sure, no problem!
i think i'd start by first fiddling around to find a good backing, based on the way i'd be feeling that day.
the backing could either be a drone with some harmonics or a sequence. then i'd add effects to that. then some bass, usually 2 bass sounds. one left, one right. if it's only one bass sound, that one goes in the center.
i use a yamaha qy70 for sequencing, syncing with the korg multitracker.
while the drone or sequence is playing, i sing something that fits the feeling. i then play the melody i was just singing. i sing bass, melody and effects voices.
the melody is usually based on traditional modes, or something i made up.
then, if applicable come the effects voices, which i prepare by voicing or singing them.
the latter is fun, like making zap or wind noises. the idea behind the voicing is that you get an idea of what sounds good and feels attached to the tracks already present. sometimes it takes a bunch of takes to fine-tune it. that takes a lot of practice! be warned...
then i'd record that voice with the moog. my singing is just kind of 'placeholder'. i never record my real singing - the technique works great because singing means that you feel very honestly what you sing. am i making sense? when singing, i memorize the melodies until i'm ready to record, and the moog is properly warmed up.
then i add some effects, usually a bit of reverb or echo.
sometimes i will jam more or less aimlessly on my kantele or the moog. (you can listen to a short experimental jam under "work in progress" on my site).
after the aimlessness, i usually come up with something usable, and then i play it on a multitracker, the very reliable and user-friendly korg d3200, which, by the way, has wooden sides!
in the process, i like to leave some space for surprises (and i'm lazy), so that means i don't compose every note. a lot of it is 'planned improvisation', meaning that you have a melody (or a mode) in mind, but can meander and variate, kind of like a comedian will adjust a routine to an audience. similar to that, any melody can either bomb, hit spot-on, or even go way better than planned!
at the end i transfer my files into audacity, trim here, cut there and normalize, like a barber. some fade-ins or outs and silencing of intros or silent passages etc. takes place. that's it!
the whole thing can take between an hour and four, after which i stop and do something else.
martin
