Odd time signatures! Chromatic bass lines! It's like I'm reliving 1989!
I'm not very satisfied with this; too many ideas, and none of them developed. It's just a sketchbook, so the fact that there's no transition from the A to the B part--and it really does sound lame, due to the rhthymic lurch discussed below--doesn't bother me; that's easily enough fixed. (See sketch #8 before and after here for an example.) But even if I did put in a transition, I'm not sure the parts really belong together at all.
Anyway, I started this recording on bass, as I often do for sketchbooks, since I tend to get riff ideas on bass more easily than on guitar (where I just end up playing chords or noodling). And I got this idea to do this descending chromatic line from a D, while droning the open D at the same time; and not an octave apart, but actually letting things get nasty and muddy in the way that, you know, you don't play minor thirds down low on a bass, much less flat seconds.
So I toyed with the riff, playing around with the rhythm, and for a bit I was playing it in some odd time signature, probably five, but I don't know, because as I tried to understand what rhythm I was improvising, it changed out from under me and became 4/4 again. So ok, I found a rhythm I liked, and then I put down the bass and went into another room for a minute. When I came back, and picked it up, I couldn't remember the rhythm I'd just had and liked, but the rhythm on the recording immediately popped out.
I don't know if it's unique to rock music--it seems like maybe it is--this extreme syncopation you do where you change chords before the measure changes, an eighth note early. It's a cool effect, and ubiquitous in rock music. Back in 1989 when I was doing a bunch of recordings in 5/8 I tried to construct one that did that change-chords-an-eighth-early, but in 5/8 it was too subtle, it became 4/8 = 2/4, which didn't sound at all interesting. Chalked it up to experience and moved on.
And gosh darn it if the thing that just spontaneously popped into my head wasn't in 5/4 with the chord changes an eigth ahead of the measure change. Instead of the classic 3+3+2+2, it's 2+2+2+3+1 (I count it as 3+2+2+3 with the first 3 ahead of the beat... actually, if I count it, I count it as '1... 2 1 2 1 2 1 2 3 1... 2 1 2 1 2 3'). So the drum at the beginning is to try to establish where the beat really is, since otherwise you might parse it as 3+2+2+3, which is, admittedly, a fine 5/4 and probably my favorite 5/4 nobody uses, but not what this one is supposed to sound like. (Ask me some time about how I misparsed a Janet Jackson song (I think) into a bizarre freakazoid 6/4 when it was actually just a 12/8 with one minor syncopation.) So hopefully this familiar rock metaphor just translates fine into 5/4, and won't sound awkward or off-kilter at all, but I don't know, having spent so much time with odd time signatures they always sound intuitive to me.
The B section is sort of a parody of the cliche 3-3-3-3-2-2 (think Phil Collin's drum fill in In The Air Tonight, or the accents in Robert Plant singing "been a long LONEly LONEly LONEly LONEly LONE-LY time'), being 3-3-3-3-3-3-2, or 20 eight notes long. It doesn't work though, since it's too long, and I didn't set up the kind of melodic tension and resolution to make it payoff I guess. But it's interesting because the SAME track on which I tried to do the change-the-chord-ahead-of-the-measure thing in 5/8 back in the day, I also on there did a 3-3-3-2-2-2 thing which didn't really work either. But I don't think I was thinking of that consciously when I tried it again here. I was just already syncopating in 5/4 so tried to think of additional ways to do so.