ddunn - thanks!
I received some offline questions about how I approach doing fast string runs and rips. I thought it would be nice to share on the forum directly.
Some examples of runs I've made with this approach, most of these using the VSL Special Editions:
So here's my thought process - it starts by considering the psychology a performer playing a fast run. The notes will be faster than a performer can track rhythmically, so they will be estimating how fast to play the notes to start and end at the right times - I call them "anchor notes". If the run is long enough, there will also be some anchor notes in the middle of the run too, usually on a beat. A performer just estimates the speed of notes between these anchors - specifically I think performers will usually play those between notes slightly faster than necessary, until they're about to arrive at the next anchor note. This gives them room for error to make sure they will can arrive to the next anchor note correctly. Then, the one or two notes before an anchor will be slightly longer to compensate for going slightly faster on previous notes, and finally the next anchor note lands closer to the desired beat - maybe not entirely accurately, but close enough. The anchor also might get held just slightly longer, or slightly accented, or both, before continuing to the next string of fast notes.
Example - same youtube video I linked earlier in this thread, at 0:26, the A-major scale is two octaves. I put one octave roughly per beat, and used the A note every octave as the "anchor notes". If you listen carefully, you'll notice a slight lingering just around completing the first octave of the run, right in the middle. It also has an accent - though I think I may have overdone that accent a little =)
For what it's worth, this is what I think I've observed about myself when playing string runs, and I've also seen similar results in piano pieces, when I've recorded myself playing the keyboard.
OK so, putting this analysis to use, here is what my process usually looks like to fine tune a strings run:
- identify what the anchor notes would be, place those coarsely where I want when recording/entering the MIDI
- tweak the in-between MIDI note positions, lengths, overlaps, and velocity CC, to reflect all these principles mentioned above.
- Layering instruments or microphone positions can also help, because it gives you the opportunity to shift MIDI note positions differently for each instrument, which is a much more natural blur effect than just overlapping MIDI notes.
- experiment with different articulations and velocities on the anchor notes, and pick an articulation that captures the style I want. Especially the beginning and end notes - staccato, fortepiano, sfz, marcato - all have slightly different types of attack which can greatly change the character of a run. The attack can usually be further sculpted by fast velocity xfades, too.
- experiment with blurring the run, between *and including* anchor notes. I usually try to make the beginning of the run slightly less blurry than later into the run.
Finally, for very dedicated, obsessed MIDI tweakers, it also helps to think about three different legato transition types on string instruments: (a) performer only needs to move a finger, (b) performer needs to shift the position of their hand along the fingerboard, and (c) player needs to move their bow to a different string. When a performer only needs to change fingers, they can be fast and less blurry. When a performer needs to shift their hand, they can have a slower but more messy transition. When a performer needs to move their bow to a different string, the transition may not be as messy, but can still be slower and inaccurate in timing. I usually don't go this far, but actually with Synchron Strings I found it useful to think about, in order to decide when to use fast legato versus slurred/fast portamento legato.
Hope at least one person finds this useful. I enjoyed the challenge of trying to put this in words, anyway. Cheers!