Seeking prototype sound for computer music compositions...

    Mar 25 2010 | 4:38 pm
    Over the years I have slowly been moving towards making Max my composition environment. A place that lets me rapidly build and develop improvised and algorithmic form and project that into various pitch spaces.
    The other half of the computer music equation is the world of spectromorphology. Normally I do this in a later audio production stage, where the piece is rendered. Ideally I want to experiment with elements of spectral language at the same stage as I would the tonal harmony elements. That way I consider them equally and understand how they contribute to the composition. Having composed the spectral changes in the sound, I could then replace it with a target timbre in the audio production stage.
    So that leads me to my question. I was wondering if someone could recommend the ideal timbre to use for prototyping spectromorphology in the early stages of a piece. One that is easy to use but has great expressiveness. It is easy to get sucked in to the world of sound design and totally lose track of the composition. I feel having a timbre that you use to compose with lets you make smart composition decisions that way you do not waste a lot of time on the audio production side.
    I welcome any thoughts...

    • Mar 25 2010 | 10:20 pm
      I'm not quite sure what spectromorphology means. Regarding timbre you could try feeding noise into fffb~ as a prototype. fffb~ is a bank of bandpass filters. Check out the help patch - it's really simple. By setting the resonancy of the filters you can easily morph between noise and tone. (the multislider for the Q's in the help patch is limited to 100 but you can go above that).
    • Mar 25 2010 | 10:57 pm
      Spectromorphology is basically the language that is created from the spectrum of the sound. That is to say, the forms that we perceive as the spectrum of a sound evolves and changes over time.
      Denis Smalley wrote a great article on Spectromorphology...
      In it he explains the basic building blocks of a spectral language and how we perceive these elements.
      It is a must read for all those who are interested in electroacoustic music.
    • Mar 27 2010 | 7:04 am
      Guess I had to chime in (pun intended) given my user name ; )
      I read this post and had a think about it for a while. But it strikes me that there really isn't an (easy) answer to this question.
      For tonal music organized via rhythmic, melodic and or harmonic means I think the piano is the closest thing to a template sound, and this in large part is really down to its keyboard ( or interface) - polyphonic and to a degree reflects western music notation rather than its sound as such.
      By dint of the fact that, in theory at least, spectromorphological 'analysis' could ascribe any sound event with a unique set of qualities and map the possible parametric variations in the descriptive characteristics of even a 'single" sound event, it would be far more of a presupposition to assume a "neutral" sound in this context than having a piano sound as a stand-in for some other (melodic/ harmonic) musical instrument.
      I'm aware there are differences in this comparison, but its kind of like asking for a word that will stand in for all the words that you may use while writing a sentence or paragraph or..
      To that end, the only vaguely viable 'stand-in' would be a general method that can approximate a similar timbral and temporal profile as the intended sound(s). This comes back to broad-band sound sources (noise?) with (ideally) flexible control over harmonic emphasis (or de-emphasis) and temporal evolution. in other words - generalized subtractive synthesis. This in turn implies it's conceptual opposite - additive synthesis -as a possible alternative approach.
      If nothing else, this approach is suggested by the materials that spectormorphologically inclined composers (like Smalley) use as the source material for much of their work - generally timbrallly rich and prone to inharmonicity. (Ie metallic textures and sharp attacks) from which it may be arguably technically easier to soften, smear, sustain and modulate, thereby getting more spectromorpholical bang for their buck...
      Just some thoughts...
    • Mar 27 2010 | 9:37 am
      I would say that any step to describe in words music a-priori is getting away from the strength of studio composition, i.e. working with the sound(ing) objects themselves and using your ear to judge of the music.
      Also, an intuitive approach in the studio, and loosing studio time, is very important in a composition process I think: it allows one to trash unsatisfactory results.
      So to reply to your question, I tend to draw very simple curves not to loose macroscopic perspective on my music: energy curves, dramatic tensions, leitmotiv reappearance and variations, volume, density. I do the same with all music, rap producing included.
      Spectromorphology is a good analysis tool, but the main question is therefore: is analysis a good composition tool. I share Varèse's perspective on the matter, and say no.
    • Mar 28 2010 | 5:16 pm
      Thank you guys for the insight. After giving it some thought myself, I think there is question within the question I am asking...
      I think the real question I am trying to get at has more to do with the structure of language. And by language, I mean patterns of sounds that we give meaning to. Music for me, is that language. So what I am really trying to do with the prototype timbre is trying to understand the elements of grammar that make up the language of spectromorphology. Those language elements exist regardless of what aesthetic decisions ultimately determine what timbre is used.
      Pierre, you bring up a good point, is analysis a good composition tool. Can understanding another language help me understand my own? Our awareness of our own compositional language has been shaped by the analysis of other composers musical language. I think it is impossible to be in a musical vacuum. Perhaps what Varese was really saying was to truly communicate, you must forget what you think you know.
      In the end the result may be the same, the language of music is experiential. Studying or optimizing any one of its parts does not in the end mean you will have a better product. I need to learn to turn off the left side of my brain, sometimes it really gets in the way.
    • Mar 28 2010 | 9:18 pm
      Perhaps what Varese was really saying was to truly communicate, you must forget what you think you know.
      Maybe. But I also think that we need to fight against that knowledge, with a fair bit of intuition in the equation. An intuitive, esthesic, personal analysis is also very valid as an anchor, yet probably idiomatic and impossible to transfer...
    • Mar 28 2010 | 9:25 pm
      Yep. And watch out for the lure of those false economies of efficiency - no hay caminos hay que caminar, as the say in Toldeo. :-)