Sunday, April 25, 2021

2-bit Music

 I've been catching up on the excellent 8-bit Music Theory set of videos. The real eye-opener was his analysis of "Rain" from Halo 4: ODST. It sounds like sparse, almost random noodling on piano and sax but on analysis every single note is an important and integral choice to the final effect.

I've dabbled in this before. I'll revisit again, and each time hopefully learn a little more. In really simple form, there's three (maybe four) ways of looking at a harmonic progression. First is the quality of the chords themselves and their relationship to the key. This is where things like the ability of the V chord to be a pivot to the relative key (I think that's how it works!) or the fact that the II and IV are minor/diminished even within a major scale, come in.

Then there's the voicing, and this is how the intervals within the chord get brought out. Simple example; play the root triad of C major. Add the 7th degree to make a C7. Now do the first inversion. That top B that was originally heard in the context of a stack of major thirds is now a dissonant minor 2nd away from the root C.

So particularly in orchestral writing, where the grand clef can easily (even, commonly) extended 8va on either side, how you voice the chords can range from tight cluster chords to open voicing. Or can stress the powerful fifth or the more sombre fourth; the top note of a triad is both of these, being a fifth from the root but a fourth from the octave of the root!

And then there's voice leading. Our brains are very good at tracking a specific line through a pattern of chords, especially if it is in a distinct voice; say, a trumpet part within a mixed/lower brass section. So even if the net movement of the chord progression is going down, and even if the last cadence was from minor all the way grim to diminished...if the notes assigned to the trumpet are making a major fifth ascension at the same time...

Which is part of my "maybe four" above; the strings are quite consonant and of similar timbre and can play in any voicing, but because of the harmonic content brass speaks better in certain voicing, becoming more powerful or more dissonant than the intervals alone would suggest, and the winds have such distinctive characters they almost always throw focus onto the actual line played by each instrument or section.

***

So basically I'm into reharmonization. My first take on the arrangement was to check basic ideas and tone colors, and see if I had a workable pallet. The next sanity check is to record some sections and see if I can actually get a blend. To that end, I've found my music paper and am writing out the first string section.

Yes, Reaper does notation. But quantizing everything to make it look sensible would be a pain. Furthermore, because of how my tracks are organized the strings are all on separate systems. (The limitation of the tracks -- or more technically, the number of tracks I'm willing to try and deal with) means I couldn't really voice the harmonies fully anyhow, having things like two solo trumpets instead of a trumpet section playing divisi. 

And after those sanity checks...

Reharmonization! I need to take a further step back here and not follow the original as slavishly. I am in no way going to call Miracle of Sound a hack. It is completely appropriate and powerful the way he has done it and I know he has the skills. But the original piece was written around a defining and unchanging ostinato with a simple and fixed 3-chord progression over it. Okay, it does break slightly in the climax of the verse.

But I need to move outside, be prepared to go places that are outside of the key, and change the ostinato when needed. You can't write good jazz harmonies just by adding 7ths on top of a triad progression! 

This project is feeling more daunting by the moment. I'll feel better when I've tried recording a few lines. The bass line -- well, I've been practicing that, and it is a pain. Apparently Garritan gave me a five-string bass because I had to drop my low string a whole step to get to the repeated root D. I'm still having trouble damping the notes properly...the last time (only time!) I recorded bass, I could leave the strings open. Now I'm trying to control them to get the necessary tighter sound.


I applied some of the ideas above and cleaned up the first two melodic bars. The main error I had been making is reinforcing the melodic line by assigning as many instruments to it as I could manage. No, the trick was to make it the top voice and having distinctive motion lacking in the other voices. So now I've got that lush string sound, and I've been able to define the chords properly, and I've managed to keep the moving pedal going.

Didn't get it done in time to record at home. I've scribbled the string parts on music paper and I'll rehearse them for a day or two. Then I'll try the experiment. Not really looking forward to recording the same part from three to a dozen times to get the full string section sound...

No comments:

Post a Comment