Obsidian

I started writing this piece a year ago, and worked on it for about four months. Then I put it to one side and started on a more melodic piece (Voile d'automne). Having completed the latter I was able to return to the former piece and complete it (as far as any piece is completed). So, here it is.

I based it on a lengthy chord progression. I wanted to explore a darker soundscape than in previous compositions, use many more exotic chords (the sound of which I really appreciate), and employ several different modes. I was mildly intrigued that the locrian mode doesn't cadence, and so confronted this issue directly.

One of the (no doubt many) weaknesses of the piece is that I have not found an adequate way to make greater use of a) the marimba; b) the voices. I did not want simply to include them when they were not needed just because they were there, yet I feel uncomfortable about how little contribution they make to the piece.

As always, all feedback will be gratefully received.

Not very sure how to respond to this. I’ll try but if you disagree please disregard these comments. It is your work after all.

I couldn’t quite understand the key signatures changing each bar at the opening. And in bar 4, why the B natural rather than C flat of the key signature (ok, if strings as the fingering would probably be different or harpists who are used to weird looking chords).

The marimba may not sound out, the way you’ve arranged the dynamics – f on the piano, p dimming to pp on the marimba (bar 8 on).

When you get to the fff (bar 35) the bass holding an independent line, the texture sounds muddy. This may be the effect you want, if so please disregard. (You’re right about the choir not being much use here if singing mp against fff.)

When you use the Basses for accentuation (bar 90 onward) it’s a lot easier to listen to. I suggest a change in arrangement of the piano part, here and elsewhere. Unless you have Lisztian hands the left hand stretches are impossible. Ok, you’d play the top E flat with the right thumb but it may look better to put it on the treble staff, stem down, to show it or say something in text on the score.

I listened out for the locrian mode but couldn’t detect it – possibly because, well, you pinpointed its cadence problems base triad/final has no 5^th, “dominant” is a diminished 5^th. You may have modulated out of it quickly on occasions you used it. I know of these problems having written a symphonic piece for strings in this mode and it gets hopelessly boring unless modulating frequently (and adding 7ths 9ths and 11ths to the basic triad).

Well done on composing a fairly long piece. It comes across to me in the semi-minimalist style (if there is such a thing) of Nyman.  

Quinn, thank you for your valuable response to Obsidian. I have been thinking a lot about each of the points you have made, and I intend to respond - not defensively, I hasten to add.

1. Whilst the work is indeed mine, I am far from confident about most aspects of musical composition. Just as people sometimes comment, on seeing a person walking a large dog: "Are you taking the dog for a walk, or is the dog taking you?"

2. Nowhere is the first point more evident than in terms of key signatures. In Obsidian I tried, wherever possible, to make the notes on the line as uncluttered as possible, changing the key signature accordingly. (However, I was not even fully consistent with this principle.) As ever, I was composing 'by ear'. For me, therefore, the piece is not in any specific key or keys - probably a nightmare for the oboe and bassoon. On what principles do other people address this issue?

3. I use MuseScore both for writing and for playback. The dynamics throughout the piece are what works for MuseScore playback. I am aware, however, that this could vary were an ensemble to play the piece live. From what you have written, it sounds like that variation could be quite substantial. Unless you are able to suggest otherwise, I guess that I am only really likely to get written dynamics right if the piece were to be played.

4. I was interested that you should pick out the section from bar 35, where you consider that the texture is muddy. I had played this section in an earlier incarnation countless times over many months, knowing that it lacked something. Only very recently I wrote the 'cello line, which improved things. However, accepting your feedback, I think I need to take out one of the other lines so as to 'unmuddy' the section.

5. The Lisztian hand-stretches (aptly-termed) expose the fact that I wanted to make use of some very bleak-sounding chords - to date, all of my compositions had been bright and fairly upbeat. I am more interested in darker music (Mahler, Rachmaninoff, Shostakovitch and Sibelius, among others) as well as less rigid music (such as Vaughan Williams), and naturally would like to explore those ideas. By way of beginning, I composed Obsidian for piano, and one of the reasons for orchestrating it for a small chamber orchestra was to make those many-fingered 11th chords feasible. However, there were a few places where I failed, and left the pianist requiring too many fingers. My bad.

6. Although there are key changes, the piece is written in Mixolydian from bar 8 to bar 88; in Locrian from bar 89 to bar 134 (with bars 126 - 129 failing to cadence on A flat minor to E flat diminshed; followed by bars 130 - 133 searching in vain for an alternative cadence); alternating between Dorian  and Locrian (Dorian 135 - 138; Locrian 139 - 142; Dorian 143 - 146; Locrian 147 - 175, the last section of which arrives once again at the failure to cadence), which is resolved by moving into Mixolydian (although my memory might be a little dodgy about this); the final two sections before the cadencing coda are both in Mixolydian.

7. Philip Glass (e.g. Koyaanisqatsi), John Adams and Michael Nyman (introduced through the soundtracks of several Peter Greenaway movies) are vital to my musical ear.

8. I should be most interested to hear your symphonic piece for strings in Locrian mode. Where might I find it.

Thank you again for taking the trouble to listen to Obsidian and offering critique. I am sincerely grateful.

I think it's pretty good. I like the style and the modulation (modal).

It's true that so many key signature changes make no sense because the reading is harder. In this case, I would write the score without key signature, marking all the accidentals.

The role of the marimba is nice. Not always all the instruments have to sound. But Perhaps a solution for those thick (and sometimes not playable) chords in the piano is to lighten those chords and complete the harmony with the marimba.

Regarding the voices, they're effective. Take care of the tesitura, it's good, but sometimes the women are in the limits (high A), for a chorus.

@Hughes

Thank you for responding to some of the dynamics issues in my post. I understand now - I've heard this before about notation software rendering the sound, the dynamics and stereo difficult to balance. 

As for the piano part, if you play you'll know you can cheat with seconds on white notes and some black notes (depending what the rest of the hand is doing)! I'm sure Liszt did and I know Ravel expected it in his piano concerti. 

What I'm pretty sure of is it would sound good played live (with suitable adjustments made to the dynamics on the finished score). The choir could be given nuances of expression lost in a notation rendering. 

I meant to add that I attended the performances of koyaanisqatsi and Powaqqatsi in London, led by Philip Glass. Likewise have seen most of Greenaway's films. Drowning by Numbers was an excellent Nyman track. My introduction to his workcame through a BBC broadcast of "Think Slow Act Fast". Shame that it was never commercially recorded.