Key Features of the CDP TEXTURE set
- TEXTURE creates multi-event textures by repeating the input soundfile at specified times and pitches, along with many other ways to shape the result.
- TEXTURE can select pitches at random within specified limits, OR 'snap' them to user-defined harmonic grids.
- TEXTURE can make use of rhythmic motifs and melodic figures with user-defined pitch, amplitude, duration and onset times.
- Most parameters, including the harmonic grids, can change over time.
A variety of musical results possible with TEXTURE
- textural gestures with time-shaped edges and density control, with pitches chosen at random from a pitch range or snapped to a harmonic grid (see our soundbyte below)
- groups of textural gestures (as in the previous type) spaced in a fixed or time-varying way
- figures or groups of defined melodic/rhythmic figures which begin at times predefined by a timing grid
- a repeating rhythm articulated with notes chosen at random or from a defined harmonic grid
- a melodic line in which each note is textured with overlapping note-events
- a melodic line with one or more fully defined musical figures attached to its various notes, motifs chosen from a random pitch range or from a harmonic grid
- fully defined musical figures attached to a random pitch range or a harmonic grid, without an underlying melodic line
- figures whose pitch-relationships are warped in the process of making all their pitches fit a predefined harmonic grid
- canonic processes made from the repetitions of extended melodic phrases.
Other forms of texturing in CDP
- GrainMill Similar results can be achieved with GrainMill, but there are various important differences. The first of these is that TEXTURE plays its input file from its beginning at the onset of every note event, then playing either all or part of it as instructed by the user. GrainMill on the other hand 'granularises' the input, turning it into grains as it gradually moves through the input soundfile from start to finish. Thus, for example, a timestretched sound with significant timbral changes (such as a morphed sound) will make these changes audible in a slow-motion granularised texture. Other differences are that TEXTURE is able to lock the events onto harmonic grids (as in our example), play with defined melodic and rhythmic motifs, and accept multiple soundfile inputs.
- In the wavecycle DISTORTION set, note that DISTORT REPEAT both extends and textures the original source sound.
- The whole GRAIN set as well as forms of EXTENSION such as EXTEND LOOP and EXTEND DRUNK (see 'Segmentation') also texture the input sound.
- One more example to mention here is time-varying vibrato applied to the input sound (as analysis data).
General observations
- My view as a composer is that the TEXTURE program suite occupies important middle ground between creating sounds and creating scores. Furthermore, it does so in a semi-algorithmic way, so that a whole spectrum of possibilities from the very random to the completely defined is available. I say 'semi-algorithmic' because it doesn't use a scripting language. However, the presence of a note data file, parameter ranges and comprehensive breakpoint file options provide tremendous tools with which to shape the algorithms the programs use to create the note-event-based sound textures.
- The 'use whole duration of input sound' option is very important, opening up the possibility of cumulative sonic effects which are greater than the 'sum of the parts' as it were. Pre-processing the input sound with a given cumulative sonic effect in mind is the key. For example, imagine preparing a sound which is a long strand with a gradually increasing amount of vibrato and a pitch which wavers slightly in a time-varying way. Now imagine creating a texture with this input, using the whole duration, in which these strands are gradually packed more densely in pitch and onset time. The mixing of all these strands will potentially become a seething, turbulent flow ...
Last updated: 4 December 2003