First post here, but I've been using Pure Data for some time now.

I'm building a simple implementation of Pulsar Synthesis in Purr-Data. It has a couple pulsaret waveform options and windowing options. I hope to extend it further and upload it here if others are interested!

For some reason though, even with this basic version, I'm getting audio dropouts. Usually I can figure these problems out but I'm drawing a blank.

Basically I'm tabread4ing through the pulsaret and window tables, and I'm allowing the phasor to read past the end of the table to create the silence between the pulsarets.

I was hesitant to do it this way for fear of a buffer overrun error, but from what I understand, this is safe to do in PD. Maybe I'm wrong?

I don't know if it's that, though, because I only seem to get the audio dropouts every 10 seconds or so.

The patch is also only using a negligible amount of CPU. I've had far more elaborate patches with dozens of tables being read and I didn't get these kinds of dropouts.

I haven't ruled out a problem with JACK or something else on my computer (Linux). Investigating.

Here is the patch: pulsar.pd

Any ideas?

Hi pure_datum,
the patch you uploaded is all messed up, i'm getting:

canvas: no method for 'dropdown'
canvas: no method for 'dropdown'
arraycopy v
... couldn't create
1557172189159-pulsar-1.pd 9 0 26 0 (*~->/~) connection failed
1557172189159-pulsar-1.pd 15 0 27 1 (receive->/~) connection failed
1557172189159-pulsar-1.pd 16 0 27 0 (receive->/~) connection failed
1557172189159-pulsar-1.pd 19 0 14 0 (dac~->outlet) connection failed
1557172189159-pulsar-1.pd 19 0 20 0 (dac~->canvas) connection failed
1557172189159-pulsar-1.pd 19 0 20 1 (dac~->canvas) connection failed
1557172189159-pulsar-1.pd 26 0 25 0 (/~->???) connection failed
1557172189159-pulsar-1.pd 27 0 9 0 (/~->???) connection failed

..in pd-0.48-1.
does it work for you when you download it again?

@solipp there are some Purr-Data objects in my patch.

It looks like you can’t load some of those objects because you’re running PD-vanilla.

yes, there is one arraycopy object missing. But the rest of the patch is rendered completely useless. I am aware of some of the incompatibilities between purr data and pd-vanilla, but something like this should really not happen. There is no way to recreate the functionality of this patch in pd-vanilla..

@solipp I see what you mean. That's very strange. It doesn’t do that when I download it. Thank you for pointing this out. I'll figure out what's going on and re-upload a new patch.

I've now replaced arraycopy with a vanilla friendly array copier. It's better to do that anyway if I plan to give this to other people.

I've also added radio buttons in place of the dropdowns for vanilla compatibility.

Other than that I have no idea why things that are connected would not connect for you. If you could check if this works, I would appreciate it: pulsar.pd

Yes, it seems that the dropdowns where causing this. All objects are connected now and i don't get any error messages. But another strange thing occurs now; it seems that Pd does not accept "$" in the send symbols of the radio buttons. This seems to be a bug since it works in the number boxes and i think that i read about this somewhere..
It will definitely work when you put $0 at the start of your send symbol, like "$0-envelope".
However, after fixing this, i got no audio dropouts. Your method of switching between the envelopes is strange though, why don't you use a set message for tabread4~?

@solipp Thanks for looking at this for me.

The fact that you're not getting drop outs helps narrow it down to something other than pd. It must be something else on my computer,

Good point about the envelope switching. That was just a quick thing I cobbled together by copying and pasting some stuff from another patch. I'll optimize that later.

@solipp I have no idea why, but some GUI objects require # instead of $ for their send/receive names...... and some will change $ to # as they are saved and reopened.
Then they work as expected.....
David.

@whale-av it's a bug.
https://github.com/pure-data/pure-data/issues/558