[noise~] inlet pin???
Does anybody know what the inlet on the noise object does? It's a message / non-signal inlet, but it doesn't respond to numbers as far as I can tell. There is absolutely no documentation of it anywhere that I can find...
Maybe someone c-savvy can go over the source code???
@bocanegra Inactive...... not used......
David.
AFAIK, in the last version the float message is used to seed the PRNG.
https://github.com/pure-data/pure-data/blob/master/src/d_osc.c#L497
Oops : It is the "seed" message that does this now.
@FUD said:
AFAIK, in the last version the float message is used to seed the PRNG.
https://github.com/pure-data/pure-data/blob/master/src/d_osc.c#L497
That is exactly the behavior I was looking for, but I can't reproduce it in my PD (latest stable portable windows version).
I'm trying to reproduce the behavior of a PRNG object from another modular program (ABox), which can be stepped by a pulse signal at any frequency, seeded/reseeded and of course gives a signal output (no concept of messages in ABox, all is signals).
My workaround so far is an array loaded with white noise, but I can't step it at arbitrary frequencies to be in sync with say the cycles of a ramp going at 330 Hz unless I use samphold~ at the end of it. The sync is crucial you see... So is the ability to reseed the mother.
Also, no matter how long I make the array I get clicks when it loops
As @whale-av noticed the float message does nothing. The seed message seeds the PRNG.
@bocanegra I don't think it is for the inlet.
It looks as though for each instance the prng value (x) is changed (much like $0).
@FUD I tried doing [seed ###< with different numbers and no notable result going through the array list view (I.e. repeoducing the same sequence)
The [noise~] PRNG is a simple (weird) LCG. It MUST produce the same sequence with equals seeds.
@bocanegra I tried with pd-0.48-1test2-ia64.......
The terminal will tell you if your message is incorrect. seed /set / float etc. all produce errors.
@jancsika might have implemented it for Purr Data (or might be willing to do so.....?).
it seems the seed message to the inlet wasn't added till pd 0.49
@whale-av said:
@bocanegra I don't think it is for the inlet.
It looks as though for each instance the prng value (x) is changed (much like $0).
Ah, so an aproximation of "true random", making the object useless for my purpose...
Looks like I'll have to keep working with arrays then...
@seb-harmonik.ar ; Yep, i should have said that i use Pd-0.49-0test3, sorry.
Well, to give you an idea of what I'm trying to achieve and how I'm going on about it so far, I'm attaching my work in progress plus an ABox patch with some example of homegrown drum sounds I'm trying to redo in PD.
The NO-CTRL is the demo patch for PD and requires NoiseOsc~.pd and NZ1024.pd to run.
If you wanna check out the ABox circuit, get ABox242 here. It's a standalone windows executable. Runs just fine with wine on both linux and mac
Basically I'm modulating the amplitudes of half sine-cycles with stepped PRNs. Turns out to be extremely useful for realistically noisy sidebands found in all sorts of percussion, when tweaked right, and coupled with a suitable reverberator. Again, check the ABox examples if you don't believe me.
If I could reseed the noise~ object in PD, I'd be able to clock a couple of samphold~ objects and be done with it. For now, the array solution is ok, although not as elegant as I dreamed it up. Ideally I'd have two tabread~ objects running in parralel with -1 index offset on the second, but getting that setup to sync with the tone osc has not worked out for me yet, hence the samphold~ shift register thingy.
There is a bunch of comment text inside the NoiseOsc~ patch that further elaborates on the intricacies of the design and its shortcomings.
For explanations of the mess inside my ABox patch, feel free to ask, but dont expect a coherrent answer 
Oops! Looks like something went wrong!
