I've been digging into this a little more, it seems I can reproduce the
unexpected behavior with just the following lines.
> samphold(c, x) = x * s : + ~ *(1-s) with {
> s = (c' < c);
> };
>
> master = os.phasor(1.0, masterFreq); // masterFreq is something like 4Hz
process = abs(no.noise) : samphold(master);
When I compile this with the `plot.cpp` architecture and look at the output
samples, it's effectively just an abs noise stream. So obviously, something
seems wrong with my `samphold` implementation.
I'm comparing again to PureData's `samphold~` implementation, and Faust's
`ba.latch':
https://github.com/pure-data/pure-data/blob/07a14c540630bc2344a80da30464405900d556fb/src/d_filter.c#L533
https://github.com/grame-cncm/faust/blob/master-dev/libraries/basics.lib#L723
And oddly enough, my `samphold` implementation there is working as expected
in the other parts of my dsp. So now I'm really confused by it!
Would love to hear your thoughts if you see what I'm doing wrong.
Thanks,
Nick
On Thu, Jul 20, 2017 at 10:22 AM, Nick Thompson <[email protected]> wrote:
> Hi all,
>
> I'm working on a project similar to the Granulator effect in the Faust
> Playground modules:
> https://github.com/grame-cncm/faustplayground/blob/master/fa
> ust-modules/effects/Granulator.dsp
>
> But I'm struggling to adapt what I see in the Granulator implementation to
> fit my use case. Specifically, I have an audio stream, and a "master clock"
> phasor, and what I want is that on every new cycle of the master phasor, a
> gate for the audio stream either opens or closes randomly (i.e. multiply by
> 1.0 or 0.0).
>
> Here are the important bits of my implementation:
>
> samphold(c, x) = x * s : + ~ *(1-s) with {
>> s = (c' < c);
>> };
>>
>
>
>> master = os.phasor(1.0, masterFreq); // masterFreq is something like 4Hz
>>
>
>
>> ratio_env = 0.4;
>> pulsar = ((master < ratio_env) : @(100)) * (probability > (abs(no.noise)
>> : samphold(master)));
>> process = myStream * pulsar;
>
>
> The `pulsar` line is borrowed almost verbatim from the Granulator.dsp
> file, except that I use the `master` phasor here explicitly instead of the
> `pulsar` input stream. Also, Granulator.dsp uses `ba.latch` where I'm using
> my own `samphold` implementation that more closely matches PureData's
> implementation, as I've had success with this patch in PD but not in Faust.
>
> With my implementation here, the output result is basically `myStream`
> mixed with noise, and I don't understand why? I have a similar result when
> I just use `pulsar = (probability > (abs(no.noise) : samphold(master)))` as
> well... why would that not work alone? What is the other part, `((master <
> ratio_env) : @(100))` supposed to be doing in Granulator.dsp anyway? I
> don't really understand that part.
>
> Any help here would be greatly appreciated!
>
> Thank you,
> Nick
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Faudiostream-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/faudiostream-users
