This piece is derived almost entirely from the voice of my daughter, Ysolde Megumi Okumura-Jones, at age 12 months. In its original form, the piece was written for 5 channel surround. This present version is a 2 channel stereo reduction.

The piece is not just about cows — my daughter shared the typical toddler fascination for many different types of animal, and especially the noises they make. She has also always been fascinated by language, and quite happily creates her own. Even now, at age 5, she bewilders Japanese people by speaking in a language of her own invention, insisting that she is speaking Japanese!

A range of techniques were applied to the voice to achieve different types of transformation — speed/pitch changing (through sample rate conversion), application of reverb, reversing, level matching, and so on. A particularly important textural tool has been the application of granular synthesis through the use of the program Thonk. This program applies preset algorithms with randomised variables to alter grain size and density, transposition, panning (left-right), location of grains from the source file, length of an outputted ‘piece’, as well as speed and direction of movement through the source file. The different algorithms offered are not user-alterable, although a number of algorithms with quite different results are supplied. Thonk was applied to a select number of particular samples which then serve as marker points in the piece.

Altogether, 200 voice samples form the core of the source material, and the first transformation used is to multiply the sample rate by 1.5 (from 44,100 to 66,150) to achieve a drop in pitch of a perfect 4th. The effect on the voice of a 12-month-old is quite disturbing! Listen out for the perfect cadence at the end of the piece, engineered again by the manipulation of sample rate.