[PD] sigmund~ track analysis errors

[João Pais]

Hi list,

Related to the bug report https://github.com/pure-data/pure-data/issues/872
, I wanted to drop some questions about sigmund:
- when using the "-npeak x tracks" settings, the X stands for the maximum
number of tracks, and no more are allowed. Is this the expected behavior?
It's not documented in the help file.


More important, I was getting inconsistent track analysis results with
sigmund~, and I think I found a way of confirming this. Basically, after
the first analysis lots of low frequencies disappear - only
resetting/recreating the object makes it possible to do the analysis.

The patch attached should be clear, can anyone confirm if this also happens
with other systems?
(done with windows 7, pd 64b)

Best,

Joao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20200301/2fff4b01/attachment.html>
-------------- next part --------------
#N canvas 123 72 516 603 10;
#X obj 19 398 print peak;
#N canvas 0 50 450 300 (subpatch) 0;
#X array test 155944 float 2;
#X coords 0 1 155944 -1 200 140 1 0 0;
#X restore 73 446 graph;
#X obj 304 533 phasor~;
#X obj 304 445 loadbang;
#X obj 304 481 440;
#X floatatom 303 508 5 0 0 0 - - -;
#X obj 280 536 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144
-1 -1;
#X msg 350 481 \; pd dsp 1;
#X obj 19 259 until;
#X obj 20 238 20;
#X obj 20 217 t b b;
#X obj 20 196 bng 15 250 50 0 empty empty empty 17 7 0 10 -4034 -1
-1;
#X obj 313 325 soundfiler;
#X obj 313 285 openpanel;
#X obj 313 258 bng 15 250 50 0 empty empty empty 17 7 0 10 -262144
-1 -1;
#X obj 19 326 t f f;
#X obj 76 316 + 204.8;
#X obj 38 286 0;
#X obj 19 306 f;
#X obj 295 564 tabwrite~ test;
#X msg 19 346 list test 2048 \$1 44100 0;
#X msg 313 305 read -resize \$1 test;
#X text 44 194 analyse;
#X obj 19 372 sigmund~ -t -npts 2048 -npeak 100 tracks;
#X text 243 235 (try it with non-regular signals as well);
#X text 23 12 Sigmund~ track analysis bug;
#X text 19 34 1 - do one analysis by clicking the green bang. Save
the results in a spreadsheet or similar \; 2 - do another analysis
\, and compare the results with the first one. There will be lots of
differences \; 3 - any further analysis results will match the 2nd
analysis (with very few deviations). These results won't be correct
\, as after the 1st analysis many lower frequencies will be ignored
- this is noticed in a patch such as doc4.data.structures14.partialtracer.pd
;
#X text 19 144 The only way to get another correct analysis is to reset
sigmund~ by editing it. The results of different analysis will match
100% \, and will have all frequencies.;
#X connect 2 0 19 0;
#X connect 3 0 7 0;
#X connect 3 0 4 0;
#X connect 4 0 5 0;
#X connect 5 0 2 0;
#X connect 5 0 6 0;
#X connect 6 0 19 0;
#X connect 8 0 18 0;
#X connect 9 0 8 0;
#X connect 10 0 9 0;
#X connect 10 1 17 0;
#X connect 11 0 10 0;
#X connect 13 0 21 0;
#X connect 14 0 13 0;
#X connect 15 0 20 0;
#X connect 15 1 16 0;
#X connect 16 0 18 1;
#X connect 17 0 18 1;
#X connect 18 0 15 0;
#X connect 20 0 23 0;
#X connect 21 0 12 0;
#X connect 23 0 0 0;