Hmmm. I made a moog-style filter last year, but it's an external of course.
Totally ripped from source code I found elsewhere, but it may help if you look
at the algorithm for the filter's operation.
Note that this is an IIR not an FIR filter, so it can burst and screw up your
DSP chain (but only if you feed it bad values). Everything is signal-based -
i.e. cutoff frequency and resonance.
Cheers,
Ed
while (n--) {
i1=(*in++);
fc1 = (*fc++);
/* This failsafe line stops the filter bursting
* ...but it is expensive! */
// if(x-safety) {
// fc1 = fc1 = 1 ? fc1 = 0 ? fc1 : 0 : 1;
// }
res1 = (*res++);
q = 1.0f - fc1;
p = fc1 + 0.8f * fc1 * q;
fcoeff = p + p - 1.0f;
q = res1 * (1.0f + 0.5f * q * (1.0f - q + 5.6f * q * q));
i1 -= q * xb4; //feedback
t1 = xb1;
xb1 = (i1 + xb0) * p - xb1 * fcoeff;
t2 = xb2;
xb2 = (xb1 + t1) * p - xb2 * fcoeff;
t1 = xb3;
xb3 = (xb2 + t2) * p - xb3 * fcoeff;
xb4 = (xb3 + t1) * p - xb4 * fcoeff;
xb4 = saturate(xb4);
xb4 = xb4 - xb4 * xb4 * xb4 * 0.01f;
xb0 = i1;
*out++ = xb4*1.414; // lowpass mode
// *out++ = i1 - x-b4; // highpass mode
// Lowpass output: xb4
// Highpass output: in - xb4;
// Bandpass output: 3.0f * (b3 - xb4);
}
Ninja Jamm - a revolutionary new music remix app from Ninja Tune and Seeper,
for iPhone and iPad
http://www.ninjajamm.com/
Gemnotes-0.2: Live music notation for Pure Data, now with dynamics!
http://sharktracks.co.uk/
On Tuesday, 15 October 2013, 13:16, Billy Stiltner billy.stilt...@gmail.com
wrote:
martin, I have not tried the moog but did use one of your abstractions using
the fexpr~ as a starting point to build the original dsp cookbook
filters(frequency thats where its happenin man) , I had been wanting to hear
them in realtime since 1998 or so. I'm not sure I have them in there correct as
they are over resonant but make a nice oscillator for drums if properly
limited.
On Fri, Oct 11, 2013 at 5:33 AM, martin brinkmann m...@martin-brinkmann.de
wrote:
On 10/11/2013 08:32 AM, Dan Wilcox wrote:
[vcf~] sounds close, but of course, it's a band pass so it's not a real
replacement.
the undocumented 2nd output of vcf~ is a lowpass. and there are a few
methods to build the basic (cookbook) filters with pd-vanilla-objects:
you could use fexpr~ to make your own signal-rate biquad. it works very
well, but needs a lot more cpu than necessary. or you could use
cpole~/czero~. like in the beequad-abstractions. i have used the same
method in all my filters, with signal-rate. (in my instruments
collection on my homepage). of course this is still not exactly
moog~. it should be possible though to make a moog-abstraction
in the same way, but i have not tried (yet)...
maybe anyone else has already?
bis denn!
martin
___
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management -
http://lists.puredata.info/listinfo/pd-list
___
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management -
http://lists.puredata.info/listinfo/pd-list
/*
* moog vcf, 4-pole lowpass resonant filter
*
* (c) Edward Kelly 2012
* BSD License
*/
#include m_pd.h
#include math.h
#define _limit 0.95
static t_class *mvcf_tilde_class;
typedef struct _mvcf_tilde {
t_object x_obj;
t_float b0, b1, b2, b3, b4; //filter buffers to keep (beware denormals!)
t_float token, debug, safety;
t_outlet *lp;
} t_mvcf_tilde;
/* We could have a mode where the fc and res are only registered at the start of the block (quick) or are registered in signal mode (slow) - i.e. a flag */
static inline float saturate( float input ) { //clamp without branching
float x1 = fabsf( input + _limit );
float x2 = fabsf( input - _limit );
return 0.5 * (x1 - x2);
}
t_int *mvcf_tilde_perform(t_int *w) {
t_mvcf_tilde *x = (t_mvcf_tilde *)(w[1]);
t_sample *in = (t_sample *)(w[2]);
t_sample *fc = (t_sample *)(w[3]);
t_sample *res = (t_sample *)(w[4]);
t_sample *out = (t_sample *)(w[5]);
int n = (int)(w[6]);
float t1 = 0;
float t2 = 0;
float xb0 = x-b0;
float xb1 = x-b1;
float xb2 = x-b2;
float xb3 = x-b3;
float xb4 = x-b4;
float i1 = 0;
float fc1 = 0;
float res1 = 0;
float q = 0;
float p = 0;
float fcoeff = 0;
// while (n-=4) {
while (n--) {
i1=(*in++);
fc1 = (*fc++);
/* This failsafe line stops the filter bursting
* ...but it is expensive! */
//if(x-safety) {
// fc1 = fc1 = 1 ? fc1 = 0 ? fc1 : 0 : 1;
//}
res1 = (*res++);
q = 1.0f - fc1;
p = fc1 + 0.8f * fc1 * q;
fcoeff = p + p - 1.0f;
q = res1 * (1.0f + 0.5f * q * (1.0f - q + 5.6f * q * q));
i1 -= q * xb4; //feedback
t1 = xb1;
xb1 = (i1 + xb0) * p - xb1 * fcoeff;
t2 = xb2;
xb2 = (xb1 +