HITHER GATE MUSIC Archer (Tom) Endrich Composers' Desktop Project

Workshop on How to Compose with CDP TEXTURE

~ CDP Workshop 2 by Archer Endrich ~

Also see Readtxws.txt

Topics

Part I – Information & Observations
Pieces of the Jigsaw:
source sound, note data file, parameters, breakpoints
Musical Possibilities Visualising the Outcome A Useful Toolbox
Group Formations Rhythms Old & New Subtle Differences
Designing Decorations Designing Motifs Designing Ornaments
Combining Multi-Events & Rhythms Combining Multi-Events, Harmony, Motifs & Rhythm
Part II – Focused Examples
1. One Pitch 2. Pitch Range 3. Harmonic Set
4. Time-Varying 5. Decorations & Ornaments
Play all Sound Examples – & links to Play Lists with Parameter Data


This workshop summarises key information and practical observations about using the CDP TEXTURE software created by the amazing Trevor Wishart, gleaned from many years of documenting and composing with TEXTURE. I then work through a stepped series of examples focused on specific TEXTURE facilities, aiming to give you an understanding of the software and the ability to use it to advantage in a variety of compositional projects.

A good place to start is to play through the 21 sound examples in Txwsplay-PC.htm and then look more closely at how they are done.

To unpack and set up the Texture Workshop on your hard disk, please see Readtxws.txt.


Part I – Information & Observations

Summary of Musical Possibilities

Texture in music is a big subject! In this introductory section, we will focus on the compositional issues: what 'texture' is, what forms it takes, how our new computer-based tools open out new possibilities.

1. Texture is a rather general reference to constructional types – not so much musical forms as the overall layout of vertical and horizontal components. A short list of some familiar types will help to illustrate what is meant:

The keywords are then: line (i.e., melody or tune), figures (motifs, embellishments, ornaments), density, overlapping, layers, patchwork, chords (harmonies, harmonic fields).

Motifs/embellishments/ornaments attached to or embedded in a melody (a 'tune') can be using the melody mainly as an outline to be filled in – a familiar aspect of some forms of variation technique. I like to think of the notes of such an outline as nodes – an underlying structure of nodes: a nodal substructure – to which other elements are attached. This concept provides a way of clarifying the relationship of melodic components and can help in the generation of linear structures. This is relevant to the TEXTURE programs, which also attach 'motifs' and 'ornaments' to 'lines' (or to Harmonic Fields/Sets, or to random pitch selections).

I personally prefer the term 'nodal substructure' because both the concept and application are somewhat abstract: in the manner of a steel framework rather than the finished surface of a building. Thus a pattern of nodes may have some 'tuneful' qualities, but often is not really a 'melody' or a 'tune', which are quite special linear forms. Trevor uses the term 'line'. The nodes are where other components are attached.

A texture is made up of multiple events, often with some form of repetition, similarities, &/or transformations, as well as a mix (or 'collage') of ingredients and an overall constructional pattern which may give it a sense of visual shape and/or dramatic gesture.

2. We can take another step in our understanding of 'texture' and its musical possibilities by considering its use in the music of the 20th century.

One way to start is to remind ourselves of the many textures with which we are familiar, sonic textures from Nature, human life and machinery. Textures such as the sound of: wind through leaves, waves, rain, the spring 'frog' chorus, the dawn chorus, groups of human voices speaking, shouting or singing, many feet walking, airplanes, traffic, motors, insects, factories, the battlefield, mixtures of all of these, etc., etc.

Thus, just as 'collage' has been used in contemporary visual art to capture the myriad components of life today, so collages of sounds and sonic textures generally have been used by musicians for the same reason. In music, textures are achieved by using voices, pitch complexes, sound samples and effects, and mixes of all of these to form 'soundscapes' of every description. From seminal works such as those of the Futurists and composers such as Stockhausen (e.g., Momente), to improvised or freely-notated sonic environments, we are in fact now very familiar with the presence of sonic textures in the musical forms or our time.

3. It has been a natural development, therefore, for computer systems to assist in the creation of sonic textures. Hardware and software 'samplers' have been important tools in this regard. CDP itself began in 1986 as a direct-to-disk system to avoid the limited RAM capacity of the first hardware samplers, and so the idea of sonic texture and soundscape has been a part of CDP from the first. CDP's powerful set of sound transformation tools have been designed to help 'develop' the sound material and integrate it into musical forms.

If the goal is to realise texture types with every pitch and timing fully-defined, then the composer can write the music out using standard notations, or make use of a sequencer &/or mixer (hardware or software). But to introduce the flexibility of random selection (within user-defined constraints) as well as fully defined chords or melodic components, and to mix and balance the relative presence of each of these in a musical texture, software of quite a different nature is required.

One solution is to make use of software for algorithmic composition. This requires considerable expertise in programming but is ultimately the most powerful approach that can be used. Trevor Wishart's CDP TEXTURE Program Set establishes a middle ground between conventional sequencer and algorithmic software. It rings the changes on combinations of different musical features by means of a set of pre-established algorithms. You can choose among the components, design some of them fully and set up constraints for random selections for others. The textures can therefore contain aspects of conventional notation, proportional (space = time) notation, and graphic-gestural shapes – the whole complex of computer-generated events is in fact best comprehended by the ear rather than the eye.

Thus the CDP Texture Set provides powerful semi-algorithmic tools with which to realise a variety of ordinary and extraordinary musical situations. Besides its ability to constrain multi-events to a pre-defined harmonic lattice, perhaps its most important musical feature is the ability to mix the random and the efined.

4. How can the computer help? While it is useful to bear in mind the standard texture constructs first listed, it is important to realise how the computer can make the realisation of textural configurations considerably freer, by balancing computer selection (random choices constrained within [time-varying] user-defined limits) and fully user-defined components. Thus the CDP Texture Set can for example randomise (within user-defined limits):

5. Thus with the CDP Texture Set, you can, for example (the list cannot possibly be complete!):

The Workshop will focus on 20 specific texture configurations in order to show how the software operates in a step-by-step manner, for these are the tools with which to realise the above list of possibilities, as well as many more which only you can dream up.

6. Sounds! There is another vitally important factor. The full potential of these semi-algorithmic mechanisms is realised when they are used as a design-framework on which to attach specially formed sounds, sounds transformed in order to take advantage of a specific framework shape. Our 21st example illustrates how, in a sense, new sounds are created by texturing compatible inputs.

You could also try running Example 20 using whirrdtgd.wav, which is provided with the Workshop. It is a transformed version of marimba.wav:

The result gives some limited insight into how the TEXTURE components can be designed to combine with a certain sound in order to derive new and different sonic complexes, or v.vs. When the sound is relatively unpitched in nature, pitch structures can be used to shape the result without pitch itself becoming too prominent. When the sound has a fair degree of pitch content, as does whirrdtgd.wav, you have to be careful about the use of pitch, as too much of it can make the result sound unnatural or trivial. You can see what I mean by trying it with some of the other Examples in the Workshop – but perhaps what one feels about the balance of 'sound' and 'pitch' is really a matter of personal preferences and taste.

The examples for the workshop remain somewhat more prosaic partly in order to illustrate the operation of the software with sufficient audible clarity, and partly to show how far one can go with pitch design with the TEXTURE software. But please bear in mind that the fun begins, as it were, when transformed sounds are matched to texture design-frameworks.

Pieces of the Jigsaw

The TEXTURE Set is so rich and diverse that it is like a big jigsaw puzzle. There are many different components to use and fit together. These are the main pieces of the puzzle:

Visualising the Outcome

The Texture Set lends itself to visual imagery. The multiplicity of parameters means that this must often be fairly abstract and diagrammatic, but nevertheless there is considerable scope for using the visual imagination. In particular, applying time-varying upper and lower pitch limits automatically results in a visual image. (These are the time-varying upper and lower pitch contours for Example 13.) Adding dots for changing note-event densities, inserting staves with (changing) chords, and using a center-right-left zig-zag for spatial movement, and you can begin to see how an image can be assembled.

Another factor is how the sonic character of the input sound is portrayed. Here we need to activate our aural imaginations. As we imagine the sound (or actually listen to the [transformed] source) we will get some kind of visual image based on its pitch trace, tonal qualities, volume and surface texture.

We can then superimpose the sonic image of the source onto the multi-event texture diagram and imagine the aural result when the source sound is spread, compacted, harmonised, panned etc. according to the shape(s) in the diagram.

Taking the time to activate the visual and aural imaginations and to see and listen internally would seem, therefore, to be an unusually appropriate activity when working with Texture.

A Useful Toolbox

A number of supplementary tools can facilitate using the Texture Set. If these are charts included with the CDP HTML reference documentation, it is recommended that you print them out and have them laminated so that they will be readily available and easy to handle. Some of the other tools mentioned are suggestions regarding an effective working method. The Texture Set is complex: experience shows that it is easy to lose the thread, and to come back to work which is ambiguous (you can't identify which parameter values gave which result).

Group Formations

Note-events organised into groups are handled by GROUPED and by the DECORATED set. Their primary characteristic is that the groups are shaped internally by random selections within user-defined limits. As the groups repeat, therefore, they will always be slightly different, due to the random choices being made.

In GROUPED the time interval between groups is set by the packing and scatter parameters, with the option to set regular time intervals with tgrid. In DECORATED, the time interval between groups is set by the times of the 'line' (Nodal Substructure).

The main internal shaping parameters within the groups are:

In the DECORATED set there are two additional controls for how the decoration relates to the linear node to which it is attached. The first of these relates to time and the second to position. The decoration can take place before (use PREDECOR) or after (use POSTDECOR) the node, or the node can occur in the midst of the decoration (use DECORATED). The centring parameter handles position, and there are 7 options. 1 places the decoration (on and) above the node. This makes the pattern of nodes – the horizontal 'line' – clearly audible. 0 centres the decoration on the nodes, which therefore get lost in the midst of the decoration, leading to a more fused texture.

I feel that GROUPED is covered sufficiently in the Reference Documentation and only have a POSTDECOR example below ( Example 17).

Rhythms Old & New

TEXTURE allows you to set precise times for the note-events of melodic figures ('motifs' or 'ornaments'). These are the 'old' rhythms. It also allows you to randomise event onsets in various ways. These are the 'new' rhythms, often essential when working with sound, when a more 'natural' feel is appropriate. Having both options is a key feature of the CDP TEXTURE software, allowing you to mix and balance the fully-defined and the random as appropriate for the compositional task in hand.

As a rule, it is useful to define rhythms at crotchet = 60, and then use a tempo control (mult) to make the figures faster or slower (or both). TIMED does not have a tempo control.

It is also useful to know that mult will change the speed of motifs or decorations, but does not affect the timings of the nodal substructure. This is both a limitation and an opportunity. The limitation is that you have to hand-calculate nodal starttimes if you have motifs which need to remain in sync at an altered tempo. You do this by multiplying the original nodal starttimes by the mult factor. The opportunity is that you can alter the temporal relationship of the figures to the nodes: a faster tempo for the figures may cause them to finish before the next node arrives, thus causing a gap; a slower tempo may cause them to overlap the next node, and therefore overlap the onset of the next motif. Our final example (Example 20) does this.

For convenience, here are a few timings for standard rhythmic figures, at crotchet = 60 = 1 sec.

Subtle Differences: Motifs, Decorations & Ornaments

The three program groups, MOTIFS, DECORATED and ORNATE at first seem to be virtually the same. There are in fact, subtle and important differences between them, which point towards a variety of musical applications.

MOTIFSdoes have fully defined motifs, but does not make use of a 'line'/'nodal substructure'.

DECORATED – does not have fully defined motifs, but does have a 'line'/'nodal substructure'.

ORNATEdoes have both a 'line'/'nodal substructure' and fully defined motifs.

Designing Decorations

DECORATIONS is really a form of GROUPED, with the difference that it applies the group formations to timed 'line' nodes. The first thing to think about is why you might want to do this. For example, is it to

One key here is to balance random selection from the pitch parameters (gpsize and gprange with the construction of the Harmonic Set, or choosing not to use a Harmonic Set. As there are no fully defined motifs or ornaments, the sonorous character of the Harmonic Set is important. For example, the Set can be designed using only one interval type, or a symmetric pitch configuration. More or less of it is used according to the range given to gprange.

Designing Motifs

Motifs are 'fully defined', meaning start time, pitch, velocity and duration. Durations may overlap start times, creating legato efffects, or end before the next start time, creating staccato effects. Velocity is important to create accents which will bring out the shape of the rhythms – without this, the rhythms run together and become imperceptible: think performance technique.

The next step is to think about whether you want the motifs to be attached at random to any pitch in the pitch range (Mode 5) or to be attached at random to any pitch in a defined Harmonic Set. The timed nodal substructure is not available here, so the placement of the motifs is relatively free. The only way to tie it down is to set the low and high pitch parameters to the same pitch. This forces the motifs to start only on this pitch, thus making it possible to create canons or phased arpeggios, for example. The motifs themselves need to be designed with these purposes in mind.

Consider, for example, creating motifs which use only one or two pitches. Thus they can begin on different pitches selected at random, creating a wash of rhythms without causing dissonant clashes, especially if attached to Harmonic Set pitches separated by intervals.

Collage effects increase when several motifs are employed, or several input sounds. The program selects from the motifs and/or soundfiles at random.

The timed motifs (TMOTIFS) option makes it possible to play with the rhythmic placement of the motifs. And the mult tempo control makes it possible to vary the speed of the motifs, thereby altering their relationship to the timed nodes: finishing sooner or overlapping.

Designing Ornaments

The ORNATE group is capable of some of the most fully user-specified textures possible with the TEXTURE software, as it puts together the (timed) Nodal Substructure available in DECORATED with the fully defined figures available in MOTIFS. The observations about both of these program groups therefore also applies to ORNATE. We need to look at both the highly specified designs and freedoms which are nevertheless still present.

The three examples below which use POSTORNATE (18-20) give an initial glimpse into what is possible. The first (18) develops a canonic idea with a very long 'ornament', really a whole melodic passage. This repeats on different start pitches as defined by the nodes, and virtually simultaneous start times create parallel intervals. However, we find that the transpositions in Mode 5 are exact (sometimes causing clashes), and those in the other Modes warp the ornament's intervals, as in Example 19. The technique of writing 'rounds' may have a role when designing canonic passages.

The third example (20) is closer to the concept of an 'ornament' as a short embellishing figure. One could keep the ornaments close to the line, maintaining the prominence of the line, by creating short melodic embellishments such as upper and lower neighbors, turns, slides and trills. Or one could go for a more textural effect by creating ornaments with a harmonic dimension, such as the tremolo figure used in Example 20. In this case, the design of the line is placed in service to the type of texture implied by the character of the ornament.

Combining Multi-Events & Rhythms

TIMED makes it possible to 'populate' a textural space with a rhythm, drawing the pitches for the repeating rhythmic figure at random from within the pitch range in Mode 5 or from the pitches of a user-defined Harmonic Set. The width of the pitch range is therefore an important factor. This can be dramatically extended by using Modes 1 or 2, which draw pitches from all octaves, not just those defined in the Harmonic Field portion of the note data file.

Combining Multi-events, Harmony, Motifs & Rhythm

There is not much else that can be done with TIMED, but a rhythmic template can be used with groups of note events (TGROUPED) and with motifs (TMOTIFS or TMOTIFSIN). TMOTIFSIN is quite a package, as it combines Rhythmic Template, (changing) Harmonic Set(s), and (a list of) fully defined motifs, not to mention the possibility of multiple input soundfiles. Given that there can be changing harmonic sets as well as time-varying outer pitch contours, chordal material can be laid down with clear rhythmic patterns, filled with melodic/rhythmic figures, and the whole complex gradually unveiled or covered up via the outer pitch contours.

Time spent planning textures which make use of so many features is therefore time well spent. The process of visualising the texture as described above is recommended, for the visual and aural imagination needs to ponder the characteristics of the whole before tackling the specific implementation.

Part II – Focused Examples
Illustrating a Variety of Musical Possibilities

For an overview or for comparative listening, you can Play all Sound Examples – & link from there to the Play Lists with Parameter Data, or back to this text.

For the practical work, you can work faster if you use the SoundShaper presets provided. To do this, load the Presets file txws.dat. Simlarly, copy the Sound Loom patches found in the \Txws\SLpatches subdirectory to your _cdpatch folder. Open the specified input soundfile(s) – or soundfile(s) of your choice – before going to the appropriate TEXTURE function and loading the preset/patch.

Please note that the Presets/Patches expect to find the files in c:\Txws\. This is why your installation should ensure that the files of the Texture Workshop are placed on Drive C: and in this directory name.

Section 1 – Examples 1 to 4, illustrating SAME PITCH: – repetitions on the same pitch; uses Mode 5 ('None'), with min & max pitch parameters both set to the same pitch. See txwspla1-PC.htm for the play list with all parameter values shown.

Example 1 – note events repeat on one pitch, with time-scatter and spatialisation

Play txws1.wav  Parameters: txwspla1-PC.htm  Preset: txwsex1

TEXTURE SIMPLE, Mode 5 We start simply! To achieve this result, we mainly need to provide a value for packing, add a bit of scatter, and give the same value to the lower and upper pitch parameters. To make it more interesting, spatial positioning is spread to the full limit.

Practical work: Alter the packing rate, either faster or slower, naming the output txws1b.wav. Do NOT make the packing value less than the value for scatter. Visualising the results can be a helpful exercise, in this case simply note events which are spaced closer or more widely.

Example 2 – regular repetitions of a rhythmic template on one pitch, with scatter and spatialisation

Play txws2.wav  Parameters: txwspla1-PC.htm  Preset: txwsex2

TEXTURE TIMED, Mode 5. The definition of a rhythmic figure is achieved by defining a rhythmic template in the note data file. In this case, the note data file txws2nd.txt reads:

60
#6

0.00 1 0 0 0
0.17 1 0 0 0
0.34 1 0 0 0
0.50 1 0 0 0
1.25 1 0 0 0
1.75 1 0 0 0

Practical work: Alter skiptime to 0.25, making the soundfile txws2b.wav. Notice how hard it is to pick out the rhythmic pattern, because the velocities are not (cannot) be adjusted. This can work to your advantage to create a rapid smooth flow. You might also try creating your own rhythmic template by editing the note data file.

Example 3 – regular repetitions of a melodic motif, each starting on the same pitch; overlap option

Play txws3.wav  Parameters: txwspla1-PC.htm  Preset: txwsex3

TEXTURE MOTIFS, Mode 5. The melodic motif is fully defined in the note data file, i.e., with pitch, velocity and duration. The pitch range in the parameters is again the same pitch (60), and this constrains the motif to begin each time on that same pitch, but does not inhibit the performance of the other pitches in the motif. The note data file txws3nd.txt reads:

60
#10 (Motif)

0.00 1 60 80 0.5
0.25 1 62 60 0.5
0.50 1 63 50 1.0
1.50 1 65 70 0.5
2.00 1 63 80 0.5
2.25 1 65 80 0.5
2.50 1 67 80 1.0
3.25 1 70 90 0.5
3.50 1 69 80 0.5
3.75 1 67 70 0.5

Practical work: Exploring the relationship of packing (time between repeats) and mult (tempo).

Example 4 – repetitions of a melodic motif, each starting on the same pitch, according to a rhythmic template

Play txws4.wav  Parameters: txwspla1-PC.htm  Preset: txwsex4

TEXTURE TMOTIFS, Mode 5. Here we need the rhythmic template and the motif definition in the note data file txws4nd.txt:

60
#5 (Rhythmic Template)

0.00 1 0 0 0
2.00 1 0 0 0
4.00 1 0 0 0
5.00 1 0 0 0
7.00 1 0 0 0

#5 (Motif)

0.00 1 60 88 0.5
0.25 1 62 84 0.5
0.50 1 63 76 0.5
0.75 1 65 72 0.5
1.00 1 67 66 2.0

Practical work: Try this rhythmic template: 0.0 2.0 4.0 4.5 5.0 5.5 6.0 by editing the original note data file for this example in a text editor and saving it as txws4n2.txt (you can't SAVE AS within SoundShaper). Having loaded the Preset, now put the cursor in the note data file parameter and click on 'Open' next to the box displaying the file. You can now select your new file and open it. Make txws4b.wav and see how the new template causes the figure to overlap.

Now alter mult low to 0.625 (for crotchet = 96), leaving mult high at 1 (crotchet = 60), making txws4c.wav. Now each repeat is at a different tempo and the overlaps go out of sync. A mult range of 0.5 to 2 throws it out completely.

A position of 0.5 and spread of 1 causes the motifs to move about in the horizontal field.


Section 2 – Examples 5 to 8, illustrating PITCH RANGE: – pitches selected at random from a pitch range; uses Mode 5, with minpitch set lower than maxpitch. See txwspla2-PC.htm for the play list with all parameter values shown.

Example 5 – note events drawn at random from a pitch range, with time-scatter and spatialisation

Play txws5.wav  Parameters: txwspla2-PC.htm  Preset: txwsex5

TEXTURE SIMPLE, Mode 5. The main task here is to establish a pitch range from within which the note events will be selected. This is done with the pitchlo - pitchhi parameters, which here are set to an 8ve range: 60 - 72. The note data file just requires the (possibly arbitrary) MIDI Pitch Value (MPV) as a pitch reference point. Thus txws5nd.txt contains only '60'.

Practical work: The randomness of the pitch selection is (deliberately) very apparent in our original example due to the wide pitch range. Make the pitch range very narrow (e.g., 60-64) and the packing rate very fast, creating txws5b.wav. You will need to edit the time-varying file txws5pk.brk to put in the smaller values, saving it to a new name. Then place the cursor in the packing parameter and click on 'Open' next to the box displaying the (old) file to open your new file. These changes will result in a dense, churning texture.

To explore the other end of the spectrum, make the pitch range very wide (e.g., 48-86) while making the packing very slow, creating txws5c.wav. Now the note-events are much separated in pitch and time. They call out for a much more distinctive sound, which could then 'ring out' with distinctive, transposed iterations.

Example 6 – regular repetitions of a rhythmic template on pitches drawn at random from a pitch range

Play txws6.wav  Parameters: txwspla2-PC.htm  Preset: txwsex6

TEXTURE TIMED, Mode 5. The definition of a rhythmic figure is achieved by defining a rhythmic template in the note data file. In this case, the note data file txws2nd.txt reads (same as Ex. 2):

60
#6 (Rhythmic Template)

0.00 1 0 0 0
0.17 1 0 0 0
0.34 1 0 0 0
0.50 1 0 0 0
1.25 1 0 0 0
1.75 1 0 0 0

This time the rhythmic template will be played on different pitches selected from within the range specified on the command line.

Practical work: This function can develop more than at first meets the eye. The pitch range could be much tighter or spread wide, as we tried with Example 5. This time, re-do the rhythmic template to be much livelier. Edit the rhythmic template in the note data file, save to a new name and, having loaded the preset for this example, open the new note data file, as described in Example 4.

Example 7 – regular repetitions of 3 melodic motifs, each starting on pitches selected at random from a pitch range

Play txws7.wav  Parameters: txwspla2-PC.htm  Preset: txwsex7

TEXTURE MOTIFS, Mode 5. 3 motifs are defined in txws7nd.txt, which reads:

60
#10 (Motif 1, C-B-F-G-C#-D Ab-F#-Bb-E)

0.00 1 60 84 0.3
0.17 1 71 70 0.3
0.34 1 65 74 0.3
0.50 1 67 56 0.3
0.67 1 61 74 0.3
0.84 1 62 60 0.3
1.00 1 68 84 0.3
1.17 1 66 60 0.3
1.34 1 70 66 0.3
1.50 1 64 78 0.3

#1 (Motif 2) (C)
0.00 1 60 60 2.0

#10 (Motif 3, C-Db-G-F-B-Bb E-F#-D-G# - Mirror Inversion of Motif 1)

0.00 1 60 84 0.3
0.17 1 49 70 0.3
0.34 1 55 74 0.3
0.50 1 53 56 0.3
0.67 1 59 74 0.3
0.84 1 58 60 0.3
1.00 1 52 84 0.3
1.17 1 54 60 0.3
1.34 1 50 66 0.3
1.50 1 56 78 0.3

Practical work: The example in its original form is solid and driving. We can loosen it up nicely by varying the tempo. Load the Preset and give mult a 0.5 (crotchet = 120) to 0.7 (crotchet = 84) range, creating txws7b.wav. Now each motif will be played at a tempo somewhere within this range, each one different. Note that the mult tempo parameters can also be time-varying!

You can take it one step further by adding a time-varying breakpoint file for the packing parameter. Think how you would like the texture to speed up or slow down over its 12 second output duration, write the times and values in txws7cpk.brk, open it in the parameter page and create txws7c.wav.

Example 8 – repetitions of a melodic motif, each starting on pitches drawn at random from a pitch range, according to a rhythmic template

Play txws8.wav  Parameters: txwspla2-PC.htm  Preset: txwsex8

TEXTURE TMOTIFS, Mode 5. The rhythmic template, designed to emphasise how overlapping figures can be created, and the motif are defined in txws8nd.txt, which reads:

60
#11 (Rhythmic Template)

 0.00 1 0 0 0
 1.00 1 0 0 0
 1.50 1 0 0 0
 6.00 1 0 0 0
 6.25 1 0 0 0
 6.50 1 0 0 0
 6.75 1 0 0 0
 7.00 1 0 0 0
 8.00 1 0 0 0
 9.00 1 0 0 0
11.00 1 0 0 0

#14 ('Trill' Motif)

0.00 1 60 84 0.4
0.10 1 61 80 0.4
0.20 1 60 76 0.4
0.30 1 61 72 0.4
0.40 1 60 68 0.4
0.50 1 61 76 0.4
0.60 1 60 72 0.4
0.70 1 61 68 0.4
0.80 1 60 64 0.4
0.90 1 61 60 0.4
1.00 1 60 84 0.5
1.25 1 63 72 0.5
1.50 1 66 78 0.5
1.75 1 64 72 0.5

Thus the relationship between the time between the time points of the rhythmic template and the length of the motif(s) is a crucial factor in the design of the texture – as is the design of the motif itself.

Practical work: A good way to illustrate the relationship between template and motif would be to alter the rhythmic template for this example so that it started slow, with gaps, and then gradually placed the onsets of the motifs closer and closer together. You can do this by editing txws8nd.txt, saving as txws8n2.txt and opening it on the parameter page as the note data file to be used, creating txws8b.wav.

The effect of this time-compression and increasing overlap can be intensified by making the pitch range time-varying, moving from wide to narrow. Here we need create and open breakpoint files for the low and high pitch parameters: txws8pl.brk and txws8ph.brk, creating txws8c.wav.


Section 3 – Examples 9 to 12, illustrating HARMONIC SET: – pitches selected at random from a user-defined harmonic set (uses only the named pitches); uses Mode 3, with minpitch set lower than maxpitch in the pitch parameter breakpoint files. See txwspla3-PC.htm for the play list with all parameter values shown.

Example 9 – note events with pitches drawn at random from a user-defined harmonic set, with scatter

Play txws9.wav  Parameters: txwspla3-PC.htm  Preset: txwsex9

TEXTURE SIMPLE, Mode 3. This differs from Example 5 in that the randomly selected pitches are now restricted to a chord/harmonic configuration. This is defined in txws9nd.txt as follows:

60
#5 (Harmonic Set - C-E-F#-A-C')

0 1 60 0 0
0 1 64 0 0
0 1 66 0 0
0 1 69 0 0
0 1 72 0 0

Practical work: This might be a good point at which to create your own harmonic set. Edit txws9nd.txt accordingly, saving as txws9n2.txt and opening it on the parameter page, creating txws9b.wav.

Another area to explore would be the way this simple process can create a harmonised sonic texture. A quick way to get an idea of what can happen would be to load whirrgdt.wav, a transformed version of marimba.wav supplied with the Workshop, and then the Preset for Example 9 and remake the sound, creating txws9c.wav. If this interests you, try creating your own sound, perhaps long with slow envelopes and with a reasably defined pitch. Remember to select 'whole input' (the -w flag so that your whole sound is used for every note-event in the texture.

Example 10 – regular repetitions of a rhythmic template on pitches drawn at random from a harmonic set

Play txws10.wav  Parameters: txwspla3-PC.htm  Preset: txwsex10

TEXTURE TIMED, Mode 3. The harmonic set defined in txws10nd.txt is the minor 7th chord C-Eb-G-Bb:

60
#6 (Rhythmic Template)

0.00 1 0 0 0
0.17 1 0 0 0
0.34 1 0 0 0
0.50 1 0 0 0
1.25 1 0 0 0
1.75 1 0 0 0

#4 (Harmonic Set - C-Eb-G-Bb)

0 1 60 0 0
0 1 63 0 0
0 1 67 0 0
0 1 70 0 0

Practical work: In Mode 3 ('Harmonic Set'), only the pitches defined in the note data file are used (as long as the low and high pitch parameters accomodate the range of this set). In all the examples, I use only the Harmonic Set Mode, so try remaking this example with Mode 1 selected ('Harmonic Field') and the pitch range extended to 36 - 96, creating txws10b.wav. Now the program selects the pitches defined in the note data file from any octave within the pitch range. The marimba occasionally sounds like a prepared piano!

Example 11 – regular repetitions of a melodic motif, each starting on pitches drawn at random from a harmonic set

Play txws11.wav  Parameters: txwspla3-PC.htm   Preset: txwsex11

TEXTURE MOTIFSIN, Mode 3. Here the motif in txws11nd.txt is designed so that it maintains the overall harmony as it transposes to the different pitches of the harmonic set and overlap in time. The note data file reads:

60
#11 (Harmonic Set - IIº Mode of Limited Transposition)

0.0 1 60 0 0
0.0 1 61 0 0
0.0 1 63 0 0
0.0 1 64 0 0
0.0 1 66 0 0
0.0 1 67 0 0
0.0 1 69 0 0
0.0 1 70 0 0
0.0 1 72 0 0
0.0 1 73 0 0
0.0 1 75 0 0

#10 (Motif - C-Db-Eb-F#-Eb-Db-C-Bb C-Db)

0.000 1 60 84 0.3
0.125 1 61 60 0.3
0.250 1 63 66 0.3
0.375 1 66 72 0.3
0.500 1 63 78 0.3
0.625 1 61 66 0.3
0.750 1 60 72 0.3
0.875 1 58 78 0.3
1.000 1 60 84 0.3
1.500 1 61 84 0.3

Practical work: As tried before, this texture can be 'loosened up' by providing a range for the mult tempo low-high parameters. Also, the 'wash' effect is increased when the sound has a softer attack transient – try using whirrgdt.wav as an input.

But to tie in with the discussion about the relationship of the motif to the harmonic set, let's see what happens when the harmonic set is restricted by making the parameter pitch range 62 - 67, a perfect fourth. Using the marimba sound as the input, load the Preset and later the pitch parameters, creating txws11b.wav. The results are remarkable.

Refer to TEXTURE ORNATE as a way to 'tie down' the motif start pitches to specific pitches of the harmonic set, i.e., those defined in the 'line' (nodal substructure) of the note data file.

Example 12 – repetitions of 4 melodic motifs, each starting on pitches drawn at random from a harmonic set, according to a rhythmic template; 2 soundfile inputs

Play txws12.wav  Parameters: txwspla3-PC.htm  Preset: txwsex12

TEXTURE TMOTIFSIN, Mode 3. The new source is gtrcdt.wav (a guitar sound, MPV 57), and marimba.wav is given as the second source. The note data file txws12nd.txt reads:

57 60
#9 (Rhythmic Template)

0.00 1 0 0 0
0.50 1 0 0 0
0.75 1 0 0 0
1.00 1 0 0 0
1.25 1 0 0 0
1.75 1 0 0 0
2.00 1 0 0 0
2.25 1 0 0 0
2.50 1 0 0 0

#8 (Harmonic Set)

0.0 1 60 0 0
0.0 1 61 0 0
0.0 1 63 0 0
0.0 1 66 0 0
0.0 1 67 0 0
0.0 1 69 0 0
0.0 1 70 0 0
0.0 1 72 0 0

#7 (Motif 1, C-Db-Eb-F#-G-A-Bb)

0.000 1 60 84 0.3
0.125 1 61 84 0.3
0.250 1 63 84 0.3
0.375 1 66 84 0.3
0.500 1 67 68 0.3
0.625 1 69 72 0.3
0.750 1 70 68 0.3

#11 (Motif 2, C'-A-Bb-A-G-F# G-A-G-F#-Eb)

0.000 1 72 80 0.3
0.250 1 69 78 0.3
0.500 1 70 78 0.3
0.625 1 69 78 0.3
0.750 1 67 78 0.3
0.875 1 66 78 0.3
1.000 1 67 78 0.3
1.125 1 69 68 0.3
1.250 1 67 72 0.3
1.375 1 66 68 0.3
1.500 1 63 76 0.3

#5 (Motif 3 - 'strummed' chord 1, C-Eb-F-A-C')

0.00 1 60 80 1.0
0.01 1 63 80 1.0
0.02 1 66 80 1.0
0.03 1 69 80 1.0
0.04 1 72 80 1.0

#5 (Motif 4 - 'strummed' chord 2, Db-F#-G-A-Bb)

0.00 1 61 80 1.0
0.01 1 66 80 1.0
0.02 1 67 80 1.0
0.03 1 69 80 1.0
0.04 1 70 80 1.0

Practical work: This time, edit the note data file txws12nd.txt replacing the first motif with one of your own design, saving as txws12n2.txt and creating txws12b.wav. This will raise the compositional issues of how to relate the motif to the harmonic set and what rhythm might be effective.

It would also be interesting to 'loosen up' this very rhythmic texture by providing a fairly wide tempo range, e.g., mult = 0.5 to 1.4.


Section 4 – Examples 13 to 16, illustrating TIME-VARYING Parameters: – e.g., pitches selected at random from a changing harmonic set, with time-varying pitch contours. Uses Mode 4 (changing harmonic fields or sets), with minpitch set lower than maxpitch in the pitch parameter breakpoint files. See txwspla4-PC.htm for the play list with all parameter values shown.

Example 13 – note events are selected at random from harmonic sets which change at a specified time-point, and which unfurl by means of time-varying pitch parameters.

Play txws13.wav  Parameters txwspla4-PC.htm  Preset: txwsex13

TEXTURE SIMPLE, Mode 4. Now, we not only select pitches from a harmonic set, but also change the set at the half-way point, thus introducing a chord change. These are defined txws13nd.txt

60
#10 (Changing Harmonic Set)

0 1 60 0 0 (1st chord, C-E-F#-A-C')
0 1 64 0 0
0 1 66 0 0
0 1 69 0 0
0 1 72 0 0
6.5 1 62 0 0 (2nd chord, D-E-F-A-B)
6.5 1 64 0 0
6.5 1 65 0 0
6.5 1 69 0 0
6.5 1 71 0 0

The number of parameters now altering over time in Example 13 is beginning to produce more supple and 'musical' results.

Practical work: We can explore the overlay of time-varying pitch parameters and the underlying harmonic set by reversing the way the contours work. In the original example, we start narrow and widen, so that the full harmonic set is being used at the point of changeover to the 2nd chord. Let's start wide and move to narrow at the time of changeover, then back to wide again. Thus the texture will focus in on a specific pitch, move to the next chord and open out again – like going through a gateway into a different landscape.

To do this, you can edit the existing pitch parameter breakpoint files txws13pl.brk and txws13ph.brk, saving to new names (e.g., txws13lp.brk and txws13hp.brk) and opening them as the files to be used as the new pitch parameters.

Example 14 – repetitions of a rhythmic template on pitches selected at random from a changing harmonic set, with time-varying pitch parameters

Play txws14.wav  Parameters: txwspla4-PC.htm  Preset: txwsex14

TEXTURE TIMED, Mode 4 A more complex rhythmic template is defined in txws14nd.txt, along with a changing harmonic set:

60
#28 (Rhythmic Template)

0.000 1 0 0 0 (First 3 beats)
0.170 1 0 0 0
0.340 1 0 0 0
0.500 1 0 0 0
0.670 1 0 0 0
0.840 1 0 0 0
1.000 1 0 0 0
2.000 1 0 0 0
2.125 1 0 0 0
2.250 1 0 0 0
2.375 1 0 0 0
2.500 1 0 0 0
2.625 1 0 0 0
2.750 1 0 0 0
2.875 1 0 0 0

3.00 1 0 0 0 (Last 2 beats)
3.34 1 0 0 0
3.67 1 0 0 0
4.00 1 0 0 0
4.10 1 0 0 0
4.20 1 0 0 0
4.30 1 0 0 0
4.40 1 0 0 0
4.50 1 0 0 0
4.60 1 0 0 0
4.70 1 0 0 0
4.80 1 0 0 0
4.90 1 0 0 0

#10 (Changing Harmonic Set)

0.0 1 60 0 0 (1st chord, C-Eb-F#-A-C')
0.0 1 63 0 0
0.0 1 66 0 0
0.0 1 69 0 0
0.0 1 72 0 0
6.5 1 58 0 0 (2nd chord, Bb-Eb-Gb-A-Db)
6.5 1 63 0 0
6.5 1 66 0 0
6.5 1 69 0 0
6.5 1 73 0 0

Although limited, a sharply defined and vigorous rhythmic template is possible, and the time-varying handling of changing harmonic configurations is available.

Practical work: Once again, let's remake this example using Mode 2 (changing harmonic field). We can replace the time-varying pitch parameter files with constants so that the pitch range will be 36 to 96. Ticking 'whole sound' (-w flag) ensures that we don't unnecessarily reduce the length of the sounds. Creating txws14b.wav and listening to the result, perhaps you will be enticed to try this again with different sound inputs.

Example 15 – repetition of 4 melodic motifs, each note of which randomly selects a pitch from within the changing harmonic set, within time-varying packing and a tempo range

Play txws15.wav  Parameters: txwspla4-PC.htm  Preset: txwsex15

TEXTURE MOTIFSIN, Mode 4. A changing harmonic set and 4 motifs are defined in txws15nd.txt, which reads:

60
#11 (Changing Harmonic Set)

0.0 1 60 0 0 (harmonic set, chord 1)
0.0 1 62 0 0
0.0 1 65 0 0
0.0 1 68 0 0
0.0 1 70 0 0
0.0 1 72 0 0
6.0 1 65 0 0 (harmonic set, chord 2)
6.0 1 66 0 0
6.0 1 68 0 0
6.0 1 70 0 0
6.0 1 71 0 0

#9 (Motif 1, C-D-AB-C'-Bb-F D-D-D)

0.00 1 60 80 0.6
0.17 1 62 72 0.5
0.34 1 68 78 0.6
0.50 1 72 76 0.5
0.67 1 70 78 0.6
0.84 1 65 72 0.5
1.00 1 62 80 0.8
1.34 1 62 76 0.8
1.67 1 62 72 0.8

#4 (Motif 2, C-D Ab-C')

0.00 1 60 80 1.0
0.50 1 62 80 1.0
1.00 1 68 80 1.0
1.50 1 72 80 1.0

#8 (Motif 3, C-Bb-C'-Bb-D Ab-Ab-Ab)

0.00 1 60 80 0.5
0.20 1 70 76 0.5
0.40 1 72 72 0.5
0.60 1 70 76 0.5
0.80 1 62 72 0.5
1.00 1 68 80 1.0
1.50 1 68 76 1.0
1.75 1 68 72 1.0

#11 (Motif 4, C-D-Ab C-Bb-Ab-F-Ab-F-D-C)

0.000 1 60 80 0.7
0.340 1 62 80 0.7
0.670 1 68 80 0.7
1.000 1 72 80 0.7
1.125 1 70 80 0.7
1.250 1 68 80 0.7
1.375 1 65 80 0.7
1.500 1 68 80 0.7
1.625 1 65 80 0.7
1.750 1 62 80 0.7
1.875 1 60 80 0.7

One could hope for a better warping of motif to harmonic set, but within its limitations, this function can produce some delightful results. Designing the motifs to work with the harmonic set is a key aspect of the compositional process.

Practical work: We have a changing harmonic field, time-varying packing, time-varying tempo and 4 motifs in this example. The result is nicely supple. Now make several new outputs, using exactly the same settings as in the Preset, saving to txws15b.wav etc. What we learn from this is how much the program is making random choices, for each output is quite different from the one before, mainly due to the choices about what tempo and which motif for each note event. Sometimes when you are making a texture and it doesn't seem to turn out quite right, try making it again – a fresh run with the same parameters may produce a useful result, closer to what you intended. These trials provide a good illustration of how the Texture software incorporates and can balance the random and the defined.

Example 16 – repetitions of 2 melodic motifs, each starting on pitches drawn at random from a changing harmonic set, with time-varying pitch range contours, according to a rhythmic template

Play txws16.wav  Parameters: txwspla4-PC.htm  Preset: txwsex16

TEXTURE TMOTIFSIN, Mode 4. Guitar [MPV-57] and marimba [MPV-60] are the two inputs into a changing harmonic set with 4 motifs on a rhythmic template. The rhythmic template is a 'hemiola' (sextuplet [in two groups of 3] followed by a triplet over the same time). All this is defined in the note data file txws16nd.txt, which reads:

57 60 (MPV's for two soundfiles)
#9 ('hemiola' rhythmic template)

0.00 1 0 0 0
0.17 1 0 0 0
0.34 1 0 0 0
0.50 1 0 0 0
0.67 1 0 0 0
0.84 1 0 0 0
1.00 1 0 0 0
1.34 1 0 0 0
1.67 1 0 0 0

#10 (Changing Harmonic Set)

0.0 1 60 0 0 (Harmonic Set, chord 1, C-Eb-F#-A-C)
0.0 1 63 0 0
0.0 1 66 0 0
0.0 1 69 0 0
0.0 1 72 0 0
10.0 1 60 0 0 (Harmonic Set, chord 2, C-Db-F#-G-B)
10.0 1 61 0 0
10.0 1 66 0 0
10.0 1 67 0 0
10.0 1 71 0 0

#5 (Strummed chord 1, C-Eb-F#-A-C)

0.00 1 60 80 1.0
0.01 1 63 80 1.0
0.02 1 66 80 1.0
0.03 1 69 80 1.0
0.04 1 72 80 1.0

#5 (Strummed chord 2, C-Db-F#-G-B)


0.00 1 60 80 1.0
0.01 1 61 80 1.0
0.02 1 66 80 1.0
0.03 1 67 80 1.0
0.04 1 71 80 1.0

#14 (Melodic figure 1)

0.000 1 69 80 0.3
0.125 1 72 78 0.3
0.250 1 69 78 0.3
0.375 1 72 80 0.3
0.500 1 69 78 0.3
0.625 1 66 78 0.3
0.750 1 63 78 0.3
0.875 1 60 78 0.3
1.000 1 63 84 0.3
1.170 1 66 72 0.3
1.340 1 69 80 0.3
1.500 1 66 72 0.3
1.670 1 69 80 0.3
1.840 1 72 76 0.3

#14 (Melodic figure 2)

0.000 1 67 80 0.3
0.125 1 71 78 0.3
0.250 1 67 78 0.3
0.375 1 71 80 0.3
0.500 1 67 78 0.3
0.625 1 66 78 0.3
0.750 1 61 78 0.3
0.875 1 60 78 0.3
1.000 1 61 84 0.3
1.170 1 66 72 0.3
1.340 1 67 80 0.3
1.500 1 66 72 0.3
1.670 1 67 80 0.3
1.840 1 71 76 0.3

Although much in this texture is very precisely defined, the randomised motif selection and time-varying pitch contours make the result unpredictably complex and wonderfully fluid.

Practical work: Variants on this example will yield much information. For example, if you keep the lower pitch range constant at Middle C (60) and allow the upper one to move gradually from Middle C upwards to B (71), you do hear the texture opening out, but also notice that all (most?) of the pitches of the strummed chords are heard whenever they play. Make a new upper pitch range breakpoint file (e.g., txws16hp.brk), set the lower one to 60 and create txws16b.wav.

Why should some motifs be restricted by the outer pitch contour and others not? Remember that each motif will begin on a pitch randomly selected from the harmonic set. The strummed chords play because their bottom pitch is always within the range. Compare the result you get when the pitch range is 66-67, creating txws16c.wav.

Also instructive is what happens when the rhythmic template uses a 'Latin' rhythm which has longer note values, such as a samba or rumba. Having altered txws16nd.txt to form txws16n2.txt, create txws16d.wav with your own rhythmic template. Remember to adjust the number of note events in the template and the skiptime parameter – and possibly the tempo as well.

Finally, alter the Harmonic Set so that it consists of 2 dyads: 63-66 for the first 10 seconds, and 67-71 for the second 10 seconds, creating txws16e.wav. Now the full motifs are no longer played. The strummed chords have fewer notes and the other motifs play with many repeated notes, as the motifs are warped to the pitches defined in the Harmonic Set.


Section 5 – Examples 17 - 20, illustrating NODAL SUBSTRUCTURE: – attaching a decoration or an ornament (or several selected at random) to a 'line' (nodal substructure), with or without a harmonic field or set. See txwspla5-PC.htm for the play list with all parameter values shown.

Example 17 – repetitions of a (timed) nodal substructure, on the pitches (and at the times) of which 'decorations' are attached, decorations made up of note-events randomly shaped within the constraints of group parameters.

Play txws17.wav  Parameters: txwspla5-PC.htm  Preset: txwsex17

TEXTURE POSTDECOR, Mode 5. The Nodal Substructure and a Harmonic Set are defined in txws17nd.txt, which reads:

60
#13 (Nodal Substructure)

0.0 1 66 0 0
1.0 1 72 0 0
2.0 1 67 0 0 (Crotchets: F# C' G C)
3.0 1 60 0 0
4.0 1 66 0 0 (Quaver Triplet: F# G F)
4.34 1 67 0 0
4.67 1 65 0 0
6.00 1 66 0 0 (Semiquavers: F# G F F#)
6.25 1 67 0 0
6.50 1 65 0 0
6.75 1 66 0 0
8.00 1 60 0 0 (Dotted crotchet: C)
9.50 1 66 0 0 (Dotted crotchet, with skiptime: F#)

#13 (Harmonic Set)

0.0 1 60 0 0
0.0 1 62 0 0
0.0 1 63 0 0
0.0 1 64 0 0
0.0 1 65 0 0
0.0 1 66 0 0
0.0 1 67 0 0
0.0 1 69 0 0
0.0 1 70 0 0
0.0 1 72 0 0
0.0 1 74 0 0
0.0 1 75 0 0
0.0 1 76 0 0

Practical work: The motifs are created on the fly according to the ranges given for the 'group' parameters: size, packing and range. The potential of this function can be explored by using more extreme values and doing so in a time-varying manner. Therefore, write a time-varying breakpoint file for the low and high limits of each of these parameters, introducing considerably more contrast. When doing this, note the timing of events in the rhythmic template. Then load the Preset and then your new breakpoint files, creating txws17b.wav. If you compare this result with the original, you should find that your time-varying decorations are considerably more interesting and supple.

Example 18 – repetitions of a (timed) nodal substructure, on the pitches (and at the times) of which a long fully defined 'ornament' is attached, transposing with intervallic exactness.

Play txws18.wav  Parameters: txwspla5-PC.htm

TEXTURE POSTORNATE, Mode 5. The note data file txws18nd.txt contains only the Nodal Substructure and a rather extended 'ornament'. In earlier examples I was stressing the need to use the same pitch as min and max values so that the motif would repeat canonically with the same start pitch. Here I am using the Nodal Substructure mechanism to achieve canonic repeats with transpositions.

57 60
#6 (Nodal Substructure)

0.000 1 60 0 0
4.980 1 60 0 0
4.981 1 63 0 0
9.960 1 63 0 0
14.940 1 66 0 0
14.941 1 60 0 0

#38 (Extended Melody 'Ornament')

0.00 1 60 80 1.0
0.50 1 60 80 2.0
1.50 1 62 80 1.0
2.00 1 63 80 0.4
2.25 1 65 72 0.4
2.50 1 63 76 0.4
2.75 1 62 72 0.4

3.00 1 60 80 0.4
3.25 1 60 76 0.6
3.50 1 60 80 2.0
4.50 1 62 80 0.4
4.75 1 62 76 0.4
5.00 1 63 80 0.4
5.25 1 65 72 0.4
5.50 1 62 76 0.4
5.75 1 63 72 0.4

6.00 1 60 80 2.1
7.00 1 58 80 2.1
8.00 1 57 73 0.4
8.25 1 58 74 0.4
8.75 1 60 76 0.4

9.000 1 62 77 0.4
9.125 1 65 78 0.4
9.250 1 63 79 0.4
9.375 1 67 80 0.4
9.500 1 69 81 0.3
9.600 1 72 82 0.3
9.700 1 70 83 0.3
9.800 1 74 84 0.3
9.900 1 75 75 0.3
10.00 1 74 88 0.3
10.17 1 72 87 0.3
10.34 1 70 86 0.3
10.50 1 66 86 0.4
10.75 1 66 83 0.4
11.00 1 66 88 1.0
11.50 1 66 84 0.4
11.75 1 66 80 0.4

Practical work: May I suggest working on a 'round' as a way of getting deeper into the potential of this program. Once you can do something as controlled as that, it is easier to start to loosen it up. Here are some pointers:

Example 19 – repetitions of a (timed) nodal substructure, on the pitches (and at the times) of which a long fully defined 'ornament' is attached, with intervals constrained (warped to) a Harmonic Set when tranposed.

Play txws19.wav  Parameters: txwspla5-PC.htm

TEXTURE POSTORNATE, Mode 3. The note data file txws19nd.txt just adds a harmonic field to the file used for Example 18. It reads:

57 60
#6 (Nodal Substructure, same as Ex. 18)

0.000 1 60 0 0
4.980 1 60 0 0
4.981 1 63 0 0
9.960 1 63 0 0
14.940 1 66 0 0
14.941 1 60 0 0

#7 (Harmonic Set: D-E-F-G-B-C#-D)

0.0 1 62 0 0
0.0 1 64 0 0
0.0 1 65 0 0
0.0 1 67 0 0
0.0 1 71 0 0
0.0 1 73 0 0
0.0 1 74 0 0

#38 (Extended Melody 'Ornament', same as Ex. 18)

0.00 1 60 80 1.0
0.50 1 60 80 2.0
1.50 1 62 80 1.0
2.00 1 63 80 0.4
2.25 1 65 72 0.4
2.50 1 63 76 0.4
2.75 1 62 72 0.4

3.00 1 60 80 0.4
3.25 1 60 76 0.6
3.50 1 60 80 2.0
4.50 1 62 80 0.4
4.75 1 62 76 0.4
5.00 1 63 80 0.4
5.25 1 65 72 0.4
5.50 1 62 76 0.4
5.75 1 63 72 0.4

6.00 1 60 80 2.1
7.00 1 58 80 2.1
8.00 1 57 73 0.4
8.25 1 58 74 0.4
8.75 1 60 76 0.4

9.000 1 62 77 0.4
9.125 1 65 78 0.4
9.250 1 63 79 0.4
9.375 1 67 80 0.4
9.500 1 69 81 0.3
9.600 1 72 82 0.3
9.700 1 70 83 0.3
9.800 1 74 84 0.3
9.900 1 75 75 0.3
10.00 1 74 88 0.3
10.17 1 72 87 0.3
10.34 1 70 86 0.3
10.50 1 66 86 0.4
10.75 1 66 83 0.4
11.00 1 66 88 1.0
11.50 1 66 84 0.4
11.75 1 66 80 0.4

Practical work: This example illustrates the warping effect of the Harmonic Set on an ornament, providing new, serendipitous intervallic relationships. To explore this further try realising a familiar round with each phrase as a separate ornament.

The problem with separating the tune into different ornaments is that either they all start on the same pitch (all nodes are at the same pitch) or any ornament (selected at random) could start on any pitch of the chord (the nodes use the different pitches of the chord). This poses a challenging compositional problem! It may be resolved by designing the phrases within a harmonic field concept (not just a chord & scale concept) and then specifying that field as a Harmonic Set (Mode 3).

NB: All ornaments start on the pitch of the specified nodal time point, even if the ornament itself starts on a different pitch. Sometimes it is useful to define all the ornaments as starting on the same pitch and then use the same or different pitches for the nodal time points.

Example 20 – repetitions of a (timed) nodal substructure, on the pitches (and at the times) of which fully defined 'ornaments' are attached, constrained to a Harmonic Set, with overlap of ornament and nodal substructure and time-varying tempo.

Play txws20.wav  Parameters: txwspla5-PC.htm
(Play txws20b.wav  uses whirrdtgd.wav as the input soundfile.)

TEXTURE POSTORNATE, Mode 3. The note data file txws20nd.txt with Nodal Substructure, Harmonic Set and Ornament reads:

60
#17 (Nodal Substructure)

0.00 1 67 0 0
1.50 1 70 0 0
2.50 1 67 0 0
4.00 1 70 0 0
4.34 1 69 0 0
4.67 1 67 0 0
5.00 1 67 0 0
6.00 1 72 0 0
7.00 1 72 0 0
7.25 1 70 0 0
7.50 1 69 0 0
7.75 1 67 0 0
8.00 1 66 0 0
8.25 1 65 0 0
8.50 1 63 0 0
8.75 1 62 0 0
9.00 1 60 0 0

#13 (Harmonic Set)

0.0 1 57 0 0
0.0 1 58 0 0
0.0 1 60 0 0
0.0 1 62 0 0
0.0 1 63 0 0
0.0 1 65 0 0
0.0 1 66 0 0
0.0 1 67 0 0
0.0 1 69 0 0
0.0 1 70 0 0
0.0 1 72 0 0
0.0 1 74 0 0
0.0 1 75 0 0

#15 ('Tremolo' ornament)

0.0 1 69 80 0.3
0.1 1 72 76 0.3
0.2 1 69 78 0.3
0.3 1 72 76 0.3
0.4 1 69 76 0.3
0.5 1 70 78 0.3
0.6 1 69 76 0.3
0.7 1 72 78 0.3
0.8 1 69 76 0.3
0.9 1 70 76 0.3
1.0 1 72 80 0.3
1.1 1 69 76 0.3
1.2 1 70 78 0.3
1.3 1 69 76 0.3
1.4 1 72 76 0.3

Practical work: It is important to consider the relationship of the design of the figure and that of the nodal substructure – not to mention that of the input sound itself. It is possible to create all manner of clearly shaped textural effects in which the figure is a contributor but is no longer heard as such.

Explore this with:

The mult tempo parameter is a very important tool. Making use of several ornaments – which will be selected at random – and supplying multiple input soundfiles expand the complexity and richness of the textures which can be created. Try this with the 3rd recommended exercise above (with the granulated sound).


Example 21 – a transformed sound is matched to a texture framework design.

Play txws21.wav  txwsplay-PC.htm

TEXTURE SIMPLE, Mode 3. The note data file ndftz5s2.txt with a 3-note Harmonic Set reads:

67
#3 (Harmonic Set)

0 1 67 0 0
0 1 64 0 0
0 1 62 0 0

This example is taken from my composition Crossing the Dark Rift (2002), Section 2 of Tzolkin cycle 5. A visual image of obliquely descending parallel squiggly lines was the starting point. A sequence of sound transformations created the squiggly lines, and then TEXTURE SIMPLE put them together.

This is a good example of how a sound is adapted in order to help realise the intended shape.

The input soundfile was pre-prepared so that it was long and wiggly. Some 'waver' was added to the original 5 sec Tibetan bowl sound, then it was stretched 2.3 times and subjected to a time-varying inharmonic shift by which its frequencies were first pushed very high and then gradually brought down to near its original pitch; finally, time varying vibrato was added, moving from 1 to 30 over 11 sec. This is the result: bdtg5wavrstrx2&3ssvib.wav.

The TEXTURE program is now employed to pack these descending streams close together. The Harmonic Set is G-F-D, and the pitch range is set to 62-67 to keep the output within this range. The note events are selected at random from the Harmonci Set and packed at 0.1 seconds over the 10 sec. output duration, with a bit of scatter. The whole soundfile is used. The onsets end at 10 sec., and then the whole soundfile note-events play out for another 14 seconds.

The combination of the complex source, high density, chord formation due to the tightly overlapping note-events, and use of the entire length of the input soundfile for every note-event serves to create a very dense texture. The texture descends because the sound itself does so. The descent could have been achieved by giving changing times to the Harmonic Set, but this would have made the texture thinner (no chord).

Last updated: 16 November 2021


©2002-2021 Archer Endrich, HITHER GATE MUSIC, Plymouth, Devon
PL4 7NL
archerhgm@gmail.com