Land of the Sirens (2017)




Compositional Analysis


Section 1


Section 2

Land of the Sirens (2017, 27'00, stereo, fixed – media)

Land of the Sirens (27 minutes’ duration, stereo, fixed media) is about the Odyssey, one of Homer’s two epics, along with the Iliad. It describes the exploits of Ulysses (Odysseus), King of Ithaca, following the mythological Trojan War. My work focuses on Ulysses and his men’s adventures on his raft, on the island of the Sirens, the mythical island of Anthemoessa in the Tyrrhenian Sea. The work includes sound objects that mimic sounds from the original narrative (crackling boat sounds, water sounds, and pointillistic micromelodies that allude to the presence of the Sirens’ chants). The work also addresses a key research question: how can variations in real-world source sound materials be transformed into new identities?

Escape Route – Scylla & Charybdis (2014) [14], a live-electronic music improvisation piece (Freeman, Styring and Conway) exploring Ulysses’ adventures in Scylla & Charybdis, and Cook’s Siren Song (2012) [15] for live Soprano voice (representing the Sirens’ chants) and Electronics are two electroacoustic works that directly relate to The Odyssey . My work, in contrast to these, is for fixed-media, and the Sirens’ songs are represented by more abstract pitched resonant sound objects. Other acousmatic works also use Greek mythology as a starting point to explore the concept of transfiguration. Carman’s Metamorphosis I (2008-2009) [16] focusses on the metamorphosis of Pygmalion’s statue to a female figure and Stavropoulos’ Atropos (2003) [17] uses the mythological Atropos, as one of the piece’s main structural elements: “[…] the direction of energy […] is supervised […] by the intrinsic morphology of the sounds. [18] The latter explicitly transforms the mythological aspect into a compositional methodology. In contrast, my work is more directly related to symbolising Homer’s themes and characters.


[14] Alan Freeman, Steve Freeman, Chris Conway, and Simon Styring, Escape Route – Scylla & Charybdis, (Leicester: Auricle, 2014).

[15] C. Cook, Christopher Cook’s stream on Longleaf Music (BMI) (2012), Longleaf Music [accessed: 17 Apr. 2017].

[16] O. Carman, Metamorphosis I, (unpublished music work, 2008-2009): (email to the author, 5 October 2017).

[17] N. Stavropoulos, Nikos Stavropoulos stream on SoundCloud – Atropos (stereo reduction) (2011), Sound-Cloud [accessed: 7 May 2017].

[18] "In Greek mythology, Atropos was one of the three Moirae (the Fates), female deities who supervised fate rather than determining it. Atropos was the fate who cut the thread or web of life. She was known as the 'inflexible' or 'inevitable' and cut this thread with the 'abhorred shears'. Although the title is not directly related to the content of the work, it was chosen to reflect compositional processes and their relation to sound materials. Here, the direction of energy, and the movement and positioning in time and in more general structural relationships, is supervised and characterized by the intrinsic morphology of the sounds, as opposed to being deterministically formulated. In this respect, the choice of a Moira name metaphorically indicates the acousmatic processes involved in the work's composition. Atropos is a highly abstract work and does not refer to anything outside of itself. Original recordings are not traceable in the work's sound world and although most of the material has been synthetically generated it exhibits physicality in content, character and behaviour"; N. Stavropoulos, Nikos Stavropoulos stream on SoundCloud – Atropos (stereo reduction) (2011), Sound-Cloud [accessed: 7 May 2017].

Land of the Sirens is divided into two interconnected main sections that are largely contrasted by the uniqueness of soundworld. Section 1 from 0:00 – 11:50 and Section 2 from 11:50 – 27:00.

Section 1 of Land of the Sirens is divided into eight subsections. Subsection 1 (0:00 – 1:00) is the introductory section, which includes chaotic drones as well as pointillistic sound objects. Subsections 2 (1:00 – 3:03), 4 (4:36 – 6:05), 6 (7:42 – 9:20), and 8 (10:52 – 11:50) are distinguished primarily by the presence of authentically recorded and transformed forest wood sounds. Subsections 3 (3:03 – 4:36), 5 (6:05 – 7:42), and 7 (9:20 – 10:52) sonically contrast with subsections 2, 4, 6, and 8: subsections 3, 5, and 7 contain more abstract and gesturally active sound materials, which are overlapped by discrete proximate and distant rhythmic schemes (e.g., 3:30). Transitional passages are also present to achieve a smoother transition between two consecutive subsections. Low-frequency continuous dragging motion, for example, characterized by smooth fluctuations in spectral space occupancy (5:50 – 5:57), aids in the transition from subsection 4 to subsection 5.

I used real-world tree branch sounds in subsections 2, 4, 6, and 8 of Land of the Sirens, which resulted in the creation of continuous heavy crackling pointillistic textures. The sounds of fallen tree branches being broken, recorded at Entwistle Reservoir (North Manchester), represent the cracking sounds of the wooden boat. Some of the dry tree branch sounds were transformed into aggressive water sound textures by using a Logic Pro delay filter called “wet splinters”, which added a watery ambiance to the source sound material and completely changed its texture.  I emphasised the effect of sea water contact with a wooden surface, Ulysses' raft, by using branch sounds transformed into water soundworlds, and attempted to depict this in a hyper-realistic manner. This answers the key research question: how can variations in real-world source sound materials be transformed into new identities? The presence of pitched materials heightens the contrast between the sections of recorded tree branch sounds and the sections that intervene between them.

The recorded tree branch sounds were subjected to a series of transformation processes. The use of a low-pass filter in Avid Pro Tools EQ3 7 Band to eliminate the outdoor environmental sounds and to focus on the sound-object attributes. The use of Cecilia to impose pitched resonance on non-pitched sounds. The use of sound objects in a variety of transpositions to reveal the intrinsic properties and natural internal gestures of the sound. Batchelor’s Clatter created new complex gestures. Finally, using Gersic’s Atomic Cloud – Grain Cloud Generator created repetitive dry pointillistic textures in rapid motion. [19] These techniques sonically enhanced the wooden boat’s crackling sounds and created multidirectionality [20] in gestural motion. 

When many different layers of transformed wooden sounds gradually and causally create an enhanced filled-up spectral space surrounding the originally recorded sound objects, dilation (becoming wider or larger) in the multidirectional gestural motion is created (e.g., 1:15). The higher a sound’s frequency in space, the higher its pitch is perceived by the ear. The superimposition of sounds of varying frequencies (from low to high) over each other resulted in the formation of a filled spectral space. The use of primarily low-frequency sounds resulted in the formation of less-occupied spectral spaces.

Using the original branch sounds (e.g., 1:00 – 1:12) creates a sense of intrinsic aggressiveness (corresponding to the character of the narrative scene) of sound motion. This is due to the real-time gestures applied to the branches during the recording process, as well as how these gestures were later shaped during the compositional process. The transformed versions of these sounds, including the ‘watery’ sound textures (e.g., 1:37 – 1:38, 1:39 and 2:06), also imply aggressive behaviour. The watery sounds are thus given a hyper-realistic dimension because the water texture appears to be original, but the gestures are attributed by a completely different entity (tree sounds).

Subsection 1 is re-elaborated as a background layer over subsections 2, 4, 6, and 8. It includes structural units of harmonicity (e.g., 0:50 – 1:10). This helps to create a sense of coherence throughout the work. Formal contrasts are created by differences in activity levels between subsections: 1 and 3 (with incidences of soft drag motion, e.g., 4:22 – 4:25); 5 (with compacted multidirectional discontinuous motion with causal appearance and disappearance, e.g., 7:13 – 7:33, and onsets followed by smooth low-frequency growling prolongations [21] which fade out, e.g., 7:01 – 7:02 and 7:03 – 7:06); 7 (less activity); 2 and 4 (featuring smooth spiral texture motion, e.g., 4:41 – 4:43); and 6 and 8 (increased activity through onsets followed by decelerated prolongations which vanish gradually, e.g., 10:52 – 10:54 and 10:55 – 10:56).

This allows the piece to move gently between filled-up spectral spaces (subsections 2, 4, 6, 8) and less-occupied spectral spaces (subsections 1, 3, 5, 7), though a sense of continuity throughout the entire part is also present to ensure its identity. Furthermore, the contrast between subsections allows the work to breathe as the activity and tension created dissolves into slower gestural motions, resulting in a rising/falling flood-tide textural movement as tension is built and released. 


[19] Atomic Cloud. Grain Cloud Generator by Tom Gersic. [accessed: 4 Apr. 2017].

[20] “Bi/multidirectional motions create expectations, and most have a sense of directed motion. They can be regarded as having both gestural and textural tendencies, and could be large structures in themselves”; (Smalley, 1997).

[21] Streaming of low-frequency repetitive dry sound elements in semi-rapid dragging motion.

Dhomont’s use of instrumental soundworlds in Citadelle Intérieure (1981) [22] is an important point of reference for the approach I took in Land of the Sirens Section 2. Dhomont’s soundworlds are sometimes characterized by harmonicity, slow motion, and continuity, and they provide a mysterious atmosphere and a sense of expectation; the use of ascension and descension in textural motion contributes to this. Furthermore, his use of distinct soundworlds creates strong contrasts, seen in spectromorphologies consisting of ‘human voice' (attack – impulse) onsets, where dry rapid-motion spectromorphologies are balanced by intervals of silence or anacrusis.

Normandeau’s Erinyes (2001) [23] used male and female actor voices, as well as tape delay and amplification techniques, to depict a mythological entity: the Erinyes. According to Greek mythology, the Erinyes protected human subsistence and penalized evildoers. [24]

The textural motion in this work is characterized primarily by slow pace and prolonged continuants, but the voices are also ‘explored from within’. [25] In contrast to Normandeau’s work, apart from using a traditional Greek instrument to represent the voices of the Sirens (rather than recording human voices), the presence of the transformed floghera soundworlds in subsection 9 of Section 2 has a double meaning in my work. The transformed floghera soundworlds could be perceived as ‘bird’ soundworlds (indication of space/place) or as the Sirens’ voices (presentation of mythological action).

Section 2 of Land of the Sirens reinterprets aspects of Section 1 by evoking new soundworlds and soundworlds that reappear as leitmotifs, giving the work a sense of identity within each section and overall coherency. The following are the primary sound types: floghera textures, slippery/dinging micro-elements, water elements, unprocessed dry wood crackling sounds, bell sounds, pitched drops, beeping sounds, transformed church organ sounds, and repetitive dry pointillistic ticking [26] sound textures in rapid motion. Notably, original water sounds from Entwistle Reservoir play a larger role in Section 2, and a highly contrasting section contains transformed sounds from a traditional Greek instrument called a floghera (a traditional wind instrument that belongs to the Hellenic pastoral music instruments and has its origins in an ancient Greek instrument called a syrighx, originating in the Classical period). The floghera material was created with the BEAST Tools Clatter module, followed by a stereo down-mix in REAPER [27] and then transposed in a variety of pitch ranges with MAGIX Sound Forge.

In contrast to Section 1Section 2 did not adhere to the A – B – A repetitive structure (where [A] was composed of transformed watery/branch sound subsections, fading in/out among abstract reverberant subsections [B]). Section 2 was created in a non-repetitive manner, with materials from each subsection rarely being reused in different subsections, and even when they were, they did not make appear/disappear at regular time intervals, as opposed to Section 1. The beeping sounds were reused in the work’s epilogue.

In Section 2, the following characteristics of sonic and musical materials are described (with examples):

  1. Dissipation (disintegrating/dilution/dispersing) 17:08 and 22:56 – 23:01.
  2. Harmonicity (harmonic sound textures) 24:53 – 25:20. Background piano sound textures in A minor.
  3. Endogeny (smooth growing from the inside) 26:14 – 26:22. Transformed church organ textures using delay effects resulting in 3rd order surrogacy.
  4. Flocking (collective motion of ‘slippery/dinging’ microelements) 16:54 – 17:00 and 23:32 – 23:50.
  5. Emptiness (spectral gaps) 18:07 – 18:20.
  6. Plenitude (filled spectral space) 25:47 – 26:12.
  7. Insertion of smooth pointillistic background sound objects 24:44 – 24:50.
  8. Pressured Onsets 15:38 – 15:46, 15:53, 15:59, 16:01 and 24:12 – 24:28 (tree branches).
  9. Flow (emergence as if the motion has always existed) 12:04 – 12:12 and 24:54 – 25:22.
  10. Contraction (becoming smaller) 11:53 – 12:07.
  11. Pointillistic textures 15:11 – 15:29, 15:47 – 15:58 (dry wood) and 25:37 – 25:42.
  12. Cyclic motion 15:29 – 15:36 (dry wood branches) and 24:51 – 25:00.


[22] F. Dhomont, Sous le regard d'un soleil noir, (Montréal: empreintes DIGITALes, 1996).

[23] R. Normandeau, Claire De Terre, (Montréal: empreintes DIGITALes, 2011).

[24] Talfourd Ely, The Gods of Greece and Rome (Mineola, New York: Dover Publications, 2003), p. 208.

[25] “The principal sound treatment was designed to bring out the primitive nature of the voice — the interior resonance that is so deeply rooted in the human unconscious. This treatment is called ‘freeze.’ At first glance this may seem absurd, given that music is something that exists in time, but the computer allows the composer to stop time. Voices can be ‘frozen’ and thoroughly explored from within. Erinyes is the fourth piece in the Onomatopœia cycle (the three preceding pieces being Éclats de voixSpleen, and Le renard et la rose)”; R. Normandeau, Robert Normandeau’s commentary on Erinyes > ElectroCD (2001) [accessed: 7 May 2017].

[26] Semi-rapid motion of dry microelements characterized by brief and semi-aggressive throw onset/attacks.

[27] Cockos Incorporated. REAPER | Audio Production Without Limits [accessed: 17 May 2017].


[Figure 3] The Floghera


Section 2 is divided into fifteen subsections, some of which overlap. When a new subsection (N) has already begun while the previous subsection (P) is coming to an end, the letter ‘o’ (overlap) is used next to the subsection (P) end time indication to clarify that a new subsection has already begun. Simultaneously, when a new subsection (N) begins before a previous subsection (P) ends, the letter ‘ω’ (omega) is placed next to the subsection (N) start time indication to clarify that, even though a new subsection has begun, the previous subsection is still audible. The letter ‘c' is placed next to the subsection (N) start time indication when the sound event of a new subsection (N) is a continuation of the sound event of the previous subsection (P), with no observed overlapping. The abbreviation ‘σ’ (meaning, to be continued) is placed next to the end time indication for subsection (P).

The transformed floghera textures in Section 2 are classified into four types: 1) static drones superimposed by birds’ soundworlds attacks–onsets  (high-pitched sounds/fast-speed movement); 2) static drones superimposed by birds’ soundworlds attacks–onsets (low-pitched sounds/slow-speed movement); 3) static drones superimposed by birds’ soundworlds attacks–onsets (high-pitched sounds/slower-speed movement); and, 4) static drones superimposed by birds’ soundworlds attacks–onsets (high-pitched sounds/moderate-speed movement).

Subsection 1 (11:50 – 11:56). Transition between Section 1 and Section 2; spectral emptiness; static movement; allows the work to breathe and proceed to Section 2. It allows for the gradual implementation of a completely new soundworld environment.

Subsection 2 (11:56 – 12:08). Transformed floghera bird-like textures.

Subsection 3 (12:03ω – 12:14). Water textures.

Subsection 4 (12:14 – 12:41). Rain falling on dry leaves (foreground) / subtle continuous floghera sound with superimpositions of transformed bird textures (background).

Subsection 5 (12:41 – 13:03σ). Church organ upbeat steps, transformed sound textures (smooth anacrusis) leading to whistling-sound shapes emerging from transformed church organ textures that gradually fade-out while subsection 6 is already underway. The transformed floghera textures are still present and gradually become more noticeable (from 12:43 onwards) as they serve as the foundation for subsections 6: (12:49 – 12:56) and (12:58 – 13:01). Owl-like sound textures. [28]

Subsection 6 (13:03c – 15:57o). Transformed floghera sounds with a high proportion of low-frequency material: (13:09 – 13:14). Slow motion owl-like sound textures at low frequencies. The main sound is spectromorphological emptiness, with occasional smooth high-pitched floghera soundworlds layer additions. However, spectromorphological plenitude is not achieved in this subsection because the smooth high-pitched floghera layer additions revert back to spectromorphological emptiness. Smooth spectral fluctuations are defined as gradual transitions between spectromorphologically empty and slightly more filled-up spectral spaces, and their duration ranges from one second (instant) to a few seconds (continuous). This subsection is enhanced by smooth spectral fluctuations (e.g., 13:29 (instant) and 14:19 – 14:25 and 14:50 – 14:55 (continuous)), adding an element of suspense/expectation. Low-pitched sounds dominate. This subsection was created to contrast with the following subsection (9) which contains higher-pitched transformed floghera sound textures with spectromorphological plenitude.

Subsection 7 (15:10ω – 16:07). Dry wood crackling sounds (natural/unprocessed) and floghera fadeout. This subsection reintroduces the wood crackling sounds from Section 1 in their natural form, with no transformations. This subsection is distinguished by structural stasis (created by repetitive wood crackling sounds of similar volume level and occupancy in spectral space), with intensity increasing each time a wood crackling sound event is presented. The introduction of subsection 7 shows cyclic motion through repetition (15:27 – 15:36). The sound of a dry branch hitting a group of dry branches in equal time intervals was captured during the recording process in Entwistle Reservoir for this part, to imply inherent energy.

Subsection 8 (16:07 – 18:24σ). This is the only subsection made up of micro-chaotic spectromorphologies:

  • Scratching (16:11 – 16:55). [29]
  • Crackling in gelatin is presented alongside scratching sounds.The word “gelatin” refers to the fact that the sounds appear to be moving within a space filled with a dense liquid (16:27, 16:33, 16:37 – 16:40).
  • Crackling from the inside (16:44 – 16:49). 
  • Slippery [30]/dinging microelements (rapid motion) (16:54 – 17:02).
  • Divergence/Convergence and simultaneous linear ascent/descent (17:02 – 17:07).
  • Smooth Anacrusis (16:50 – 16:54) leading to the rapid motion of the slippery/dinging microelements (16:54 – 17:02).
  • Smooth water drop texture (17:20 and 18:00).
  • Pointillistic ticking sound events in non-rapid motion (17:18 – 17:19).
  • Friction/Resistance (mutually rubbing spectromorphologies) (17:32 – 18:00). These spectromorphologies exist in tandem with other layers.
  • Ascending contour (17:44 – 17:50 and 23:59 – 24:07). 
  • Descending contour [31] (17:15 and 17:24, 17:46 – 17:47 [32] and 23:50 – 23:56).
  • Ascending bubbling spectromorphologies. Because these do not sound like water bubbles, this term simply refers to their movement (17:49).
  • Pitched drops (17:54).
  • Static smooth background pitched layer (18:05 – 18:08 and 18:12 – 18:20).
  • High frequency resonant throw/fling (18:22).

Subsection 9 (18:24c – 23:37o). A significant subsection of Section 2 and the one with the most spectral space occupancy. The transformed floghera material sounds like a mix of birds and mysterious voices. Furthermore, the bird-like textures inform the listener about the location (place/space) and represent mythological action (Sirens’ voices). The floghera textures are distinguished by their high spectral density and extended continuants. The superimposition of higher frequency fading in and out in rapid motion over the high spectral density floghera spectromorphologies of extended continuants creates resistance as well as bird-like textures/Sirens voices. Exogeny is caused by the superimposition of higher frequency floghera textures in rapid motion, which results in spectral growth by adding to the exterior (e.g., 19:35 – 19:56).

Subsection 10 (20:15ω – 20:37σ). The gradual appearance of water textures in rapid motion, which leads to subsection 11. This lays the groundwork for the reappearance of Section 1 rapid motion spectromorphologies in subsection 11.

Subsection 11 (20:37c – 23:03σ). The reappearance of rapid motion spectromorphologies from Section 1. These spectromorphologies are the result of a series of superimpositions of various layers from Section 1. Because of the direct reference to material used in Section 1, this unifies the piece as a whole. Furthermore, floghera soundworlds from subsection 9 remain present but gradually fade out (dissolving). The rapid motion spectromorphologies of Section 1 and the floghera soundworlds contrast. Section 1 spectromorphologies become significantly sparser towards the end of this subsection, as they are superimposed over high frequency floghera spectromorphologies. This causes a counter-endogeny (deflation) in the textural motion (22:55 – 23:03).

Subsection 12 (23:03c – 23:33). The final appearance of floghera sound textures. The characteristics are the same as in subsection 10, but the volume level is lower in this subsection.

Subsection 13 (23:33 – 25:27σ). This subsection begins with the application of the characteristic slippery/dinging spectromorphologies (23:32 – 23:50) which were previously presented (16:54 – 17:02). Micro-textures from Section 1 and Section 2 subsections can be observed: repetitive dry pointillistic ticking sound textures in rapid motion from Section 1 (24:06), and pointillistic ticking sound events in non-rapid motion from Section 2 (24:07). This subsection is characterized by ascending and descending textural motion (e.g., descending 23:50 – 23:55 and ascending 23:59 – 24:06). Dry unprocessed crackling wood sounds can be heard from 24:11 – 24:28. Background piano soundworlds with no attack – onset (e.g., 25:03) create a feeling of release while smooth background textures from Section 1 are explored.

Subsection 14 (25:27c – 25:38). A brief subsection characterized by spectromorphological emptiness, with soundworlds from Section 1 heard in the background at a distance. This subsection leads to Subsection 15 which is highly contrasting.

Subsection 15 (25:38 – 27:00). The work’s conclusion made use of superimposed material from the entire piece. The use of processed church organ intervallic pitches (25:40 – 25:50) creates dilation (becoming wider or larger) and contraction (becoming smaller, 26:05 – 27:00). The beeping sounds that appeared at the beginning of Section 1 reappeared near the end of the piece (26:39 – 26:47).

In terms of structure, both Section 1 and Section 2 have the following sound objects in common: water soundworlds (e.g. Section 1: (1:37 – 1:38), 1:39 and 2:06 and Section 2: 12:04 – 12:14), bell sounds (e.g. Section 1: 3:03 and Section 2: 26:00), beeping sounds (e.g. Section 1: 0:19 – 0:56 and Section 2: 26:39 – 26:47), dry crackling wood soundworlds (e.g. Section 1: 4:36 – 6:05 and Section 2: 24:11 – 24:28) and repetitive dry pointillistic ticking sound textures in rapid motion (e.g. Section 1: 1:22 – 1:27 and Section 2: 24:06). On the other hand, the following sound objects are only present in Section 2: floghera textures (18:30 – 23:30), pitched drops (17:54), transformed church organ soundworlds (25:40 – 25:50) and slippery/dinging microelements (23:32 – 23:50). In addition, in Section 1 only, stone soundworlds are present (e.g. 3:18 – 3:27), as well as guitar sound objects (e.g. 3:46) and dry wood sound objects with pitched resonance applied through Cecilia’s Harmonizer Module (e.g. 0:50 – 1:07).


[28] Rhythmic sound textures comprised of a short-duration repetitive inharmonic sound element. Textural motion is used to distinguish each element. These textures are strikingly similar to owl soundworlds.

[29] Soundworlds characterized by reciprocal motion. These soundworlds appear to be dragged onto a dry surface.

[30] Flocking with a hint of resonance.

[31] The inverse of ascending contour (downward motion).

[32] Present at the same time as the ascending contour (17:44 – 17:50).