I have a Biquad II filter implementation in my embedded environment and I'm trying to find some equations to generate the coefficients for it.

I have already found the Audio Eq Cookbook, which simply describes how to calculate all of the coefficients for Type I.

LPF: (Taken from Audio Eq Cookbook, Type I Equations)

omega = 2*PI*frequency/sampleRate

sin   = sin(omega)
cos   = cos(omega)

alpha = sin/(2*Q)                                     (if Q is specified)
= sin*sinh[ ln(2)/2 * bandwidth * omega/sin ]   (if bandwidth is specified)

b0 =  (1 - cos)/2
b1 =   1 - cos
b2 =  (1 - cos)/2
a0 =   1 + alpha
a1 =  -2*cos
a2 =   1 - alpha


What are the equations for type II coefficients?

• Why would the coefficients depend on biquad type ? Mar 22 '19 at 16:29
• Can I use biquad I coefficients for II type as well? Mar 24 '19 at 12:25
• hay, i didn't see this until now, 2 weeks later. the coefficients for the Direct Form 1 and Direct Form 2 are the same. but these coefficients have to be massaged a bit if your implementation is a form like Lattice or Ladder or some other State Variable Forms. Apr 5 '19 at 22:01
• @robertbristow-johnson hey robert, please elaborate since I'm desperately wanting to implement a ladder filter. Can you lead me to a reading source? or even better, a math like this simple? Apr 5 '19 at 23:24
• @robertbristow-johnson lol, robert it is you! you are the author of audio eq cookbook, huge fan! such an honour! :) thank you for this lovely source, It saved me like a years work maybe. Thank you! Your page in musicdsp is down by the way, is there any other sites hosting your original writing? While I find you here, would like to ask another question. When my filter drops down below 200-300 hz, some distortion/noise begin to rumble in low end, especially with high Q. Is there a way to eliminate this? Using direct form II transposed. Apr 5 '19 at 23:38