Constant-slope bounded sweep ugen

Oohhh... the normal ugen set is making this really difficult for me.

I am trying to measure the weight of a pitch class relative to other
pitch classes (which are coming in via pitch-following -- so far I'm
getting wholly usable results there with Qitch and Tartini both). The
idea is that a valid pitch will "add energy" to its pitch class bin,
with a certain fixed slope. If the pitch is recognized only for a
split-second, it would not add enough energy to trigger a change in
behavior in the client. A note that is held longer or repeated many
times would add more energy and eventually cross a threshold.
Something similar would happen with brief interruptions in the pitch
recognition -- a tiny glitch would not pull the value below the
threshold, but if the note stops for some time, the pitch class would
be deactivated.

I was thinking of a linear response, but the problem is that this
calculation depends on the slope and basically every linear UGen we
have is time driven, not slope driven. I could use Ramp but if the
trigger input changes before the ramp reaches the final value, I would
have to adjust the time somehow and that is proving to be really,
really hard.

Note that Sweep.kr(...).clip(...) will NOT work for this because
changing the rate would just bring the value down from some value far,
far above the desired maximum, so notes would take an indefinite
amount of time to deactivate.

If I didn't have to leave my Mac behind for this stupid business trip
I might adapt one of the existing UGens in c++, but likely I won't
have time for that and I wouldn't have any way to compile it in
Windows -- leaving me stuck for whole week without being able to work
on this logic.

While thinking out loud here, it occurred to me I might use Decay2
with a Schmidt trigger (the latter of which I didn't discover until
just today :).

Any other ideas?
hjh


--
James Harkins /// dewdrop world
jamshark70@...
http://www.dewdrop-world.net

"Come said the Muse,
Sing me a song no poet has yet chanted,
Sing me the universal."  -- Whitman
_______________________________________________
sc-users mailing list
sc-users@...
http://lists.create.ucsb.edu/mailman/listinfo/sc-users