Peiman

I have another question about the looper you posted the other day 

(sorry!!). Basically I was wondering why the pitch slides up and down 

while the loop lengths are being changed and if it's possible to 

I have another question about the looper you posted the other day  
(sorry!!). Basically I was wondering why the pitch slides up and down  
while the loop lengths are being changed and if it's possible to  
eliminate this somehow?

loop-points-no-click

On 11 Jan 2007, at 19:35, peiman khosravi wrote:

> I have another question about the looper you posted the other day 

> (sorry!!). Basically I was wondering why the pitch slides up and 

> down while the loop lengths are being changed and if it's possible 

the loop length and position are only sampled at the loop-points 

the phasor frequency depends on the loop length, but it isn't sampled 

only at the loop points but is sent directly to phasor~.

this causes the little glissandi when changing the loop length.

you have to sample the new freqency values only when phasor is 

jumping back to the loop beginning - and this requires a feedback 

loop, which is not easily possible when working with signal objects.

you can use send~ and receive~ objects to fake it.

it's not 100% accurate, cause it depends on the signal vector size, 

actually it would be best to have a dedicated external to do this, a 

slight modification to phasor~ that only changes its frequency at the 

beginning of a new ramp. i think i've done something like that when 

first experimenting with writing externals.

> Hello Volker,
>
> I have another question about the looper you posted the other day  
> (sorry!!). Basically I was wondering why the pitch slides up and  
> down while the loop lengths are being changed and if it's possible  
> to eliminate this somehow?

the loop length and position are only sampled at the loop-points  
(sah~ controlled by phasor~)
the phasor frequency depends on the loop length, but it isn't sampled  
only at the loop points but is sent directly to phasor~.
this causes the little glissandi when changing the loop length.

you have to sample the new freqency values only when phasor is  
jumping back to the loop beginning - and this requires a feedback  
loop, which is not easily possible when working with signal objects.
you can use send~ and receive~ objects to fake it.
it's not 100% accurate, cause it depends on the signal vector size,  
but it works quite ok in this context.
see below.
actually it would be best to have a dedicated external to do this, a  
slight modification to phasor~ that only changes its frequency at the  
beginning of a new ramp. i think i've done something like that when  
first experimenting with writing externals.
let's see if it's still around...
volker.

loop points no click