Six-note chords

Hello composers !

A few weeks ago I published this piece in Soundcloud. I know that there are splendid pianists in this Forum, let's see what you think of my composition. . .


As usual, I have made a (long) theoretical explanation about the work, that attached below.

Thanks for your attention.




Six-note chords

This work is the third in a series of piano pieces that I have composed following the harmony that defines the Tonnetz. The first was Parsimonious Trichords and the second was Walking by four-note chords. As I explained in the introduction to these two works, the Tonnetz comes from the German term Tonnetzwerk which means network of tones. The prototypical case is the network or graph in whose nodes each of the 12 notes of the chromatic scale are placed in such a way that the arcs that join these nodes form the chords of three notes, major and minor, distributed in such a way that the chords that are contiguous are only differentiated by a single note with only a semitone or tone difference. This little variation in the transit of one chord to its neighbor has caused them to be called parsimonious chords, in the sense of scarce, thrifty. On this network we can chain the chords following different itineraries or paths, thus defining the harmonic progression. In the piece Parsimonious Trichords I used this Tonnetz with three note chords. But it is also possible to define other Tonnetz with chords of more notes that maintain this parsimonious structure in which the contiguous chords have this minimum difference of half or one tone. This process of definition has been well studied by Dmitri Tymoczko in his excellent article The Generalized Tonnetz ( In the second piece Walking by four-note chords I worked with a Tonnetz of four-note chords and in the present piece I have used the Tonnetz with six-note chords. It is only possible to build Tonnetz that meet the condition that the number of notes that form the chords is a divisor of the total number of notes available in the musical scale used; therefore, in the case of the 12-note chromatic scale, only four different Tonnetz can be built, those with 2-, 3-, 4-, or 6-note chords. (In my case, I would need to compose a piece with the 2-note chord Tonnetz, which I will probably do in the near future).

 The evolution of the size and complexity of the Tonnetz is remarkable. We are going from the 24 major and minor chords that make up the Tonnetz of 3-note chord, to the 42 chords of the Tonnetz of 4-note chord (dominant seventh, half-diminished seventh, minor seventh and French sixth chords), to reach the 124 chords that make up the Tonnetz of 6-note chords. In this latter case, the chords do not have their own name, but they can be classified into 12 different types according to the distribution of the amplitudes of the intervals that appear between the contiguous notes that vary between one, two or three semitones. Another significant difference is that the harmony provided by 6-note chords is more dense and more dissonant than that obtained by seventh chords and much more than the diaphanous harmony obtained by triads. The third difference to highlight is the notable increase in parsimony, of the little variation that exists between the contiguous 6-note chords; this increase is due to the fact that we went from a proportion of 1/3 to another of 1/6 of different notes with respect to the total number of notes of the chord.

 As I did in the first two works, I have defined the harmonic progression on the Tonnetz by choosing the contiguous chords that follow a Hamiltonian path, that is, a path that passes through all the chords of the Tonnetz but only once. Specifically, I have used 3 different Hamiltonian paths, one for each of the three parts of the work.

 The construction work of the Tonnetz of 6-note chords is laborious and I have done it by programming diverse algorithms in Python language. I have done the same for the search for the Hamiltonian itineraries, but in this case the difficulty has been much greater because this involves solving a problem that, in mathematics, is called a NP-complete type problem that requires practically infinite time to be solved, when the size of the set where it arises has a considerable size. What was very easy in a set of 12 elements (Tonnetz of 3-note chords) and also in the set of 42 chords of 4 notes, has become practically impossible in the current set of 124 chords of 6 notes. It has not been possible to write or find an algorithm that computes a Hamiltonian path in a 124-element graph in a time that is practicable with the computing tools at my disposal. This has forced me to make some approximations to get paths that were not purely Hamiltonian, but almost. The itineraries used have an 8% deviation from the Hamiltonian itineraries; but perhaps this small deviation has a favourable effect: to give a little more variety to the harmony and alleviate the possible excessive monotony due to the great parsimony that the 6-note chords would have if they were part of a perfect Hamiltonian itinerary.

You need to be a member of Composers' Forum to add comments!

Join Composers' Forum

Email me when people reply –


  • Interesting concept, but I have to admit that merely a sequence of chords doesn't really draw me. What about using your chord sequences as a basis upon which to construct something more interesting? Like melodic lines that trace out these chords, accompaniment figures based on them, etc.. I know you did some of that, but the majority of the piece sounds to my (probably biased) ears like just a sequence of block chords. Would've been more interesting if you did something more with them than just block chords.

    • I appreciate your sincere criticism. I understand that my results are not very of your taste, as a consequence, obviously, in your way of understanding music. But my general musical approaches as well as the demands of the characteristics of the Parsimonius Chords, have taken me to these results that are quite far from your considerations, but that personally, they do convince me.

      Thank you for being interested in listening and reflecting on my music.

  • I listened to it, and it seemed to me (as someone who is fairly conventional or even old fashioned as a music listener and composer) an interesting experiment, which might point a new direction of development for atonality.  I thought though that it would be much easier for many of us to understand the notes on it, and the discussion of it, if we could see the score.

    • You are absolutely right that the score is essential to analyze a work in depth.

      But I am sorry to say that the version of the score that I currently have is still not presentable for the human interpreter since, as a consequence of my working methods, my "music sheet" on which I base the creation of the composition is what I call, the "machine-score", which the Sibelius program (editor-playback engine) understands perfectly but, has passages, which are not easily decipherable by a human interpreter as there is still quite a lot of underlying MIDI code. Obtaining the final score is a job that I leave pending for the end and, sometimes, a bit parked, given its industriousness and its low probability of being interpreted. If in this case I finished it quickly in the near future, I would certainly publish it.

      I really appreciate your interest.

  • Hi Ramon, I find your mathematical approach and the result fascinating. I don't fully understand the mechanisms you describe but I will be listening again to the music.

    I'm wondering if a quantum computer could speed up finding a pure Hamiltonian path through your graph?  I guess algorithms for quantum computing are somewhat limited at this point.

    I notice you have a large number of followers on SoundCloud, I find that very encouraging for the future of experimental music.

    Thanks for posting!

    • I'm glad you found my work fascinating ! I appreciate this comment and the effort to read the theoretical explanation. Perhaps I should apologize for not having completely successful in choosing the mathematically correct terms and at the same time they were understandable through a standard language, but it is not an easy task without the existence of common previous background.

      Undoubtedly, a quantum computer (even one of the current digital electronic supercomputers) could solve the problem of correctly finding Hamiltonian itineraries, but we are still far from the normalized use of this type of tools. But, as I said in the introduction, it may be that the imperfection of the itineraries used, though it decreases formal coherence, have favorably influenced the final result of the work, by giving a little more variety to the very high parsimony of the harmonic progression of the 6-note chords.

      Regarding Soundcloud, I think that I have achieved some follow-up because I have tried an active attitude participating in the comments of the authors that I follow and also exchanging messages and in the reposts, however it is difficult to get audience in this medium for the type of music that I do.


This reply was deleted.