Hello Ricardo
> You mean the voices mixed into the S&H and new
code can't change that?
Its like that:
The CPU provides all parameter values one after the other parameter on a
so called data bus (this is called multiplexed transmission). On the so
called address bus at the same time, the information is provided for
WHICH of the 32 parameters the actual value is meant for (Cutoff,
Resonance, Waveform,...).
The digital data get converted into analog voltages in the D/A converter
- but still all values(=voltages) are on the same line.
Now a so called demultiplexer comes and distributes this single line on
32 separate lines according to the destination info from the address bus
(you may imagine this as a rotary switch). But this separate lines get
the voltage only for a nano second. This would not help to control the
synth. So the mentioned Sample&Hold stages pick this short signal and
store it until the next short signal appears. The output of these S&H
stages is now sent to various voltage receivers (Cutoff to Filter,
Resonance to FeedbackVCA, Waveform to DCO,...).
So if you have one sound with 32 parameters, then you need 32 S&H
stages. That is, what the JX3P can provide.
If you want to have two different sounds in the same synth at the same
time, you need for each sound separate 32 S&H-stages. In total 64 S&H
stages. But these do not exist. So you cannot convert the JX3P into a
multi timbral synth.
Florian