main content

Interactive DSP Parametric Filter Sandbox

Standby
Live Audio Filter H(f)

Interactive Filter Sandbox Ready

Click any Preset below or click Pink Noise / Filter Mic to audition live.

1.0
1,000 Hz

Instant DSP Filter Presets

Click any preset card to instantly load parameters and audition live spectral filtering.

Click card to audition

Understanding Audio Biquad Filters & Frequency Response

Biquad filters are second-order Infinite Impulse Response (IIR) digital filters implemented according to Robert Bristow-Johnson's Audio EQ Cookbook ↗ and the W3C BiquadFilterNode specification ↗:

  • High-Pass Filter (HPF): Passes frequencies above cutoff while rolling off low rumblings below cutoff at -12 dB/octave. Essential for eliminating microphone handling rumble, HVAC noise, and subwoofer port over-excursion.
  • Low-Pass Filter (LPF): Passes low frequencies while attenuating high treble hiss, ultrasonic interference, and digital DAC aliasing harmonics.
  • Band-Pass Filter (BPF): Isolates a narrow musical frequency band determined by the center frequency and resonance Q factor.
  • Notch Filter: Surgically removes a pinpoint frequency spike (such as 50Hz or 60Hz AC electrical power hum) with minimal side attenuation.

The Q Factor (Quality Factor / Resonance)

In parametric filtering, Q = f₀ / Bandwidth. A low Q (0.5 to 0.707 Butterworth) creates a broad, musically smooth slope. A high Q (5.0 to 15.0) focuses filter energy into an ultra-sharp surgical cut, perfect for notching out ground loop hums or acoustic feedback without affecting nearby musical harmonics.

Real-Time Spectral Verification

The live spectrum visualizer above displays the 4096-point FFT transfer function in real time. Compare the raw signal against the filtered output using the Filter Active / Bypass toggle to evaluate phase impact and attenuation depth.

DIGITAL SIGNAL PROCESSING FAQ

Frequently Asked Questions on DSP Audio Filters

Comprehensive breakdown of filter types, Butterworth vs Chebyshev vs Bessel, FFT/DFT transforms, and DSP pros & cons.

1DSP(数字信号处理)中的滤波器是什么?
DSP 滤波器是一种数学算法,通过修改数字音频信号的频率、振幅或相位,让所需的频段通过,同时衰减不需要的成分(如噪声、蜂鸣声或混叠)。
2DSP 的主要应用领域有哪些?
广泛应用于音频均衡(EQ)、降噪处理、混响效果、回声消除、动态压缩、无线通信、语音识别以及医学成像等领域。
3DSP 中的 DFT 与 FFT 是什么?
DFT(离散傅里叶变换)是将离散时域信号转换为频域能量谱的数学算法(计算量 $O(N^2)$)。FFT(快速傅里叶变换)是将 DFT 计算加速至 $O(N \log N)$ 的高效算法,实现了实时音频分析。
4什么是切比雪夫(Chebyshev)滤波器?
切比雪夫滤波器在通带与阻带之间具有比巴特沃斯更陡峭的过渡带衰减,但代价是在通带(I 型)或阻带(II 型)内引入振幅波纹(Ripple)。
5经典的三种基本滤波器类型是什么?
1) 低通滤波器(LPF):允许低于截止频率的信号通过; 2) 高通滤波器(HPF):允许高于截止频率的信号通过; 3) 带通滤波器(BPF):仅允许特定频段通过。
6数字信号处理(DSP)有哪些局限性?
主要包括缓存计算引入的延迟、量化与舍入误差噪声、超过奈奎斯特频率 $f_s/2$ 时引发的混叠(Aliasing)风险,以及计算资源的消耗。
7DSP 中的低通滤波器如何工作?
数字低通滤波器(如 Biquad 双二阶 IIR)使低于截止频率($f_c$)的信号无损通过,并以固定的斜率(如 -12 dB/oct 或 -24 dB/oct)滚降衰减高于 $f_c$ 的信号。
8巴特沃斯与切比雪夫滤波器的区别?
巴特沃斯滤波器通带内具有最大平坦特性,完全无波纹。切比雪夫滤波器在相同阶数下滚降更陡峭,但会产生振幅波纹和非线性相位失真。
9什么是贝塞尔(Bessel)滤波器?
贝塞尔滤波器专为实现最大平坦群延迟(线性相位)而设计,能完好保留音频瞬态信号的时域波形,绝不产生多余振铃或相位畸变。
10为什么在滤波器设计中使用贝塞尔函数?
贝塞尔多项式能确保通带内所有谐波成分具有相同的传播时间延迟,从而彻底消除相位失真。