Re: [PD] Issue with Gaussian Noise
On Wed, 2010-08-11 at 00:17 -0400, Martin Peach wrote: > On 2010-08-10 04:32, Frank Barknecht wrote: > Also an uncertain execution order in the log divide part needs one of > those [tff]s. HAHAHA, I would have never dreamed that once someone would be saying that to Frank. (Frank, please get me right: I meant to laugh with you, not at you). Cheers Roman ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] Issue with Gaussian Noise
On Wed, Aug 11, 2010 at 09:44:13AM +0200, Roman Haefeli wrote: > On Wed, 2010-08-11 at 00:17 -0400, Martin Peach wrote: > > On 2010-08-10 04:32, Frank Barknecht wrote: > > > Also an uncertain execution order in the log divide part needs one of > > those [tff]s. > > HAHAHA, I would have never dreamed that once someone would be saying > that to Frank. > > (Frank, please get me right: I meant to laugh with you, not at you). Yeah, it's funny! Lesson learned here: If you start with DSP code and convert it to messages, look very, very closely for fanning connections. Ciao -- Frank ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] Issue with Gaussian Noise
On Wed, 2010-08-11 at 10:01 +0200, Frank Barknecht wrote: > On Wed, Aug 11, 2010 at 09:44:13AM +0200, Roman Haefeli wrote: > > On Wed, 2010-08-11 at 00:17 -0400, Martin Peach wrote: > > > On 2010-08-10 04:32, Frank Barknecht wrote: > > > > > Also an uncertain execution order in the log divide part needs one of > > > those [tff]s. > > > > HAHAHA, I would have never dreamed that once someone would be saying > > that to Frank. > > > > (Frank, please get me right: I meant to laugh with you, not at you). > > Yeah, it's funny! Lesson learned here: If you start with DSP code and > convert it to messages, look very, very closely for fanning connections. Ah, I understand. There must have been a plausible explanation for that ;-) Roman ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] Issue with Gaussian Noise
Hi, On Wed, Aug 11, 2010 at 12:17:14AM -0400, Martin Peach wrote: > Yes it is the reason for the valleys. Your random is on [-0.5,0.5] > instead of [-1,1]. Ah, of course, how silly. > Also an uncertain execution order in the log divide part needs one > of those [tff]s. Oops, I overlooked this while converting the DSP version to floats, thanks for pointing it out. Now the bell sounds like a bell! Ciao -- Frank ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] Issue with Gaussian Noise
On 2010-08-10 04:32, Frank Barknecht wrote: Hi, I made two generators for Gaussian white noise, one using DSP, the other as message based version. Both use the same algorithm, but the histogram analysis of both shows a defect with the message version: The histogram has a visible "valley" around the central expectation value, so it doesn't look like a real gaussian bell curve anymore. The message version is based on [random] not [noise~], and I "faked" a true frand by taking and scaling random values from the range 0-10. Could this be the reason for the valleys or is the reason the custom algorithm [random] uses or is it, that two [random]s are not sufficiently independent? Yes it is the reason for the valleys. Your random is on [-0.5,0.5] instead of [-1,1]. The random needs to be multiplied by 2 million before subtracting one million then dividing by one million. Then it works. Also an uncertain execution order in the log divide part needs one of those [tff]s. The DSP version looks normal to me if I sample it a few times. Martin ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
[PD] Issue with Gaussian Noise
Hi, I made two generators for Gaussian white noise, one using DSP, the other as message based version. Both use the same algorithm, but the histogram analysis of both shows a defect with the message version: The histogram has a visible "valley" around the central expectation value, so it doesn't look like a real gaussian bell curve anymore. The message version is based on [random] not [noise~], and I "faked" a true frand by taking and scaling random values from the range 0-10. Could this be the reason for the valleys or is the reason the custom algorithm [random] uses or is it, that two [random]s are not sufficiently independent? Ciao -- Frank #N canvas 96 55 817 467 10; #X floatatom 81 182 5 0 0 0 - - -; #N canvas 484 63 422 413 histogram-view 0; #X obj 119 119 int; #X obj 119 181 t f f; #X obj 119 203 tabread \$0-histo; #X obj 119 323 tabwrite \$0-histo; #X obj 119 64 * 1; #X obj 119 86 + 1; #X obj 119 245 + 0.05; #X obj 119 24 inlet; #X obj 119 272 clip 0 1; #X connect 0 0 1 0; #X connect 1 0 2 0; #X connect 1 1 3 1; #X connect 2 0 6 0; #X connect 4 0 5 0; #X connect 5 0 0 0; #X connect 6 0 8 0; #X connect 7 0 4 0; #X connect 8 0 3 0; #X restore 249 235 pd histogram-view; #N canvas 0 0 450 300 (subpatch) 0; #X array \$0-histo 2 float 0; #X coords 0 1 1 0 200 140 1; #X restore 465 177 graph; #X msg 81 118 bang; #X obj 307 117 s \$0-histo; #X obj 249 150 until; #X obj 249 127 max; #X msg 307 94 const 0; #X obj 249 178 gaussian-random 0 0.01; #X msg 249 94 1; #X msg 212 45 seed1 12 \, seed2 192313; #X obj 81 145 gaussian-random; #X connect 3 0 11 0; #X connect 5 0 8 0; #X connect 6 0 5 0; #X connect 7 0 4 0; #X connect 8 0 1 0; #X connect 9 0 6 0; #X connect 10 0 8 0; #X connect 11 0 0 0; #N canvas 309 24 864 609 10; #X text 337 37 Simplified Box-Muller method as described on: http://www.dspguru.com/dsp/howtos/how-to-generate-white-gaussian-noise ; #X text 53 143 V1; #X text 211 143 V2; #X text 136 212 S; #X text 84 482 X; #X obj 320 176 inlet; #X text 325 152 variance; #X obj 108 384 sqrt; #X obj 266 178 inlet; #X text 270 153 mean; #X obj 318 210 loadbang; #X obj 318 263 select 0; #X msg 318 291 1; #X obj 318 239 f \$2; #X obj 82 144 t f f f; #X obj 146 144 t f f; #X obj 135 178 *; #X obj 101 178 *; #X obj 101 210 +; #X obj 102 237 log; #X obj 102 264 /; #X obj 102 291 * -2; #X obj 102 318 sqrt; #X obj 83 354 *; #X obj 83 409 * 1; #X obj 83 433 + \$1; #X obj 83 458 outlet; #X obj 82 77 t b b; #X obj 82 48 route bang seed1 seed2; #X msg 134 74 seed \$1; #X msg 205 76 seed \$1; #X obj 82 18 inlet; #N canvas 0 0 645 530 frand 0; #X obj 119 183 outlet; #X obj 119 41 inlet; #X obj 119 105 random 1e+06; #X obj 119 151 / 1e+06; #X obj 119 128 - 50; #X connect 1 0 2 0; #X connect 2 0 4 0; #X connect 3 0 0 0; #X connect 4 0 3 0; #X restore 146 111 pd frand; #N canvas 0 0 645 530 frand 0; #X obj 119 183 outlet; #X obj 119 41 inlet; #X obj 119 105 random 1e+06; #X obj 119 151 / 1e+06; #X obj 119 128 - 50; #X connect 1 0 2 0; #X connect 2 0 4 0; #X connect 3 0 0 0; #X connect 4 0 3 0; #X restore 82 112 pd frand; #X connect 5 0 7 0; #X connect 7 0 24 1; #X connect 8 0 25 1; #X connect 10 0 13 0; #X connect 11 0 12 0; #X connect 11 1 7 0; #X connect 12 0 7 0; #X connect 13 0 11 0; #X connect 14 0 23 0; #X connect 14 1 17 0; #X connect 14 2 17 1; #X connect 15 0 16 0; #X connect 15 1 16 1; #X connect 16 0 18 1; #X connect 17 0 18 0; #X connect 18 0 19 0; #X connect 18 0 20 1; #X connect 19 0 20 0; #X connect 20 0 21 0; #X connect 21 0 22 0; #X connect 22 0 23 1; #X connect 23 0 24 0; #X connect 24 0 25 0; #X connect 25 0 26 0; #X connect 27 0 33 0; #X connect 27 1 32 0; #X connect 28 0 27 0; #X connect 28 1 29 0; #X connect 28 2 30 0; #X connect 29 0 33 0; #X connect 30 0 32 0; #X connect 31 0 28 0; #X connect 32 0 15 0; #X connect 33 0 14 0; #N canvas 426 51 797 591 10; #N canvas 0 0 450 300 (subpatch) 0; #X array \$0-gaussian 2 float 2; #X coords 0 1 2 -1 200 140 1; #X restore 355 158 graph; #X obj 79 189 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144 -1 -1; #X floatatom 162 70 5 0 500 1 variance - -; #X obj 162 89 / 100; #X obj 217 116 tgl 15 0 empty empty switch 17 7 0 10 -262144 -1 -1 0 1; #X floatatom 108 32 5 0 0 1 mean - -; #X obj 108 51 / 100; #X obj 107 216 tabwrite~ \$0-gaussian; #X text 254 15 Gaussian White Noise; #X obj 230 141 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144 -1 -1; #N canvas 484 63 814 486 histogram-view 0; #N canvas 0 0 450 300 countto 0; #X obj 99 58 inlet; #X obj 99 177 f 0; #X obj 127 177 + 1; #X msg 142 143 0; #X obj 99 208 outlet; #X obj 99 142 until; #X obj 99 117 t a b; #X obj 99 96 max; #X connect 0 0 7 0; #X connect 1 0 2 0; #X connect 1 0 4 0; #X connect 2 0 1 1; #X connect 3 0 1 1; #X connect 5 0 1 0; #X connect 6 0 5 0; #X connect 6 1 3 0; #X connect 7 0 6 0; #X restore 119 117 pd countto; #X obj 119 162 tabread \$0-gaussian; #X obj 119 249 int; #X obj 119 271 t f f; #X obj 119 293 tabread \$0-histo; #X obj 119 383 tabwrit
