leemorgan

Hi everyone - I keep using what I suppose are glorified LFOs to provide moving variables in my patches, and have done for years - but they always seem, visually at least, to be kind of jerky. Is there a most efficient way to do them (one example enclosed)?

I try a different method every time - using counter for example, or ~peakamp - but I never know which is most CPU friendly / smooth / reliable. Or is is just the visual feedback that's a little jumpy?

number

scale

flonum

number~

cycle~

dial

slider

comment

Hi everyone - I keep using what I suppose are glorified LFOs to provide moving variables in my patches, and have done for years - but they always seem, visually at least, to be kind of jerky.  Is there a most efficient way to do them (one example enclosed)?

 I try a different method every time - using counter for example, or ~peakamp  - but I never know which is most CPU friendly / smooth / reliable. Or is is just the visual feedback that's a little jumpy? 

Hi everyone - I keep using what I suppose are glorified LFOs to provide moving variables in my patches, and have done for years - but they a

most-efficient-wave-lco-patch

next time paste compressed patch, it is no fun to scroll arround to collect all your code.

Mixing signal and non signal objects is not optimal in any case.

If you need oscillator to drive a signal,

then do it all using signal objects, and then display it on GUI using

defered representation if you really need to have visual animation.

If you need non-signal oscillator, than don't use signal objects, or make better

It is also better to check what range gets produced, in your patch 

image.png

number– has at best 20 ms report time, is that what your smooth ramp resolution is ?

Mixing signal and non signal objects is not optimal in any case.
If you need oscillator to drive a signal,
then do it all using signal objects, and then display it on GUI using
defered representation if you really need to have visual animation.
If you need non-signal oscillator, than don't use signal objects, or make better
signal to non signal conversion.
It is also better to check what range gets produced, in your patch 
you drive slider out of it's range


number–  has at best 20 ms report time, is that what your smooth ramp resolution is ?


Thanks - and apologies -I have edited my post.

The oscillator is needed to drive a variable integer. So would I be better avoiding signals altogether/using counter? I guess I always heard that signals for timing are always better...should never the two meet?

Thanks - and apologies -I  have edited my post.

The oscillator is needed to drive  a variable integer. So would I be better avoiding signals altogether/using counter? I guess I always heard that signals for timing are always better...should never the two meet? 

It is ok to use cycle~ or phasor and to drive integer ramps,

but it is better to keep it 0. - 1. float and scale to integers at end.

it depends on resolution of integers and what you drive with them

It is ok to use cycle~  or phasor and to drive integer ramps,
but it is better to keep it 0. - 1. float and scale to integers at end.
it depends on resolution of integers and what you drive with them
Using change at scale output is a must.


at a smaller vectorsize and/or such an audio LFO will be about as time exact as a data rate LFO, but the values will be less precise because number~ or snapshot~ do not interpolate values or predict the future.

i also do such things in float until the bitter end, and if you are looking to replace phasor~, that can be done using metro and line, though in this case you still have that problem of low value precision.


at a smaller vectorsize and/or such an audio LFO will be about as time exact as a data rate LFO, but the values will be less precise because number~ or snapshot~ do not interpolate values or predict the future.

i also do such things in float until the bitter end, and if you are looking to replace phasor~, that can be done using metro and line, though in this case you still have that problem of low value precision.


Most efficient Wave / LCO patch