Make Electronic Sound

Hi. I have a Gen patch where I have either 3 Sawtooth oscillators or a Square wave oscillator. A param comes into 

 to determine which is used. The problem I have is that even though I thought the 

If I disconnect the connection between the 

 does not add to the cpu overhead. I tried using a 

 object to choose between the saws and the square wave but still the square wave adds to the cpu.

Is there a way to 'switch off' the square wave oscillator or am I thinking totally along the wrong lines?

Hi.  I have a Gen patch where I have either 3 Sawtooth oscillators or a Square wave oscillator.  A param comes into in 3 to determine which is used.  The problem I have is that even though I thought the gate object would prevent the squareOsc from using any cpu, it still does.

If I disconnect the connection between the squareOsc and the * object below, the squareOsc does not add to the cpu overhead.  I tried using a switch object to choose between the saws and the square wave but still the square wave adds to the cpu.

turn-off-a-gen-subpatch

but you can't turn it off persay... however, you can zero the signals coming in, or use 'params' for things that don't need updating at sample-rate, in some cases...(or use 'fixdenorm' and 'fixnan' for other similar efficiency purposes)...

...hoping others will explain...
but you can't turn it off persay... however, you can zero the signals coming in, or use 'params' for things that don't need updating at sample-rate, in some cases...(or use 'fixdenorm' and 'fixnan' for other similar efficiency purposes)...


in gen~ patching there is simply no way to do this.

if an object / subpatch whatever is not connected to anything at all in the signal chain, the compiler will simply ignore it and it will not incur any cpu. but as long as something is connected in the chain, it will be compiled and processed no matter what gates / selectors / etc you use. (this is unlike latching in reaktor-core, for example).

this is a big feature request for gen~ patching and cycling 74 know about it.

however, all is not lost. to control cpu usage in gen~ you need to use GenExpr language in a codebox. in an if/else block, cpu usage is on or off depending on if the function / code is 'called' or not. (there are caveats if you are using data / buffer, but that is a longer story).

...you get the idea. the cool thing is, "somethingveryexpensive" could be a '.gendsp' abstraction in the search path, but i usually just use it with a declared function at the top of the gen codebox file.

there is a (very simple) example of such an if/else block in the filters i posted as a zip file in this thread:

https://cycling74.com/forums/collaborate-on-building-reaktor-filters-in-gen/

if/else in GenExpr really works. i have made very complex multiple nested if/else blocking and got away with it very well (

https://www.ableton.com/packs/dub-machines/

GenExpr is remarkably easy. it is literally like typing gen~ patches. syntax very forgiving. really worth learning.

myoutput = 0; // initialise
if (mydecision == 1) {
myoutput = somethingveryexpensive;
} else {
myoutput = somethingverycheap;
}

there is a (very simple) example of such an if/else block in the filters i posted as a zip file in this thread:
https://cycling74.com/forums/collaborate-on-building-reaktor-filters-in-gen/

if/else in GenExpr really works. i have made very complex multiple nested if/else blocking and got away with it very well (https://www.ableton.com/packs/dub-machines/ for your reference).

That's interesting, thank you. Is it only possible to block cpu usage for code inside a code box? I do prefer drawing networks in gen~. I suppose I could copy the code and paste it into a code box, but it does not seem the best way to manage a design.

It should come as no surprise to you when I tell you that it worked! Using a codebox with a simple if else solved my cpu woes.

It should come as no surprise to you when I tell you that it worked!  Using a codebox with a simple if else solved my cpu woes.

(i think 'hth' means 'hope that helps'... and it does for me too, thanks!)

'Turn off' a Gen subpatch?