Symbols can be used in data structures. However, you can't just feed [append] (or even plain [set]) with symbols. The workaround is to use the pointer of the created data structure to edit its symbols with a [set] to which you give the -symbol argument.

Well, unfortunately, as the help says, "with a 512 sample delay, it begins to attenuate at about 110Hz". To keep decent low frequencies I have to use at least 2048 according to my quick test. No easy solution to the problem I guess.

Hello,

I can't find the source to it anymore, but to remove DC offset I use

[rzero~1]
|
[rpole~ 0.995]

which is practically perfect for my use, except it messes with waveforms due to inherent phase shift. Is there any way to cancel/revert this phase shift with, say, an allpass filter ? If so, how should I tune that?

The state saving system and dynamic gui modules are worth a look. I haven't delved in the rest yet.

Maybe should you have

[loadbang}
|
[float 6]
|
[until]

instead? Anyway, you can execute pd with the -noloadbang option to edit your patch.

@oid said:

Non-linear elements in the feedback path can be fun and do some interesting things but far from a general purpose filter, no clue if pd will play well with that sort of abuse.

In my humble experience: very well. I've been working on an "acoustic feedback" patch for a while now. With allpass/diffusion, peak filter and basic distortion in the loop, it was quite lively and convincing.

Thank you, it did work here too!

Hi,

Has anybody managed to get pd spectral toolkit working?
I'm especially interested in the GEM waterfall spectrum but alas, i get

pd-externals/SpectralToolkit/pd_spectral_toolkit.l_ia64: undefined symbol: _ZGVbN2v_exp
pd_spectral_toolkit: can't load library

on startup. make linux didn't complain about anything.

I love this feedback oscillator! I don't know how relevant it is, but it reminded me of https://ccrma.stanford.edu/software/snd/snd/fm.html [cascade FM: sin(sin(sin))]:

As z increases above 1.27, we get a square wave, then period doubling, and finally (ca. 1.97) chaos.

@chaosprint said:

For example, I would like to trigger a sound playback multiple times
it layers the playback at that particular rate

You mean, like a polyphonic sampler? If so I'd load the sample in a table and have an instance of tabread~/tabread4~ for each "layer".

@jameslo said:

If you follow the algorithm in http://cgm.cs.mcgill.ca/~godfried/publications/banff.pdf, E(5,13) = 1001010010100, But if you use the much simpler one from @Stutter you get 1001001010010.

If you flip stutter's result three digits to the left you get banff's. It's a matter of a [- 3] somewhere.

If I had to buy something today I'd go for Intech EN16. Fighting my impulse buy a while back, I dug out my old dusty bcr2000, found out it had a software editor and could be set in absolute mode. And I don't need anything more.

With encoders set to absolute, no need for two-way communication, no more jumps in values, everything stays in the patch. I have a set of WIP abstractions that steal the focus of the shared unique controller, map and scale CC increments/decrements to the relevant patch parameters and deals with store/restore presets.

I just wish there were something to faders like what encoders are to knobs.

http://www.pd-tutorial.com/english/ch03s08.html was a great help for me back in time.

@ddw_music you were right, problem solved :

@ddw_music very nice! With your downward edge detection I can now divide a phasor~ by an arbitrary number. There's a thing I cannot wrap my head around though: how to reset the counter to zero every divided downward edge. So it can safely be run indefinitely.

I don't understand why feeding the rpole~ argument with ((1-(divided downward edge detection)) delayed by one sample) only resets it once in a while and not every cycle.

@RT-Chris you should have a look at how it's done in Automatonism
@jameslo absolutely brilliant!

so I totally gave up on solving my maths problems in pd. I used FAUST instead and run my tonestack~.dsp in [faustgen~].
you can access the 3 parameters with [low/med/top $1( from 0 to 1.
first time with FAUST, looks awesome.

Wew lads! What a complicated topic. Finding real roots for the denominator, finding either 3 real or 1 real and 2 conjugated complex roots for the numerator... The current filter is 100% highly unstable, I hope I'll get it working eventually.

so far my best guess would be https://en.wikipedia.org/wiki/Cubic_equation#Trigonometric_solution_for_three_real_roots

Hi,
I'd like to implement in pure data the tone stack model I found in the DAFX book. It's a linear third order transfer function. In the book they use mathlab which seems to have a routine to solve it so the code they give is of no use to me.
In pd, should I find the roots of both third order polynomials and use poles and zeros objects like in Maelstorm tutorial? Does his method transpose? Should I use fexpr~ instead? Is there another method?