Episode 4 - Blue Arabesque

The Seed

I don’t subscribe to the notion that good art comes solely from some kind of divine inspiration.  You can see the caricature in your head - the artist is driving down a road, and notices the way a beam of light hits a sign. They cock their head, pull off to the side of the road, rummage through the glove box to find some paper, and scrawl down something that cracks the world open.  I think it’s reductive, and worse gives creators license to be lazy and just wait for that lightning bolt to strike.  

The most successful artists I know are the ones that work _really_ hard.  They whittle themselves into vessels for receiving the ideas that would bounce off others.  For me, that’s what hoursongs do.  They exercise the muscles, and make me more observant and flexible.

This episode’s song, Blue Arabesque, is a clear reflection of the lessons learned from this process.

This song started like many of mine have recently, in D-Minor tuning (which is D - A - D - F - A - D).  I love alternate tunings because they turn the guitar, which is a really known instrument to me, into something wholly unfamiliar.  I started going back and forth between an unfretted open chord, and a shape that would sound as an A minor on standard tuning.  But on this tuning, it became something completely different.  

What even is that chord?  A - E - G - Bb?   I’m more of a populist about theory, so I’ve only learned enough to allow me to clearly express myself in the mode I want to work in.  But in general, unless I _have_ to learn it, I don't - unless something is really frustrating.  Which this chord ended up being.  I could talk about it for a while, but suffice to say, it’s an A Dominant 7 with a Flat 9 (I think)

So I had a nice little instrumental backdrop, and a melody ontop of it.

The Arrangement / What Would I Change?

When I played it, the rhythm of the song always had a push and a pull.  I thought that if I recorded my acoustic DI'd and sang live, I could follow the rhythm on another pass and record separate passes of the vocals and guitar to get isolation.  So I recorded a few live takes until I got one that had a nice flow, then tried to do another pass ontop of it - and, found the rhythm impossible to match.  So I kept the live take as a base.

I knew I wanted to have a fantasia at the beginning - a dreamlike section that introduced you to the song.  I downloaded a sample library that focuses on harp arpeggios, and spent some time futzing around with different runs.  You can pick different start and stop notes, in different modes, and at different speeds, so I played around with a set of them, recorded a half dozen, and then started experimenting with layering them.

PaulStretch is a tool that allows you to massively slow down the speed of audio.  It takes something that sounds like this <unprocessed> and turns it into this <processed>.  It is just this beautiful tool for sound design, and a great tool in my arsenal.  It used to just be distributed as raw code, which I compiled it using XCode and then fed audio through.  But now the nice folks at Audacity included it out of the box - so no more messing with compilers and nerdy computer science stuff that slows you down.  

I took one of the harp runs and went from this, to this:

I layered that in, and then wrote a little string part for the chorus sections, and the song felt like it was in the right place.

So - what would I change

• There’s a part of me that would like to take the opening fantasia section and really spread it out - maybe that could work as a transition between another piece?  The nice thing about hoursongs is that the format forces you to abandon stuff - more often than not the right decision.

• It would be nice to add some vocal harmonies.  I’m not positive about singing this entire piece in falsetto either.

• Finally, having some real strings would be _amazing_.