Doepfer

Hello Florian, 


your focus is magistral, again: thanks ! 


Just few things more. You wrote: 









As written before: our hearing expects exponential changes signal 
levels. So you always need one exponential part in the game. 



And this is assumed now in my comprehension. 

<blockquote>





As you have both, linear voltages (velocity or LFOs) and exponential 
voltages (envelopes), you need both types of VCAs in your system. But I 
tend to have more linear VCAs than exponential ones. 
</blockquote>

Here I have two questions. The first is more about definitions. For each note played by the midi keyboard, the midi interface will provide the system with a fixed voltage (offset) and so the “lin/exp” question is not really meaningful here: there's no evolving data to be sent, just a static value for each note which is neither linear neither exponential. As for the LFO, on the contrary, it is continuously changing and, as far as I can see, it derives more from a (cycled) envelope, which is usually exponential (unless probably the LFO is shaped on a triangle wave). 


The second question is definitely “practical”. 
How better would I combine the parts, this way? 


linear voltage --> exp VCA 

exp voltage --> linear VCA 


Kind regards, 
Diego 




Il giorno 28/nov/2016, alle ore 11:27, Florian Anwander fanwander@mnet-online.de [Doepfer_a100] < Doepfer_a100@yahoogroups.com > ha scritto: 

<blockquote>





Hello Diego 

Am 28.11.2016 um 00:36 schrieb diegora@free.fr [Doepfer_a100]: 
> but the core 
> phrase you wrote missed the essential “verb” 
You are right. Sorry for that 

> «if you mix the velocity voltage with the envelope voltage and feed 
> the sum into one VCA, it will cause the VCA all the time a little bit 
> (according to the velocity voltage) and the evnelope will add on top 
> of this basic level.» 

This should read "...it will cause the VCA to let the signal through all 
the time a little bit...". 

The important thing is, to remember that MIDI-control and 
voltage-control work different: 

The MIDI-system is event orientated. This means: the event Note-On tells 
the note value and the velocity value. It is not the job of the 
MIDI-Keyboard to keep up this information. The receiver has to compute 
and remember the status until the next event is received. 

The CV/Gate-system is status orientated. This means: in a CV-system the 
sender of the information (let it be the keyboard or the MIDI-interface) 
has to deliver continuosly the information about the actual status. The 
receiver does not have to care about it, he can rely that the sender 
provides the status all the time. 

Now back to the velocity topic. 

Lets take an example: 

We switch on the complete system. All voltages, let it be pitch / 
velocity / gate, are 0 Volt. 

From the MIDI-keyboard an noteOn event comes to the MIDI interface. 
This event says: "Note D#1 has been pressed with 50% velocity". 

The MIDI-Interface will translate this to 
* Pitch-voltage = 1.32 Volt (I'm not sure 1.32 is the right value for 
D#1, but it does not matter in this case) 
* Velocity-voltage = 2.5 Volt (assuming the velocity voltages cover 0 to 
5 Volts) 
* Gate-voltage = 5V (assuming gate is 0V/+5V) 

The next event will say: "Note D#1 has been released". 

Now the MIDI-Interface will change only the gate-voltage back to 0 
Volts, but the pitch and velocity voltage will remain: 
* Pitch-voltage = 1,32 Volt 
* Velocity-voltage = 2,5 Volt 
* Gate-voltage = 0V 

Now let us assume, we have used an organ-like envelope at the ADSR (this 
makes it easier to calculate the voltages, because we can simply use the 
gate-voltage as envelope). If we add the envelope-voltage and the 
velocity-voltage, then we would get a voltage of 7.5 Volt at the noteOn 
(2.5V velocity + 5V envelope). At the noteOff event we would get a 
voltage of 2.5 Volt (2.5V velocity + 0V envelope). 

If you feed this to the audio-VCAs control input, then the VCA still 
would send some signal through continuously after the noteOff. 

> Anyway, now I see I always need to adopt 2 VCAs, probably (I try to 
> guess again) the better quality VCA will preferably manage the audio 
> signal (and here I can choose to fire the velocity OR the EG). 
Right. 

> Another subject which probably crosses this application is the 
> difference between linear and exp VCA. Sorry, I won't you to go back 
> to the lin/exp dilemma of the EG segments :-) but if EGs segments are 
> always exp (because it's easier to design), then probably wouldn't be 
> that wise to have an amp which features exp control scales, it might 
> be redundant. It seems that linear VCAs finally are more 
> versatile... 
As written before: our hearing expects exponential changes signal 
levels. So you always need one exponential part in the game. 
As you have both, linear voltages (velocity or LFOs) and exponential 
voltages (envelopes), you need both types of VCAs in your system. But I 
tend to have more linear VCAs than exponential ones. 

> As for the AC-coupled / DC-coupled thing, I really understood very 
> little when I tried to go wiki. 
Basically everything works fine with DC-coupled VCAs, but one might get 
some "whumppp"-sounds at immediate noteOns, if you use a DC-coupled VCA 
with an audiosignal that contains a DC-offset (example: unsymmetric 
pulse wave). 

> One last thing: why do we found some VCA's who offer 2 CV Ins if they 
> cannot be used to process EG + Velocity? Do the 2 inputs affect the 
> amp in the same way or they are wired to different stages? 
Example: if you want to control the overall level of a voice with 
velocity and with an slight amplitude modulation, then you may do that 
with the mix of two voltages on one VCA. 

> I was interested in this quad VCA, what do you think about it? 
The http://www.doepfer.de/a1324.htm is the exponential counterpart to 
the http://www.doepfer.de/a132.htm (and of course the 132-4 is fourfold, 
while the 132(-1) is twofold). 

Florian 

 
</blockquote>