[PD] Analog/good-sounding oscillators in PD

Thu Apr 6 22:09:28 CEST 2006

This GOP patch takes as $1 the name of the table you want to use.  Do
a tabosc4~ and it sounds ok.

I think getting the waveformed tuned by hand and not simply by using
mathematically exact proportions gets you closer to the sound you
want, gives you a little randomness on the proportions.  For a nice
square wave for instance add just a tiny bit of the lower
even-numbered partials for a more realistic sound.  A problem you're
going to run into with waveforms with too high partials is aliasing,
though this diminishes the quality with lower fundamental
frequencies... tradeoffs of
the digital domain...

-david

On 4/6/06, chris clepper <cgc at humboldtblvd.com> wrote:
> You want band limited functions which Pd doesn't have by default.  Someone
> may have externals that do band limited saw and pulse waves available.
>
>
> On 4/6/06, cyborgk at nocturnalnoize.com <cyborgk at nocturnalnoize.com > wrote:
> > Hello everyone,
> >
> > I'm working on a set of abstractions that will allow users to easily
> > sequence rhythmic-pattern style music; basically, to make it easy to
> > create IDM/electro/minimal-techno style stuff. Anyway, there's one major
> > weakness to the whole project: PD oscillators don't sound good! Also, the
> > filters I've tried don't sound very analog at all.
> >
> > So, my question is:
> >
> > 1. Does anyone have any good, "analog-sounding" oscillators already
> > created? I guess I only need saw, PWM square, and triangle, but they need
> > to sound GOOD (like Reaktor, basically). I see how to do simple versions
> > from the example patches, but I don't know enough DSP to get rid of the
> > digital sound. For a great example of the kind of sound I want, check out:
> > Asynth vst from antti at smartelectronix.
> > http://antti.smartelectronix.com/
> >
> > 2. What would be the most analog sounding filter to use?
> >
> > 3. Is wavetable synthesis a reasonable option? How do I create a custom
> > table to use for such synthesis? For example, if I synthesize a wave of
> > one period in another program, how long should it be? I know this should
> > be easy, but my brain is broken on this question. My gut instinct is that
> > a wavetable would only sound good near original frequency or you just get
> > the typical digital artifacts from pitch-shifting a sample... is that
> > correct? I'm guessing what I'm really talking about is ROMpler style
> > multi-sampling,  and selecting the wave closest to input frequency. Don't
> > know if anyone built this already.
> >
> > Btw, I can solve the problem for myself by loading in the VST~ external,
> > but I want what I build to be cross-platform so that it is most useful to
> > interested people in the community.
> >
> >
> >
> >
> > _______________________________________________
> > PD-list at iem.at mailing list
> > UNSUBSCRIBE and account-management ->
> http://lists.puredata.info/listinfo/pd-list
> >
>
>
> _______________________________________________
> PD-list at iem.at mailing list
> UNSUBSCRIBE and account-management ->
> http://lists.puredata.info/listinfo/pd-list
>
>
>
-------------- next part --------------
#N canvas 401 557 394 168 10;
#X obj 14 20 vsl 12 96 0 1 0 1 \$0-1 \$0-1 empty 0 -8 0 8 -262131 -123526
-1 9500 0;
#X obj 26 20 vsl 12 96 0 1 0 1 \$0-2 \$0-2 empty 0 -8 0 8 -225271 -42246
-1 400 0;
#X obj 38 20 vsl 12 96 0 1 0 1 \$0-3 \$0-3 empty 0 -8 0 8 -225271 -42246
-1 3000 0;
#X obj 50 20 vsl 12 96 0 1 0 1 \$0-4 \$0-4 empty 0 -8 0 8 -225271 -42246
-1 200 0;
#X obj 62 20 vsl 12 96 0 1 0 1 \$0-5 \$0-5 empty 0 -8 0 8 -225271 -42246
-1 1600 0;
#X obj 74 20 vsl 12 96 0 1 0 1 \$0-6 \$0-6 empty 0 -8 0 8 -225271 -42246
-1 100 0;
#X obj 86 20 vsl 12 96 0 1 0 1 \$0-7 \$0-7 empty 0 -8 0 8 -225271 -42246
-1 900 0;
#X obj 98 20 vsl 12 96 0 1 0 1 \$0-8 \$0-8 empty 0 -8 0 8 -225271 -42246
-1 0 0;
#X obj 110 20 vsl 12 96 0 1 0 1 \$0-9 \$0-9 empty 0 -8 0 8 -262131
-123526 -1 600 0;
#X obj 122 20 vsl 12 96 0 1 0 1 \$0-10 \$0-10 empty 0 -8 0 8 -225271
-42246 -1 0 0;
#X obj 134 20 vsl 12 96 0 1 0 1 \$0-11 \$0-11 empty 0 -8 0 8 -225271
-42246 -1 300 0;
#X obj 146 20 vsl 12 96 0 1 0 1 \$0-12 \$0-12 empty 0 -8 0 8 -225271
-42246 -1 0 0;
#X obj 158 20 vsl 12 96 0 1 0 1 \$0-13 \$0-13 empty 0 -8 0 8 -225271
-42246 -1 200 0;
#X obj 170 20 vsl 12 96 0 1 0 1 \$0-14 \$0-14 empty 0 -8 0 8 -225271
-42246 -1 0 0;
#X obj 182 20 vsl 12 96 0 1 0 1 \$0-15 \$0-15 empty 0 -8 0 8 -225271
-42246 -1 100 0;
#X obj 194 20 vsl 12 96 0 1 0 1 \$0-16 \$0-16 empty 0 -8 0 8 -225271
-42246 -1 0 0;
#X obj 214 18 cnv 15 150 106 empty empty empty 20 12 0 14 -260818 -66577
0;
#N canvas 0 0 450 300 graph3 0;
#X array \$1 4099 float 0;
#X coords 0 1 4098 -1 140 96 1;
#X restore 219 23 graph;
#N canvas 300 135 651 325 guts 0;
#X obj 234 -60 r \$0-1;
#X obj 252 -10 r \$0-2;
#X obj 264 40 r \$0-3;
#X obj 284 84 r \$0-4;
#X obj 285 -60 r \$0-5;
#X obj 303 -10 r \$0-6;
#X obj 315 40 r \$0-7;
#X obj 335 84 r \$0-8;
#X obj 338 -60 r \$0-9;
#X obj 356 -10 r \$0-10;
#X obj 368 40 r \$0-11;
#X obj 388 84 r \$0-12;
#X obj 397 -60 r \$0-13;
#X obj 415 -10 r \$0-14;
#X obj 427 40 r \$0-15;
#X obj 447 84 r \$0-16;
#X msg 248 155 \; \$1 sinesum 4096 \$2 \$3 \$4 \$5 \$6 \$7 \$8 \$9
\$10 \$11 \$12 \$13 \$14 \$15 \$16 \$17 \; \$1 normalize;
#X obj 184 72 symbol \$1;
#X obj 248 132 pack s f f f f f f f f f f f f f f f f;
#X obj 234 -39 t b f;
#X obj 285 -39 t b f;
#X obj 338 -40 t b f;
#X obj 397 -39 t b f;
#X obj 252 11 t b f;
#X obj 303 11 t b f;
#X obj 356 10 t b f;
#X obj 415 11 t b f;
#X obj 264 61 t b f;
#X obj 315 61 t b f;
#X obj 368 60 t b f;
#X obj 427 61 t b f;
#X obj 284 104 t b f;
#X obj 335 104 t b f;
#X obj 388 103 t b f;
#X obj 447 104 t b f;
#X connect 0 0 19 0;
#X connect 1 0 23 0;
#X connect 2 0 27 0;
#X connect 3 0 31 0;
#X connect 4 0 20 0;
#X connect 5 0 24 0;
#X connect 6 0 28 0;
#X connect 7 0 32 0;
#X connect 8 0 21 0;
#X connect 9 0 25 0;
#X connect 10 0 29 0;
#X connect 11 0 33 0;
#X connect 12 0 22 0;
#X connect 13 0 26 0;
#X connect 14 0 30 0;
#X connect 15 0 34 0;
#X connect 17 0 18 0;
#X connect 18 0 16 0;
#X connect 19 0 17 0;
#X connect 19 1 18 1;
#X connect 20 0 17 0;
#X connect 20 1 18 5;
#X connect 21 0 17 0;
#X connect 21 1 18 9;
#X connect 22 0 17 0;
#X connect 22 1 18 13;
#X connect 23 0 17 0;
#X connect 23 1 18 2;
#X connect 24 0 17 0;
#X connect 24 1 18 6;
#X connect 25 0 17 0;
#X connect 25 1 18 10;
#X connect 26 0 17 0;
#X connect 26 1 18 14;
#X connect 27 0 17 0;
#X connect 27 1 18 3;
#X connect 28 0 17 0;
#X connect 28 1 18 7;
#X connect 29 0 17 0;
#X connect 29 1 18 11;
#X connect 30 0 17 0;
#X connect 30 1 18 15;
#X connect 31 0 17 0;
#X connect 31 1 18 4;
#X connect 32 0 17 0;
#X connect 32 1 18 8;
#X connect 33 0 17 0;
#X connect 33 1 18 12;
#X connect 34 0 17 0;
#X connect 34 1 18 16;
#X restore 188 213 pd guts;
#X coords 0 -1 1 1 360 125 1 10 5;