Trigger finger setup to control EXS

Question

I must be missing something: right out of the box, the M-audio trigger finger faders and knobs only control the selected track's level and pan parameters. How should I set it up to control, say, my EXS filter cutoff frequency or attack level?
I've read the chapters about automation, control surface support, and the EXS midi controller list. Yet I still don't understand how to turn control events into fader events, which are the only events that seem to alter the sampler adjustments in real time. What am I missing?
Or is there a web site from which I could download ready-made presets?

Answer 1

Well, I eventually built a cascade of transformers that convert midi control events into Logic fader events. Each transformer is set up as a track automation splitter to catch a specific CC message and send fader events to an audio instrument instead.
Here, a foot controller event is mapped to the resonance control of the ESP instrument. The unintercepted midi events are passed unchanged to the recording & trough object, i.e. to the selected track. Using the Trigger Finger default assignments, I mapped the ESP's 9 most useful parameters [filter, envelope, drive, chorus] to convenient faders and rotary pots, leaving the rest for volume, pan and modulation.
Logic internal fader events can be used in real time and record directly to track automation. I made this into a macro so it can easily be copied [or imported] from the original song file environment. The macro contains two external objects references: the audio instrument and the recording & trough object. The advantage is twofold:
• To use such a cascade, just substitute the recording & trough object (found on the click & ports layer in the environment) for the macro.
• To control many instruments, simply plug many macros in parallel — you'll need to unpack the macros and change the references to different audio instruments. Oh, and Logic doesn't frown when there are many recording & trough objects: they're all references to a unique internal object.
What I learned that isn't documented:
• All plug-ins — and audio instruments, including GB's instruments, are no exceptions — number their parameters by order of appearance in the controls view of the instrument interface. This numbering is zero-based: the first parameter could theoretically be controlled by a CC#64 event [but see above in this thread] and is actually mapped to fader 0 — for instance, the mono/poly parameter of the EXS sampler is assigned CC#71 [see the plug-in reference, page 571] because it's the 8th parameter, and therefore is internally mapped to fader 7.
• No plug-in uses the same layout for its parameters, so there's no way to build a general controller, but it takes only a few minutes to remap a handful of splitters and adapt them to the most useful parameters.
• Audio instruments only receive fader events from internal channel 2, so the transformer must be set up to fix this whatever the midi channel the data came from.
• Audio instruments also listen to internal channel 1 — the channel strip channel — for standard control events like modulation, pitch-bend, and the like... There's no need to convert these, since they can be recorded along with the rest of the midi data, but it's still possible: just remember they must be assigned to channel 1.
• Splitters can be made to react solely to a given midi channel: using a versatile control surface like the Trigger Finger, one could build a set of 16 presets, each dedicated to controlling a specific plug-in. This would mean live performance with something like 64 faders and 128 pots at your disposal, at the flip of a button... Not counting the 16 x 16 pads, of course.
• Macros are clumsy: they can only process one input to one output, unless you embed external references — contrastingly, Reaktor accepts an arbitrary number of inputs and outputs. I'm disapointed by the lack of documentation: which objects can be embedded, and which can't?