More Arduino Arpeggios

Another project in Propellerhead Software’s Reason using an Arduino Uno micro-controller mated to a Spark Electronics MIDI-Shield daughter board.

Download audio  (06:15, 14 MB). The bit-rate was left high, at 320 kbps. This makes for a pretty big file, but it also helps avoid shrill compression artifacts, especially in the virtual women’s voices. CAUTION: If you have your EQ set to boost treble, you may think you’ve stumbled into a bunch of shrieking Harpies when the ladies sound off.

As in a previous post, I’ve sent chord data via MIDI to the micro-controller,  which is running Brendan Clarke’s sequencer program. The Arduino then returns the notes in the form of an arpeggio, which I then map onto a Reason “Combinator” utility. The latter encapsulates four separate instruments, playing in unison. To the arpeggios I add virtual strings and a bass line, plus a virtual SATB mixed chorus.

The MIDI-Shield and micro-controller combined have provisions for adjusting gate length (allowing for short, precise notes or a long and rather muddy slurred effect), for tempo, for any one of eight pre-defined note patterns, and for sequence length. In this case, I simply chose the first eight-note sequence, at a brisk tempo, and with minimal gate duration. Thereafter, I made no real-time adjustments, although I could have.

This rather simplistic use of the MIDI setup is not that much different from Reason’s built-in “Matrix” pattern generator, with this exception: I can supply arbitrary chords at will in real time and without limit. No need to program every sequence I could conceive of in one of the “Matrix” pattern banks. And transposition is as simple as supplying a fresh chord to the MIDI interface.

OTOH, thinning out a dense chord while sustaining some of its notes is a little tricky. Because of the sequencer’s design, it adds/replaces notes in a FIFO fashion. This means you have to thin the chords by repeating notes you wish to keep, thus shifting out of the queue the notes you want to silence. Lots of trial and error.

I’ve learned that doing the final mix in Reason doesn’t work too well for me. Better (for me, at least) to export individual voices and apply EQ and other fixes in Adobe Audition, then produce a final mixdown. This may be due to my fondness for densely stacked harmonies. I have trouble getting a final mix in Reason that isn’t muddy and listless. Truth be told, I don’t yet have a firm grip on compression and other master effects that might make things sound crisper and brighter. So, “it’s not you, it’s me.”