Well, as the title says. This would enable smoother transitions between samples.
You can also accomplish this by using the modulation chains for each sample, but i suspect you probably want a different attack and fade-out value for the first triggered and last triggered sample in the chain… (this would require some kind of timer idea:process chain x (fast attack rate) if timer is not yet set, process chain y if triggered within 500msecs (crossfade), else process chain z (other decay / fade out transition))
A nice article explaining the idea in more detail: http://www.soundonsound.com/sos/jan06/articles/lostscience.htm
You can come a long way with creative use of modulation devices. For instance, I have created instruments that had a “quiet” velocity layer, gradually fading out as the “loud” layer took over.
This was achieved by having the quiet layer ignore the velocity as input from notes (disabling “VEL>VOL” in the keyzone), and then implementing my own method for tracking the velocity consisting of two velocity-tracking devices: one for (slightly) fading the volume in “from the bottom” and another one for (strongly) fading the volume out “at the top”.
The second layer - the loud one - was using normal VEL>VOL tracking, but with an additional velocity tracker for making the volume fade in only when above a certain threshold (to make the transition more abrupt).
The same principle can be applied not only to velocity but also pitch, but it does have it’s limitations: complex setups would be a pain to work with. A simple setup like the one I described is done in a couple of minutes, though.
@fladd: I read the article you linked to, and it warns about some of the pitfalls related to this technique: if you are not careful, you might introduce some very serious problems with phase, or muddy the sound. So this technique would best be applied to real, acoustic recordings or samples with complex timbres (avoid stuff like sine waves unless you really want trouble).
Yes, this functionality would indeed be used for sampling acoustic instruments. Your example is very nice, but also shows that a more direct way is needed for easier setting up.
++++++++++++++++++++++++ VERY WELCOME FEATURE 
/>/>/>) want this too…
I think this will be possible as soon as you can layer sample aliasses (layering the same sample multiple times). The ramping however has not much use because keyzones are attached to the key and its velocity.
But if sample aliasses are possible, it will most likely also be fairly easy to write a script to automate cross-section layering.
sample aliases are welcome, but with cross fading you can achieve not only smooth transition from one sample to another but even if you cross fade two samples on one note you can achieve velocity on mixed samples. less you press the midi keyboard key you can hear one sample more you press you can hear another… etc…
Yeah but that will look crazy stupid, like in the crossfade zone you’ll have two zones every key. And the amount/distance of crossfading won’t be adjustable at all.
So a good feature request, probably more for 3.1 or so, is 1. velocity-domain layer crossfading and 2. key-domain layer crossfading.