Files:
Introduction
It takes special care to get human-sounding intrument work from any sequencer. Everything about sequencers is inherently mechanical, whereas there is a certain "wishy-washiness" to human timing. It's not simply randomness or "human error" in timing. "Shuffle" or "Swing," the practice of shifting every odd 8th note forward, slightly captures the wishy-washy feeling a little bit, and can be important to creating some types of groove, but is only a start.
To get a real sense of the problem, imagine a human shaking a tamborine. You get a sharp downstroke, creating the primary rythm. But then you have a "lazy" upstroke, preparing for the next downstroke. What will the upstroke sound like? It will be "smeared," softer. If it is a typical 4/4 rythm, you'll also have an emphasis on the second and fourth 1/4 notes. So basically, you get:
Tshhh shh Tshhh shh TSHHHH shh Tshhh shh Tshhh shh Tshhh shh TSHHHH shh Tshhh shh
Don't worry if you can't make sense of that, just listen to the result mp3 linked to at the top of this article.
The Upstroke
So how to emulate the upstroke, working from a downstroke sample? The easiest way--which will still provide a decent result--is to give it give it an attack. If you look closely, you'll see that not only is their a slight time added, but there is also some tension, so the curve of the attack bows inward slightly. That bow, barely visible on the envelope, is important in getting a natural sound.
The second thing to notice--and this is also true of the upstroke--is that there is zero sustain, and a fairly rapid decay. This is important because you need very close control over the duration of the sample. Why? Thinking back to the human actually shaking a tamborine, the beginning of the upstroke will cut off the downstroke.
In the mp3 and zip file provided, I used two samples--one for the downstroke, an envelope modified version for the upstroke, then a sample from the same tamborine, but with different dynamics, for the emphasized downstroke.
Levels and Sustain
Adjusting levels and sustain mixes a bit of randomness with a bit of common sense. Obviously, and human shaking a tamborine isn't going to get the exact same "note velocity" and duration every time. But the pattern can establish different groove characteristics by emphasizing different notes. It helps to listen to actual recorded loops at this point (which can be found via google for just about any instrument.
Tip
Try substituting different samples into the .flp provided (i.e. shakers). The channel settings and sequence work well for other "human instruments" as well.