voice select ..zipper noise

i am having a bit of a problem here
I have set my reaktor ensemble to 50 voices , the id nr output .from the voice info module is multiplied by a factor of 100 ...which gives me 100 ..200..300...5'000 for each voice , then these values get converted to log.values for the p input of a bandpass filter
A monophonic saw wave goes in to the bandpass ,resonance set 0.98
But here's the problem , when I place the voice select module after the filter , and assign a slider to scroll/move trough the different voices ( thus each voice having a different frequency ) ...there is a horrible zipper noise ..even after applying an audio smoother and setting event rate to max, most noticeable when resonance is set high .
I don't know if this is the nature of things , since the voice select mo

here the ens.

https://www.box.com/s/70mxhdisalunm7fsbjt6

 

a few thoughts -

get rid of the voice combiner after the from voice module, it's not necessary and it's making things really really loud.

the smoother does nothing because the from voice module takes only integer inputs to the V input (2.5 is not a voice, and it does not fade between them, afaik)

why not just have a knob that controls the pitch of the filter?

 

Sure i could juste modulate the filter freq
This is just an exercise for using Theo voice module,IT is awesome for creating partials form One single sine osc,by just multiplying the voice output with the fundamental freq into the input of osc.
---
Why should i get rid ofwel the voice combine module?
It really lowers the volumes because it ís multiplied by 1/x (x=nr of voices) so it really lowers the overall output

 

when i wrote that i didn't see the 1/x module.

regardless, taking a mono signal (the output of the From Voice module) and running it thru a poly voice combiner takes that signal and multiplies it by the number of voices, in this case 50.

then multiplying it by 1/50 undoes that, but still, both modules are unnecessary.

anyway, to get the result you're looking for i guess you'd have two from voice modules from adjacent voices and the voice select knob could use a crossfader to fade between them or something like that.

 

I think that would work nicely. After smoothing the control, feed it into a div/mod module. The div and div+1 could select the two voices and the mod could crossfade between them. Exactly like crossfading between a series of adjacent waves in a wavetable.

Remember to set the crossfader to have linear action.