WebPd patch tester

[Brandon James]

Hey Guys,

I knocked together a little tool for testing WebPd patches.

It's just 8 toggles and 8 sliders already set up with receives in the PD patch.
The range for all the controls is 0 to 1.
Slider8 and Toggle8 are set to control the volume and turn the sound off by default

[Brandon James]

I should add that the code for the controller is both long winded and
amateurish and probably useless to anyone that would be here, but I
think the controller itself could be pretty handy.

>  WebPd_Tester.7z
> 32KViewDownload

[Brandon James]

And here's an FM tone thing-a-majig that I made, mostly just cause  after staring at all those sliders I wanted to make them do something.

[Spencer Kelly]

this looks great. I haven't been able to get any sound out of it though
i open the index.html and it prints the slider info, but no sound
maybe it needs a pd.play() somewhere?
i will try to get the slider code into the ui. it looks awesome. its great too that you seem to have figured out live interaction...

i just committed a bunch of objects, including:  log, sin, cos, tan, abs, <<, >>, key, keyup, keyname.

oh, and i think the select bug is a bug across all objects. nothing seems to come out of outlet 1 (the 2nd one), anywhere.

will try to get some hacking time with the UI this weekend. i have big plans for it.

[Brandon James]

> this looks great. I haven't been able to get any sound out of it though
> i open the index.html and it prints the slider info, but no sound
> maybe it needs a pd.play() somewhere?

You need to turn on toggle 8 and move slider 8 to adjust the volume.
Then move slider 1 to set a frequency.
I didn't set any default values, so you have to move a slider to
initialize it.

> i will try to get the slider code into the ui. it looks awesome. its great
> too that you seem to have figured out live interaction...

Thanks a lot. I'm still trying to figure out a good way to make an
object in processing that goes in the draw loop, but still responds
properly to mouse clicks, but I'm getting there slowly.

> i just committed a bunch of objects, including:  log, sin, cos, tan, abs,
> <<, >>, key, keyup, keyname.

I saw those. Awesome. I'm slowly working your objects into my branch
as I go over them and make unittests.
I'm not sure if this is proper version control etiquette, but they
sure make testing easier.

> oh, and i think the select bug is a bug across all objects. nothing seems to
> come out of outlet 1 (the 2nd one), anywhere.

Trigger works. If you check my branch, there's another version of
[sel] in there that I wrote that works with multiple arguments, but
still doesn't pass on rejects. Probably because I still don't fully
know what I'm doing yet.

> will try to get some hacking time with the UI this weekend. i have big plans
> for it.

Looking forward to it.

[Brandon James]

Also, after some experimenting, I got the best results keeping the Pd
buffersize as low as possible.
The default 4410 is way too high for any kind of interaction, and is
probably responsible for all that clicking in the Demo file.
256 worked great on my really old slow computer, but 128 caused it to
overflow.
It really does wonders for smoothing control signals from
processing.js.
Connect the receive to a line~ set at about 5ms and it should be
smooth and responsive.

[Spencer Kelly]

On Thu, Oct 14, 2010 at 7:51 PM, Brandon James <xbrando...@gmail.com> wrote:

Thanks a lot. I'm still trying to figure out a good way to make an
object in processing that goes in the draw loop, but still responds
properly to mouse clicks, but I'm getting there slowly.

ha, ya me too. it's totally a senseless way of programming.

I saw those. Awesome. I'm slowly working your objects into my branch
as I go over them and make unittests.
I'm not sure if this is proper version control etiquette, but they
sure make testing easier.

thanks. i never figured out that bit.

Trigger works.

good point.

=
hey,
who are you in real life brandon? its good working with you. do you have a twitter or something where I can snoop?

I'm 24, from montreal. I sell frozen yogurt in the daytime.
http://spencerwaterbed.com/soft/

[Brandon James]

> ha, ya me too. it's totally a senseless way of programming.

So I finally decided to make two objects instead of trying to cram it
all into one. One for the data and one for the picture.
I think this will be better in the long run anyway, since it will make
skinning and automation simpler.
I'll post a set of gui controls after I get things streamlined

> who are you in real life brandon?

I'm a 29 year old unemployed musician/DJ in Cincinnati.
I don't have much of a web presence, so not much snooping to be done.
I'm currently working on a project that I'm hoping to be mostly if not
entirely web based, which is how I found WebPd.
I have this: http://www.myspace.com/psykological , but don't use it
much.

[Spencer Kelly]

ya 256 works better on mine too.
I don't know anything about performance, but anyone know if jaegermonkey or any of the new javascript optimizations will improve audio speed?
performance will ofcourse be an issue with javascript audio, but what can we expect?

[chrism]

Hey Brandon,

Thanks for this, I will set that to be the default buffer from now on!

Chris.

--
-------------------
http://mccormick.cx

[Brandon James]

>i think the select bug is a bug across all objects. nothing seems to
>come out of outlet 1 (the 2nd one), anywhere.

I think the problem was that when we were creating the outlet array,
we were only creating an outlet per argument.
So the reject outlet was never created since it doesn't have an
argument.

Check my branch for [sel] and [route] objects that work.