• Register
  • Help
Results 1 to 8 of 8

Topic: Discussion: Role of Timbre in Harmony

  1. #1

    Post Discussion: Role of Timbre in Harmony

    Ok, time to open what will hopefully be a much smaller can of worms than my last discussion post.

    Part I:

    I wonder if many of you composers have considered how much the role timbre plays in harmony. Obviously, those interested in orchestration have timbre very much in mind when writing. But I have lately been giving much thought to how individual timbres affect the quality of a given harmony in ways that aren't often discussed.

    Most of you are probably very familiar with the natural harmonic series. When dissonance occurs between two or more notes, the intensity of the dissonance is determined not just by the relationships between the fundamental notes, but also the relationships of all the harmonics in each note.

    When we learn about traditional theory, this fact plays only a very small role. It usually only manifests itself as a rule of thumb regarding spacing in a chord: "user larger harmonic intervals for lower pitches." Why is this a rule of thumb? The reason lies in the spacing of the harmonic series. Since our hearing tops out somewhere around 20khz and the harmonic series is more closely spaced the higher you go, lower pitches have many more overtones that fall within the natural range of hearing. More overtones means more potential for clashes between pitches in the lower registers.

    If this isn't clear, try the following experiment: Go to your piano, and play a harmonic major third at middle C and the E above it. listen to the harmony, and then play it an octave lower to compare. Repeat at each successively lower octave, and note how much denser and less "pleasant" it becomes in the lowest piano registers. The same experiment can be done in reverse, by picking a dissonant interval, and moving it progressively higher to hear the dissonance become progressively less harsh.

    Now let's examine the next facet of this idea. We know that different instruments have different harmonic makeups (timbre). This is how we can distinguish an oboe from a flute by ear. The timbre of an instrument is determined by the distribution of the harmonic series. If we know a little bit about the harmonic distribution of different instruments' sounds, we can use that as an in-road to some interesting orchestration experiments.

    To be continued in part II
    - Jamie Kowalski

    All Hands Music - Kowalski on the web
    The Ear Is Always Correct - Writings on composition

  2. #2

    Re: Discussion: Role of Timbre in Harmony

    Interesting discussion. Go on, please. I give it some thoughts and closer investigation.


  3. #3

    Re: Discussion: Role of Timbre in Harmony

    Part II:

    Below are a few instruments with general discriptions of their harmonic distribution:

    Strings and brass are very rich in harmonics, with many of the lower partials quite audible. The amplitude each partial is directly related to its position within the series (higher pitch = lower amplitude).

    Timpani are missing the fundamental note, meaning the harmonic series begins on partial 1.

    Clarinets (and stopped organ pipes) produce the odd-numbered harmonics almost exclusively

    Harmonics on the flute drop off very sharply, while those of the oboe drop much more slowly.

    Pitched percussion are somewhat unpredictable. A xylophone is usually missing the 2nd partial, while a vibraphone is missing the 2nd and 3rd. Metals tend to have a lot of inharmonicity (partials are "out of tune" from the natural series)

    All of this information may seem terribly irrelevant to what you're doing when you orchestrate. But if you like to play with combining instrument sounds, it helps to know what sort of interaction to expect. For example, a bass clarinet playing a C one octave below middle-C will have a strong overtone at the G above middle C, and a weaker one at the E a 6th above that. Because of this wide spread, it's easy to distribute dissonances within this range and not have quite so "harsh" a sound.

    Try combining the bass clarinet C with an English horn playing the Db a minor 9th above. Take a good listen to the sound of the 9th, then replace the bass clarinet with a trombone. Why does this instrumentation sound more dissonant when playing the same interval? The reason lies in the "holes" in the bass clarinet's overtone series. The middle C and C an octave above clash strongly with the Db fundamental and strong 1st partial in the English horn. But those same Cs are almost completely absent in the clarinet.

    So onto the extended discussion. I wonder if some of you have already thought along these lines, and whether this sort of thing is of any use or even interesting to all of you. If there is interest in the topic, I'll try to provide some examples from of my work where these considerations were used.
    - Jamie Kowalski

    All Hands Music - Kowalski on the web
    The Ear Is Always Correct - Writings on composition

  4. #4

    Re: Discussion: Role of Timbre in Harmony

    I does not seem to me, a question that should be asked or "discussed"!
    I don't mean that someone unaware of the place of timbre in harmony should not ask about it, but rather that it should be self-evident to anyone else.

    I rarely orchestrate using a guideline set in stone as to the timbral predominance. The "effect" I am trying to achieve is what guides my choices.

    In other words, the same chord - let us say, C major - can be heavily predominated by double reeds in one instance, sweetly blended of flutes and strings in another, or brassily pungent in yet another.

    In my own work, I rather enjoy the effet of the exact (or nearly so) same passage of music set in multiple orchestrations at different moments.

    I am also acutely aware that some passages do not necessarily FIT all instrumentations. I am of the school that orchestration should grow out of the music itself.

    There is nothing worse, in my opinion, than music that has been plunked down willy-nilly, and gratuitously handed out to whatever instrument "happens" to be there at the time.

    By this I don't mean to praise colouristic effect for effect's sake. However, music should be written for the instruments which will play it, not against.

  5. #5

    Re: Discussion: Role of Timbre in Harmony

    Quote Originally Posted by qccowboy
    I am also acutely aware that some passages do not necessarily FIT all instrumentations. I am of the school that orchestration should grow out of the music itself.

    There is nothing worse, in my opinion, than music that has been plunked down willy-nilly, and gratuitously handed out to whatever instrument "happens" to be there at the time.
    I agree with this sentiment. I think orchestration is seen as the last step to get the pure idea of the music realised. However, it should be considered hand-in-hand with the notes themselves. Composition and Orchestration may be two separate disciplines, but one should never be used without consideration for the other. This doesn't rule out an arrangement of an existing piece, but it does put it in a different light, I think.
    - Jamie Kowalski

    All Hands Music - Kowalski on the web
    The Ear Is Always Correct - Writings on composition

  6. #6

    Re: Discussion: Role of Timbre in Harmony

    I think it goes even deeper than just timbre. I did my MA dissertation on just this subject.

    I started by looking at a mathematical theory of harmony put together by Ernst Terhardt and refined by Richard Parncutt. This model attempted to simulate the perception of isolated chords by inputting individual harmonics, with their respective 'weights' in dB. The end result gave values for a number of perceptions, such as the 'rootedness' of the chord (i.e. how much it reinforces the harmonic series of the root note), roughness, etc. Whilst it sounds a little cold and calculated, the results actually tallied with subject's reported perceptions.

    The problem with the model was that it was for isolated chords, rather than progressions (Parncutt actually had a model for short progressions, but it stopped reflecting anything musically worthwhile as far as I was concerned), so I began thinking about ways to model the same perceptions across longer musical passages. One way I thought of was by crossing the model with the results of research into 'Auditory Scene Analysis.' This is a field that is attempting to quantify how we parse a whole mess of partials (harmonic or otherwise) into groups. The most relevant strand of this research boils down to determining how much the participation of a harmonic in a horizontal stream of information can prevent it participating in a vertical group (on a gross level, in contrapuntal writing, a note's 'coherence' to a melodic line weaken its participation in vertical harmonies, and you thus hear those harmonies differently than if they were sounded in isolation. But this also applies to individual harmonics. For instance, if a sine wave is sounded in alternation with a complex tone, one of whose harmonics is close in frequency to the sine, then it will 'capture' the harmonic into a horizontal stream, and the harmonic will become more noticeable, but give a reduced contribution to the perception of the complex tone). My argument was that a successful 'musically relevant' model needed not only to feed in the actual harmonics of each sound (as opposed to the simplification of treating each sound as one pitch which we composers have used for so long) but also re-weight each harmonic, before entering it into the vertical summation, according to its participation in horizontal streams. This would then simulate the effect of counterpoint and orchestration on the perception of each chord.

    I can't really remember the details now - it was a long, long time ago, but you can probably get hold of my dissertation by inter-library loan. It's here:-


    Some of the more interesting thinking along these lines is from Steve McAdams. He did a lot of experiments on what factors will cause tones to cohere - for instance, he played the odd and even harmonics of an oboe sample from two speakers, out of phase, and found that listeners perceived two separate sounds at each speaker, until the harmonics were brought in phase, when they suddenly perceived an oboe placed centrally.

    Let me know if this becomes an area that you start reading much about, and I'll give you the titles of all the books I used.

  7. #7

    Re: Discussion: Role of Timbre in Harmony

    Hi Jamie,

    I wonder if many of you composers have considered how much the role timbre plays in harmony.
    This is an interesting issue.

    It first depends upon whether the music is predominantly homophonic or polyphonic. And secondly upon the effect one wants to achieve:
    • Homophonic and clear
    • Homophonic and obscure
    • Polyphonic and clear
    • Polyphonic and obscure
    Of course, there are gradations among these simple categories, but that’s roughly the domain. I would grant that most composers don’t think consciously in these terms, but the effect is nonetheless intuitively operative somewhere in the background as they go about their work.

    Most composers, once they have voiced their harmony at the sketch stage, routinely desire the utmost clarity such that the each element of the harmony is quite clear, and go to great lengths to ensure that the orchestration is properly weighted to achieve this end.

    Other composers have very different aims. For example, highly dissonant harmony can be greatly softened via the orchestration, assigning the minor seconds and major sevenths to “softer” instruments, or having the brass use mutes, or marking the orchestration at different dynamic levels at the same point in the score. And alternatively, that dissonance may well be emphasized by doing quite the opposite.

    I remember studying the various overtone weights when doing classes in electronic music, but this was more for mimicking a conventional instrument or creating some novel electronic timbre. I would certainly accept that the notion of relative overtone strength could be consciously useful in some fashion for orchestration tasks, but the fact is that much “interesting” music changes both harmony and sonority too rapidly for overtone considerations to be of much practical use, other than the basic voicing.

    That is to say, direct application of this concept in a uniform and continuous way may prove difficult, and have little discernible resulting effect. However, sections that are relatively harmonically and/or timbrally static may well benefit, and I believe passages of extended parallelism having a relatively constant timbre may also benefit from a conscious allocation of the overtones inherent to each instrument, provided that is the effect the composer wishes to achieve.

    Another idea, which is related to your topic, is to explore is the creation of "synthetic" instruments from existing orchestral resources by arranging them in the overtone series and scoring them in strict parallel motion. The better instruments for this purpose are those with few strong overtones of their own, such as the flute, but others certainly could be used.

    The most famous example of this is the passage from Ravel's Bolero where he combines (if I remember) horn, two flutes and celesta to create an ethereal, almost electronic timbre from this combination.


  8. #8

    Re: Discussion: Role of Timbre in Harmony

    Not only have composers thought of this, there was a whole 'movement' or 'ism' starting in the '70's: "Spectralism".


    - k
    "An artist is someone who produces things that people don't need to have, but that he - for some reason - thinks it would be a good idea to give them."

    - Andy Warhol

Go Back to forum


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts