Deprecation notice
This filter designer is no longer maintained and may be removed in the future.

For hobby and DIY embedded projects (Arduino, ESP32, RP2040, STM32 and similar), a current AI assistant does this job better today. Describe what you need in plain words: filter type (Butterworth, Bessel, Chebyshev, resonator, ...), low/high/band pass or stop, order, sample rate, corner frequencies, and the number format you want (float, or fixed point with a given bit width). Ask for the coefficients, a ready to use C/C++ function, and a check of stability and frequency response. You can then refine the result in conversation, e.g. "make it cheaper for an 8 bit AVR" or "explain the saturation handling".

For DSP building blocks in C++ (header-only, no dependencies), see my main project AbacDsp on GitHub.
character
chebyshev ripple db (<0.0)
qfactor (resonator) (>0.0)
frequency Alpha Low Alpha Up or

samplerate: Hz
Lower corner Hz Midi note (69=A440)
Upper corner Hz Midi note (69=A440)
more matched z-transform (better off, check frequency response)
prewarp (better on)
output
Signal bits (including sign): -1 bit saturation save
brutal shift, this will remove shifts up to 4 positions by changing the factor, hence reducing accuracy by maximum 4 bits but more efficient calculation
execution

//Band pass resonator order=1 alpha1=0.01 class filter { public: filter() { for(int i=0; i <= 2; i++) v[i]=0; } private: short v[3]; public: short step(short x) { v[0] = v[1]; v[1] = v[2]; long tmp = ((((x * 31928L) >> 6) //= ( 3.0449316288e-2 * x) + (v[0] * -15386L) //+( -0.9391013674*v[0]) + (v[1] * 31708L) //+( 1.9352749938*v[1]) )+8192) >> 14; // round and downshift fixed point /16384 v[2]= (short)tmp; return (short)(( (v[2] - v[0]))); // 2^ } };