posted 04 July 2006 19:10            

I would like to be able to select their order, and whether serial or parallel, from the VCS. I guess I could create 3 Sound chains and use a Crossfade to switch between them, but I was wondering if there's a dedicated switch Sound that I could could plug the two distortion Sounds into, that would combine them in any way I like (in realtime). If I use the Crossfade method I'll have to create 3 chains, each with two instances of the distortion Sounds, and that will be rather computationally expensive!

Anyone?

Thanks!

posted 04 July 2006 21:11            

Maybe you could use a timeline and jump back and forth between the different instances of the sounds with midi program changes. This would allow you to turn off the sounds that you're not using, and you could set up the timeline to deal with the crossfading between sounds. You could also try using a compiled sound grid instead.

posted 04 July 2006 21:20            

I haven't used the timeline much so far, but I think the moment has arrived for me to get into it. And I've never used a compiled sound grid, just checking it out in the manual, looks very interesting

posted 05 July 2006 14:25            

SchedulingSounds.kym

In the attached sound, I have taken a timeline and saved it as a sound, and then made it into a class. All the class does is turn the input sound on and off, with !KeyDown. (You have to wait a couple of seconds between keydowns because of the way the timeline was setup). So I have made the input grains going into a verb (trying to get some expensive CPU going, but you could put anything into it). Watch your CPU as you turn it on and off...

It only works a couple of times on and off, because the timeline I had was only a couple of instances of the input sound, separated by WaitUntils (waiting for !Keydowns). But since it is possible to use program changes to jump within a timeline, it should be possible to make a sound that uses program changes to turn the input on and off. Only I can't figure out how to do it. The program change/marker capability seems to be built into the timeline structure and doesn't translate when trying to save a timeline as a sound.

Incidentally, saving the timeline as a sound is a trick that SSC told me about a couple of weeks ago - after you've compiled a timeline, it is on your clipboard and you merely have to paste it into a sound file.

SSC - how would we be able to use program changes (or a !Gate) to loop between the input on and off?

Phil

[P.S. Due to some sloppiness in creating the sound class, I inadvertantly have euverbs contained in the class in addition to the one going into the input - anyway, it adds to the cpu which helps make the point...]

posted 06 July 2006 08:53            

Why not just use the Timeline? That way you can change the speaker configuration, edit the parameters and structure of the Sound more easily, get visual feedback on the position of the time cursor, jump to arbitrary points in time (by clicking or sending program changes), draw panning functions for surround setups, etc.

posted 06 July 2006 11:21            

quote:

---

Originally posted by SSC:
Why not just use the Timeline? That way you can change the speaker configuration, edit the parameters and structure of the Sound more easily, get visual feedback on the position of the time cursor, jump to arbitrary points in time (by clicking or sending program changes), draw panning functions for surround setups, etc.

---

The timeline definitely has its advantages, but at times it's also overkill for what I need. It's great if you want to do sequencing-type things or draw in or generate and edit parameter automation - but sometimes you just want to do something like swap out a couple of effects, and to put a sound in a timeline would mean that you either put the sound in twice with variations (and any changes you make to the common parts you then have to do twice), or you put the effects on a separate track with sends to the track, and then you get into changing the structure of your sounds to accomodate this.

I was just hoping to be able to create a class that would turn off and on (in terms of CPU usage) whatever is plugged into it. Since this is possible with the timeline, I assume that it can't be an impossibility with a normal sound.

Going back to the original question on this thread, such a sound would make it possible to turn off one branch of computationally intensive objects while turning another on - thus allowing the two different wirings of objects that Jack has.

Incidentally - this wouldn't do what Jack is asking becasue it wouldn't realtime - but wouldn't it be possible to create a script that wires up whatever is plugged into it either serially or in a parallel structure, depending on user input before compiling?

Phil