Bimetallic March2024/02/18 MSK 015 MSK 013 MSK 012 MSK 009 MSK 007 | MP3 FLAC | |||
Share on: Facebook Twitter Reddit LinkedIn | ||||
This is meant to sound something like a brass band combined with steel drums, also known as steel pans, hence "bimetallic." Except for the non-steel drum track (generated by a software synth with free samples) it's all done with three modular patches: a basic subtractive patch featuring the Leapfrog VCF, two voices of that in unison (from the two oscillators in the Middle Path VCO) mixed together for a "fatter" sound; much the same for the Coiler VCF; and the special steel drum patch. Steel drums have an unusual spectral profile because they're only partially tuned. A big part of the sound is a fairly pure damped sine wave, but then there's a wash of higher-frequency harmonic and inharmonic stuff that comes in later in the note, a little bit like the sound of a cymbal (which is physically similar) but less inharmonic while still being somewhat inharmonic. Synthesizing that is an interesting challenge. I started with the Middle Path VCO in "firm sync" mode, and passed the triangle output of the slave oscillator through a VCA (one channel of the MSK 015 Quad VCA) before returning it to the Middle Path shaper. Some of the master oscillator's triangle output also goes into the shaper through the normalled connection. There are a total of four separate ADSR envelopes (MSK 012 Transistor ADSR) in my patch, to control amplitude of the triangle before the shaper; FM on the slave oscillator; amplitude of a second VCA channel downstream of the shaper; and cutoff frequency of an MSK 009 Coiler VCF on the output. But many of these envelopes are actually quite similar to each other and someone with fewer envelope modules could probably get away with using fewer and just multing them out to the different destinations. With the inharmonic stuff from the slave side coming in relatively late in the note, as the initial pure harmonic transient is fading, it does end up sounding a lot like a steel drum. The timbre changes a fair bit across different pitches - this is not a patch that tracks consistently - but that too is authentic to real steel drums, which sound a lot different in different pitch registers. After generating MIDI for each instrumental part I recorded a bunch of separate tracks and mixed them down. This came to 17 tracks: six with the Leapfrog subtractive patch (one "high," two "bright," two "dark," and one "low" - these refer to different settings of the MSK 008 Octave Switch and the filter knob settings); six more with the Coiler subtractive patch (same breakdown); four with the steel drum patch (I composed it around the idea of one person playing a "spiderweb lead" pan and one playing a "double second" pair of pans, each player playing up to two notes at a time); and then one track of MIDI software drums. I don't have much experience of orchestrating many-part music like this, so as well as exploring the modular patching this is a compositional exercise too. | ||||