/* For information on usage and redistribution, and for a DISCLAIMER OF ALL * WARRANTIES, see the file, "LICENSE.txt," in this distribution. iem_spec2 written by Thomas Musil, Copyright (c) IEM KUG Graz Austria 2000 - 2006 */ #include "m_pd.h" #include "iemlib.h" /* -------------------------- spec2_1p1z_time~ ------------------------------ */ static t_class *spec2_1p1z_time_tilde_class; typedef struct _spec2_1p1z_time_tilde { t_object x_obj; t_float x_a0; t_float x_a1; t_float x_b1; t_float *x_begmem_forw; t_float *x_begmem_back; int x_blocksize; t_float x_msi; } t_spec2_1p1z_time_tilde; static void spec2_1p1z_time_tilde_list(t_spec2_1p1z_time_tilde *x, t_symbol *s, int argc, t_atom *argv) { if((argc >= 3) && IS_A_FLOAT(argv,0) && IS_A_FLOAT(argv,1) && IS_A_FLOAT(argv,2)) { x->x_a0 = (t_float)atom_getfloatarg(0, argc, argv); x->x_a1 = (t_float)atom_getfloatarg(1, argc, argv); x->x_b1 = (t_float)atom_getfloatarg(2, argc, argv); } } static t_int *spec2_1p1z_time_tilde_perform(t_int *w) { t_float *in = (t_float *)(w[1]); t_float *out = (t_float *)(w[2]); t_spec2_1p1z_time_tilde *x = (t_spec2_1p1z_time_tilde *)(w[3]); int i, n = (t_int)(w[4]); t_float a0 = x->x_a0; t_float a1 = x->x_a1; t_float b1 = x->x_b1; t_float *vec_forw = x->x_begmem_forw; t_float *vec_back = x->x_begmem_back; t_float f; for(i=0; i<=n; i++) { f = in[i]; out[i] = a0*f + a1*vec_forw[i] + b1*vec_back[i]; vec_forw[i] = f; vec_back[i] = out[i]; } return(w+5); } static t_int *spec2_1p1z_time_tilde_perf16(t_int *w) { t_float *in = (t_float *)(w[1]); t_float *out = (t_float *)(w[2]); t_spec2_1p1z_time_tilde *x = (t_spec2_1p1z_time_tilde *)(w[3]); int n = (t_int)(w[4]); t_float a0 = x->x_a0; t_float a1 = x->x_a1; t_float b1 = x->x_b1; t_float *vec_forw = x->x_begmem_forw; t_float *vec_back = x->x_begmem_back; t_float ff; while(n) { t_float f[16]; f[0] = in[0]; f[1] = in[1]; f[2] = in[2]; f[3] = in[3]; f[4] = in[4]; f[5] = in[5]; f[6] = in[6]; f[7] = in[7]; f[8] = in[8]; f[9] = in[9]; f[10] = in[10]; f[11] = in[11]; f[12] = in[12]; f[13] = in[13]; f[14] = in[14]; f[15] = in[15]; out[0] = a0*f[0] + a1*vec_forw[0] + b1*vec_back[0]; out[1] = a0*f[1] + a1*vec_forw[1] + b1*vec_back[1]; out[2] = a0*f[2] + a1*vec_forw[2] + b1*vec_back[2]; out[3] = a0*f[3] + a1*vec_forw[3] + b1*vec_back[3]; out[4] = a0*f[4] + a1*vec_forw[4] + b1*vec_back[4]; out[5] = a0*f[5] + a1*vec_forw[5] + b1*vec_back[5]; out[6] = a0*f[6] + a1*vec_forw[6] + b1*vec_back[6]; out[7] = a0*f[7] + a1*vec_forw[7] + b1*vec_back[7]; out[8] = a0*f[8] + a1*vec_forw[8] + b1*vec_back[8]; out[9] = a0*f[9] + a1*vec_forw[9] + b1*vec_back[9]; out[10] = a0*f[10] + a1*vec_forw[10] + b1*vec_back[10]; out[11] = a0*f[11] + a1*vec_forw[11] + b1*vec_back[11]; out[12] = a0*f[12] + a1*vec_forw[12] + b1*vec_back[12]; out[13] = a0*f[13] + a1*vec_forw[13] + b1*vec_back[13]; out[14] = a0*f[14] + a1*vec_forw[14] + b1*vec_back[14]; out[15] = a0*f[15] + a1*vec_forw[15] + b1*vec_back[15]; vec_forw[0] = f[0]; vec_forw[1] = f[1]; vec_forw[2] = f[2]; vec_forw[3] = f[3]; vec_forw[4] = f[4]; vec_forw[5] = f[5]; vec_forw[6] = f[6]; vec_forw[7] = f[7]; vec_forw[8] = f[8]; vec_forw[9] = f[9]; vec_forw[10] = f[10]; vec_forw[11] = f[11]; vec_forw[12] = f[12]; vec_forw[13] = f[13]; vec_forw[14] = f[14]; vec_forw[15] = f[15]; vec_back[0] = out[0]; vec_back[1] = out[1]; vec_back[2] = out[2]; vec_back[3] = out[3]; vec_back[4] = out[4]; vec_back[5] = out[5]; vec_back[6] = out[6]; vec_back[7] = out[7]; vec_back[8] = out[8]; vec_back[9] = out[9]; vec_back[10] = out[10]; vec_back[11] = out[11]; vec_back[12] = out[12]; vec_back[13] = out[13]; vec_back[14] = out[14]; vec_back[15] = out[15]; in += 16; out += 16; vec_forw += 16; vec_back += 16; n -= 16; } ff = in[0]; out[0] = a0*ff + a1*vec_forw[0] + b1*vec_back[0]; vec_forw[0] = ff; vec_back[0] = out[0]; return(w+5); } static void spec2_1p1z_time_tilde_dsp(t_spec2_1p1z_time_tilde *x, t_signal **sp) { int n = (sp[0]->s_n)/2; if(!x->x_blocksize)/*first time*/ { x->x_begmem_forw = (t_float *)getbytes(2 * (n+1) * sizeof(t_float)); x->x_blocksize = n; x->x_begmem_back = x->x_begmem_forw + n + 1; } else if(x->x_blocksize != n) { x->x_begmem_forw = (t_float *)resizebytes(x->x_begmem_forw, 2*(x->x_blocksize+1)*sizeof(t_float), 2*(n+1)*sizeof(t_float)); x->x_blocksize = n; x->x_begmem_back = x->x_begmem_forw + n +1; } if(n&15) dsp_add(spec2_1p1z_time_tilde_perform, 4, sp[0]->s_vec, sp[1]->s_vec, x, n); else dsp_add(spec2_1p1z_time_tilde_perf16, 4, sp[0]->s_vec, sp[1]->s_vec, x, n); } static void *spec2_1p1z_time_tilde_new(t_symbol *s, int argc, t_atom *argv) { t_spec2_1p1z_time_tilde *x = (t_spec2_1p1z_time_tilde *)pd_new(spec2_1p1z_time_tilde_class); outlet_new(&x->x_obj, &s_signal); x->x_blocksize = 0; x->x_begmem_forw = (t_float *)0; if(argc >= 3) spec2_1p1z_time_tilde_list(x, s, argc, argv); else { x->x_a0 = 1.0f; x->x_a1 = 0.0f; x->x_b1 = 0.0f; } x->x_msi = 0.0f; return (x); } static void spec2_1p1z_time_tilde_free(t_spec2_1p1z_time_tilde *x) { if(x->x_begmem_forw) freebytes(x->x_begmem_forw, 2 * (x->x_blocksize+1) * sizeof(t_float)); } void spec2_1p1z_time_tilde_setup(void) { spec2_1p1z_time_tilde_class = class_new(gensym("spec2_1p1z_time~"), (t_newmethod)spec2_1p1z_time_tilde_new, (t_method)spec2_1p1z_time_tilde_free, sizeof(t_spec2_1p1z_time_tilde), 0, A_GIMME, 0); CLASS_MAINSIGNALIN(spec2_1p1z_time_tilde_class, t_spec2_1p1z_time_tilde, x_msi); class_addlist(spec2_1p1z_time_tilde_class, (t_method)spec2_1p1z_time_tilde_list); class_addmethod(spec2_1p1z_time_tilde_class, (t_method)spec2_1p1z_time_tilde_dsp, gensym("dsp"), 0); // class_sethelpsymbol(spec2_1p1z_time_tilde_class, gensym("iemhelp2/spec2_1p1z_time~-help")); }