I agree with Florian and Bernard. I believe that most (completely) analog synths work that way, as the release capacitor was still loaded while it 'released', and when the attack part of a new note had to be played, the capacitor would leave off from the value it had for the release. It's probably a bit more complex than that, but on my Yamaha CS 60 it also works that way. It opens up all kinds of possibilities where the release time of the previous note governs the response of the attack of the next. It's a good thing, that behavior! :-)

Cheers,
Rainer

--- In kiwitechnics@yahoogroups.com , Florian Anwander <fanwander@...> wrote:
>
> Hello
>
>
> bernard.escaillas wrote:
> > On the contrary, if you play chords having less than six keys
> > with the "resume VCA level" you propose, you will get unpredictable results
> But that is how the original JX3P works in fact. And it does not disturb
> in none of the three possible original voice algorithms
>
>
> kiwitechnics wrote:
> > I orginally had the envelopes restart from the current level rather
> > than zero but I didn't like the result as notes had variable attack
> > amounts which varied from nothing to the pot setting and sounded odd.
> Did you compare your former version with the original JX3P? I think,
> this is how the original JX3P does work too.
>
> May be I am too grim, but I own two JX3P's - one is original and one is
> kiwi'fied. So I always can compare... (And I have to admit, that I'd
> like to have the envelopes as close as possible to the original version).
>
>
> Florian
>