[sdiy] Release Velocity & Poly Aftertouch -- Too Much?

[Rainer Buchty]

>I can understand why many MIDI keyboard makers didn't implement poly
>aftertouch. It could add to the complexity and cost of a keyboard
>considerably.

I object... If you look at the Ensoniq patented Polypressure keyboards, 
those things are fairly simple:

- *no* contacts at all, so no corrosion
- instead each key contains a piece of metal which together with an 
etched-on coil on the PCB forms an oscillator, and
- this oscillator frequency is counted and converted

So what you effectively need is a bunch of transistors, a little bit of
selection logic, the counter circuitry, and a controller to get the
entire keyboard scanning and interpretation done.

The keyboard controller you always need, so the extra costs are mainly 
transistors, logic and the counter circuitry -- on the cost reduction 
side, such keyboards should be easier to manufacture cause you don't 
need to fiddle with contact springs or mats.

So in the end the overall additional cost is maybe in the range of $5
per keyboard -- multiply it by 10 and the customer will happily pay that
for increased usability.

>Does it really somehow add to the cost of a keyboard? (I wouldn't 
>expect it to be a big factor.)

Someone had to write that software instead of copying the same old 
routines from Hal Chamberlin's MoM :)

>Most importantly:   I want to know if it's simply expecting too much to 
>have a MIDI keyboard with both poly aftertouch AND release velocity. Is 
>that amount of data simply too much for a MIDI stream to handle?

Release velocity comes for free, because it's already present in the
note-off command. Poly-AT indeed can be a bitch; if I record stuff with
Channel-AT on I often see numerous AT messages just from accidental
"after-pressing" -- now multiply this with the number of notes played...

Rainer